JP2020180844A - Data processing device and data monitoring system - Google Patents

Abstract

To provide a data processing device and a data monitoring system which reduces the possibility that in monitoring measured data per cycle, measured data which is in effect normal is determined as abnormal.SOLUTION: Provided are a data processing device for monitoring measured data and a data monitoring system that includes this data processing device, the data processing device being constituted so that a condition setting unit 71 sets one or more data sorting reference regions, and a feature point, a sorting level and a sorting range in the data sorting reference region; a first processing unit 62 sorts measured data on the basis of the data sorting reference region and creates first measured data after processing; a second processing unit 64 selects measured data from measured data for multiple cycles preserved in a first storage unit 61 on the basis of the first measured data after processing and creates second measured data after processing; a monitoring waveform creation unit 66 creates a second upper-limit waveform and one or more second lower-limit waveforms on the basis of the second measured data after processing.SELECTED DRAWING: Figure 1

Description

The present invention relates to a data processing device and a data monitoring system, and more particularly, a data processing device and data suitable for monitoring a time-varying amount of measurement data so as to repeat one cycle from a start time point to an end point point a plurality of times. Regarding the monitoring system.

Conventionally, for an amount that fluctuates with time so as to repeat one cycle from the start time to the end time a plurality of times, an upper limit waveform and a lower limit waveform for the time of the one cycle are set, and the upper limit waveform and the lower limit waveform are used. An apparatus for monitoring the measurement data of each cycle based on the comparison with the measurement data of each cycle has been proposed (see, for example, Patent Document 1).

For example, FIG. 28 is a graph showing an example of the measurement data of the time fluctuation amount as described above as three-dimensional data of (x, y, z). Here, (x, y, z) corresponds to (time, measured value, number of cycles). In FIG. 28, the measurement data is drawn by superimposing the measurement value waveforms over one cycle along the plane of time (x) -measured value (y) in the cycle number (z) direction.

Japanese Unexamined Patent Publication No. 06-201398

Conventionally, when the measurement data as shown in FIG. 28 is monitored based on the upper limit waveform and the lower limit waveform over one cycle, the range defined by the upper limit waveform and the lower limit waveform is typically obtained. The measurement data that deviates from the above is determined to be abnormal.
The conventional monitoring of measurement data has the following problems. FIGS. 29 (a) and 29 (b) show, in the machining process of the workpiece in the machine tool, some measurement data, the mean value waveform (m) and the positive standard deviation waveform (+ 1σ) of all the measurement data, respectively. It is a graph which is superimposed on the two-dimensional plane of time (x) -measured value (y) together with the negative side standard deviation waveform (-1σ). Here, the mean value waveform is a waveform obtained by connecting the mean values over a plurality of cycles at each time point in one cycle, and the positive standard deviation waveform (+ 1σ) is the mean value over a plurality of cycles at each time point in one cycle. The negative standard deviation waveform (-1σ), which is a series of values obtained by adding the standard deviations at time points, is a series of values obtained by subtracting the standard deviation at that time point from the mean value over multiple cycles at each time point in one cycle. It is a waveform.

For example, each measurement data shown in FIG. 29 (a) exceeds the value of the positive standard deviation waveform (+ 1σ) in the part A, and each measurement data shown in FIG. 29 (b) is positive in the part B. It can be seen that the value of the side standard deviation waveform (+ 1σ) is exceeded. Therefore, these measurement data are highly likely to be determined to be abnormal in the monitoring of the measurement data using the upper limit waveform and the lower limit waveform. However, the measurement data shown in FIGS. 29 (a) and 29 (b) actually show that there is no problem in the quality of the workpiece processed in the processing process from which the measurement data was acquired after processing. In that sense, it is the measurement data that should be judged to be normal even in the monitoring of the measurement data for each cycle. As can be seen from FIG. 29, the factor that increases the possibility that such measurement data is determined to be abnormal is the deviation of the measurement data in the time (x) direction (that is, in FIG. 29 (a), the machining start time is usually It was earlier, and in FIG. 29 (b), the processing start time was later than usual).
Such variations at the start of machining can occur due to variations in the characteristics of the machine tool and / or the workpiece even when the machining and measurement conditions are constant. For example, the variation of the machine tool includes the variation (expansion / reduction) of the dimensions of the components of the machine tool due to the fluctuation of the temperature, and the variation of the workpiece includes the variation of the size, hardness and the like. ..

If there is measurement data that is determined to be abnormal due to such factors at the actual measurement site, even if it is determined to be abnormal in the monitoring of the measurement data for each cycle, the processing related to the measurement data is immediately performed. It is necessary to perform some other inspection on the work piece to confirm whether it is really bad, instead of making the work piece processed in the process defective. Therefore, if such a situation occurs frequently, there is a problem that the production efficiency of the product related to the processing process in this machine tool is lowered.
Further, conventionally, the upper limit waveform and the lower limit waveform are waveforms obtained by statistically processing a plurality of cycles of measurement data, such as the positive standard deviation waveform and the negative standard deviation waveform as described above, respectively. Is often used. Such a waveform has an advantage that, for example, even if a small number of abnormal data is included in the measurement data of a plurality of cycles used for calculation, its influence is unlikely to appear on the obtained upper and lower value waveforms and lower limit value waveforms. However, this feature, on the other hand, makes it possible to obtain upper limit waveforms and lower limit waveforms even when a plurality of measurement data include measurement data that is substantially normal with only a time lag. , It leads to the fact that the existence of such data is difficult to be reflected, and it is a factor of the above-mentioned problem.

An object of the present invention is to reduce the possibility that the measurement data, which is substantially normal even if there is a time lag in the monitoring of the measurement data for each cycle, is determined to be abnormal, and by extension, the operation of the measurement target. It is to provide a data processing apparatus and a data monitoring system capable of improving the efficiency of the above.

The following aspects of the invention exemplify the configurations of the present invention and will be described separately in order to facilitate understanding of the various configurations of the present invention. Each section does not limit the technical scope of the present invention, and while taking into consideration the best mode for carrying out the invention, some of the components of each section are replaced, deleted, or the like. Those to which the above-mentioned components are added can also be included in the technical scope of the present invention.

(1) A data processing device for monitoring an amount of measurement data that fluctuates with time so as to repeat one cycle from the start time to the end time, and is a first processing unit, a second processing unit, a condition setting unit, and monitoring. It includes a waveform creation unit, a first storage unit, a second storage unit, and a third storage unit.
The first storage unit stores measurement data of a plurality of cycles and stores the measurement data.
The first processing unit creates a first upper limit value waveform and a first lower limit value waveform in a time range of one cycle based on the measurement data of a plurality of cycles stored in the first storage unit.
The condition setting unit sets one or more data sorting reference areas that are a part of the two-dimensional area in a two-dimensional area composed of a time range of one cycle and the entire range of measured values, and the data. In each of the sorting reference areas, one time point in the data reference area is set as a feature point, one value of the measured value is set as a sorting level, and the data sorting reference area and the sorting level are set according to the above. Set the sorting range as the time range including the feature points,
The first processing unit further, for each of the one or more data sorting reference regions, is within the range of the sorting range before the feature point from the measurement data of a plurality of cycles stored in the first storage unit. The first front measurement data in which the intersection of the waveform corresponding to the measurement data and the sorting level exists, and the intersection of the waveform corresponding to the measurement data and the sorting level within the range after the feature point in the sorting range. The first rear side measurement data in which the data is present, and the intersection of the waveform corresponding to the measurement data and the sorting level exists on both sides of the sorting range both before and after the feature point. The measurement data is selected, and the selected first front side measurement data, the first rear side measurement data, and the first both side measurement data are stored in the second storage unit.
The first processing unit further removes the measurement data determined to be abnormal based on the first upper limit value waveform and the first lower limit value waveform from the measurement data stored in the second storage unit, and the rest. A first post-processing measurement data group consisting of the measurement data of the above is created, and the first post-processing measurement data group is stored in the second storage unit.
The second processing unit selects measurement data from the measurement data of a plurality of cycles stored in the first storage unit based on the measurement data group after the first processing, and the second processing unit is composed of the selected measurement data. A post-processing measurement data group is created, and the second post-processing measurement data group is stored in the third storage unit.
The monitoring waveform creation unit is based on the second post-processing measurement data group in order to monitor the measurement data acquired after the period in which the plurality of measurement data stored in the first storage unit is acquired. A data processing apparatus characterized in that it is configured to create one or more second upper limit waveforms and one or more second lower limit waveforms in a time range of one cycle.

(2) In the data processing apparatus according to item (1), an input unit that receives measurement data stored in the first storage unit from the outside, and the second upper limit value waveform and the second lower limit value waveform to the outside. A data processing device including an output unit to be transmitted.

(3) In the data processing apparatus according to item (2), each of the measurement data received from the input unit is stored in the first storage unit together with date and time information indicating the date and time when the measurement data was received by the input unit. Saved
The first processing unit obtains the mean value and standard deviation of the measured values of the measurement data of the plurality of cycles stored in the first storage unit at each time point over the plurality of cycles at a plurality of time points within the time range of one cycle. The first mean waveform, which is calculated and the calculated mean values of multiple time points, is added to the mean value of each of the multiple time points plus the default positive multiple of the standard deviation of the time point. Create a first positive standard deviation waveform and a first negative standard deviation waveform that is a series of values obtained by subtracting the default positive multiple of the standard deviation at the relevant time point from the mean value of each of the multiple time points. ,
For each of the one or more candidate ranges set within the time range of one cycle, the condition setting unit has a fluctuation start time point in which the time fluctuation of the first mean value waveform satisfies a predetermined fluctuation start condition, and the first. The time variation of the mean value waveform determines the fluctuation end time point that satisfies a predetermined fluctuation end time condition, and the fluctuation start time point and the fluctuation end time point, the first mean value waveform, the first positive standard deviation waveform, and The data sorting reference region is set based on at least one of the first negative standard deviation waveforms, and the data sorting reference region spans a plurality of cycles at each time point from the fluctuation start time to the fluctuation end time. The feature points and the sorting level are set based at least partially based on the distribution of the values of the measurement data and one or both of the time series patterns in which the values of the measurement data are arranged according to the date and time information.
The data sorting reference region is a two-dimensional region composed of a time range of one cycle and the entire range of measured values, the value of the first negative standard deviation waveform at the start of the fluctuation and the value of the standard deviation waveform at the end of the fluctuation. The lower side that linearly connects the values of the first negative standard deviation waveform, the value of the first positive standard deviation waveform at the start of the fluctuation, and the value of the first positive standard deviation waveform at the end of the fluctuation are linearly connected. The region between the upper side connected in a shape, or the value of the first positive standard deviation waveform at the start of the fluctuation and the value of the first positive standard deviation waveform at the end of the fluctuation are linearly connected. The region between the lower side and the upper side consisting of the first positive standard deviation waveform, or the value of the first average value waveform at the start of the fluctuation and the value of the first average value waveform at the end of the fluctuation. The region between the lower side linearly connecting the two and the upper side composed of the first average value waveform, or the value of the first negative standard deviation waveform at the start of the fluctuation and the first negative at the end of the fluctuation. Includes a region between the lower side that linearly connects the values with the side standard deviation waveform and the upper side that consists of the first negative standard deviation waveform.
The sorting range is a two-dimensional region composed of a time range of one cycle and the entire range of measured values, and at two time points in which the value in the measured value direction takes the sorting level on the boundary of the data sorting reference region. A data processing device characterized in that it is configured to be a time range between.

(4) In the data processing apparatus according to item (3), the first processing unit uses the first positive standard deviation waveform as the first upper limit waveform and the first negative standard deviation waveform as the first negative standard deviation waveform. A data processing device characterized by having one lower limit waveform.

(5) In the data processing device according to any one of (2) to (4), the data monitoring device further includes a fourth storage unit.
The second processing unit receives the first front measurement data or the first front measurement for each of one or more selected data sorting reference regions among the measurement data included in the first post-processing measurement data group. Regarding the measurement data consisting of the measurement data satisfying the predetermined time fluctuation condition among the data and the measurement data satisfying the predetermined time fluctuation condition among the first rear side measurement data, the waveform corresponding to the measurement data is the said in the sorting range. The average value and standard deviation of the time values that intersect the sorting level within the range before the feature point are calculated, and the average value is obtained by subtracting the default positive multiple of the standard deviation from the average value. On the other hand, the front time threshold range up to the value obtained by adding the default positive multiple of the standard deviation is set.
Of the measurement data included in the first post-processing measurement data group, for each of the selected one or more data sorting reference regions, a predetermined one of the first rear measurement data or the first front measurement data. Regarding the measurement data consisting of the measurement data satisfying the time fluctuation condition and the measurement data satisfying the predetermined time fluctuation condition among the first rear side measurement data, the waveform corresponding to the measurement data is the range after the feature point in the sorting range. The average value and standard deviation of the value of the time when the sorting level intersects with the sorting level are calculated, and the standard deviation with respect to the average value is obtained by subtracting the predetermined positive multiple of the standard deviation from the average value. Set the rear time threshold range up to the value obtained by adding the default positive multiples of
For each of the one or more data sorting reference areas selected from the data sorting reference area in which the front time threshold range and the rear time threshold range are set, the measurement data of a plurality of cycles stored in the first storage unit. Among them, the second front measurement data in which an intersection satisfying a predetermined time fluctuation condition between the waveform corresponding to the measurement data and the sorting level exists within the front time threshold range, and the measurement data within the rear time threshold range. Second rear measurement data in which there is an intersection satisfying a predetermined time fluctuation condition between the waveform corresponding to the above and the sorting level, and measurement data in both the front time threshold range and the rear time threshold range. The second-sided measurement data in which the intersection of the waveform corresponding to the above and the sorting level exists is selected, and all the second front-side measurement data, the second rear-side measurement data, and the second-sided measurement data are subjected to the first 3 Save in the storage
The second processing unit further removes the measurement data determined to be abnormal based on the first upper limit value waveform and the first lower limit value waveform from the measurement data stored in the third storage unit, and the rest. A second post-processing measurement data group consisting of the measurement data of the above is created, and the second post-processing measurement data group is stored in the third storage unit.
The second processing unit further measures after the second processing for each one or more data sorting reference regions selected from the data sorting reference regions in which the front time threshold range and the rear time threshold range are set. The average value and standard deviation of the value of the time when the waveform corresponding to the second two-sided measurement data included in the data group intersects the sorting level within the sorting range is calculated, and the standard deviation is calculated with respect to the average value. A two-sided time threshold range is set from a value obtained by subtracting a predetermined positive number multiple to a value obtained by adding a predetermined positive multiple of the standard deviation to the average value, and from the second post-processing measurement data group, the said Measurement data having an intersection that satisfies a predetermined time fluctuation condition between the waveform corresponding to the measurement data and the sorting level within the two-sided time threshold range is selected, and the selected measurement data is stored in the third storage unit.
The monitoring waveform creation unit selects the second two-sided measurement data related to any combination of one or more of the data sorting reference regions from the measurement data included in the second post-processing measurement data group. A monitoring measurement data group consisting of selected measurement data is created, the monitoring measurement data group is stored in the fourth storage unit, and based on the monitoring measurement data group stored in the fourth storage unit. , The data processing apparatus characterized in that the second upper limit value waveform and the second lower limit value waveform are created.

(6) In the data processing apparatus according to item (5), the monitoring waveform creation unit averages the measured values of one or more measurement data included in the monitoring measurement data group for each time point over one or more cycles. The value and standard deviation are calculated, and the second average value waveform, which is a series of the calculated average values at multiple time points, and the default positive multiple of the standard deviation at that time point for each mean value at multiple time points. The second positive standard deviation waveform, which is a series of added values, and the second negative standard deviation, which is the average value of each of multiple time points minus the default positive multiple of the standard deviation at that time point. A data processing apparatus characterized in that a waveform is created, the second positive standard deviation waveform is used as the second upper limit value waveform, and the second negative standard deviation waveform is used as the second lower limit value waveform.

(7) In the data processing apparatus according to item (6), the monitoring measurement data group is one selected data sorting from the second side measurement data included in the second post-processing measurement data group. A stand-alone monitoring measurement data group consisting of measurement data in which an intersection that satisfies a predetermined time variation condition between the waveform corresponding to the measurement data and the sorting level exists within the two-sided time threshold range related to the reference region, and a measurement data group for independent monitoring. Among the second two-sided measurement data included in the second post-processing measurement data group, the measurement data corresponds to at least one range of the two-sided time threshold range related to the plurality of selected data sorting reference regions. A Japanese-style monitoring measurement data group consisting of measurement data in which an intersection satisfying a predetermined time variation condition between the waveform and the sorting level exists, and the second-sided measurement data included in the second post-processing measurement data group. From the measurement data in which, within all the two-sided time threshold ranges related to the plurality of selected data sorting reference regions, there is an intersection that satisfies a predetermined time variation condition between the waveform corresponding to the measurement data and the sorting level. It consists of a group of measurement data for product type monitoring.
The monitoring waveform creation unit has, for each of the stand-alone monitoring measurement data group, the sum-type measurement data group, and the product-type monitoring measurement data group, the second average value waveform and the second positive standard. The deviation waveform and the second negative side standard deviation waveform are created, and at least one of the created plurality of the second positive side standard deviation waveforms is set as the second upper limit value waveform, and the created plurality of said A data processing apparatus characterized in that at least one of the second negative standard deviation waveforms is used as the second lower limit value waveform.

(8) In the data processing apparatus according to any one of (5) to (7), the second processing unit has the front time value range and the front time value range and each of the selected one or more selected data sorting reference regions. Using a plurality of the predetermined positive numbers used when setting the front time threshold range and the rear time threshold range, the rear time threshold range is used, and a plurality of plurality of positive numbers corresponding to the predetermined positive numbers are used. Set for each stage,
The monitoring waveform creation unit creates the monitoring measurement data group for any combination of one or more of the data sorting reference regions at each of the plurality of stages and stores the data in the fourth storage unit. (4) Based on the monitoring measurement data group stored in the storage unit, the second upper limit value waveform and the second lower limit value waveform are created for each of the plurality of steps. Data processing device.

(9) In the data processing apparatus according to any one of (5) to (8), waveform analysis configured to be able to display measurement data with the time direction as the horizontal axis and the measurement value direction of the measurement data as the vertical axis. And the display, including
The waveform analysis and display unit includes the first mean value waveform, the first positive standard deviation waveform, and the first negative standard deviation waveform in the candidate range for each of the one or more candidate ranges. It is configured to be able to display a vertical parallel line that can move in the time direction automatically or by input from the operator, passing through any time point within the candidate range.
The waveform analysis and display unit further displays a distribution diagram of the values of the measurement data of the plurality of cycles stored in the first storage unit at the time on the vertical axis parallel line and the values of the measurement data of the plurality of cycles as the date and time information. The time-series pattern diagram arranged according to the above is configured to be displayable, and based on the calculated mean value and standard deviation of the values over the plurality of cycles, the horizontal axis parallel line passing through the mean value, the standard with respect to the mean value. The horizontal axis parallel line passing through the value obtained by adding the default positive multiple of the deviation and the horizontal axis parallel line passing through the value obtained by subtracting the default positive multiple of the standard deviation from the mean value can be displayed. Being done
The waveform analysis and display unit further includes measurement data satisfying a predetermined time variation condition among the first front measurement data or the first front measurement data with respect to each of the one or more data sorting reference regions. Regarding the measurement data consisting of the measurement data satisfying the predetermined time fluctuation condition among the first rear side measurement data, the time when the waveform corresponding to the measurement data intersects the sorting level within the range before the feature point in the sorting range. It is configured to be able to display a distribution map of the values of and a time series pattern diagram in which the values of the time are arranged according to the date and time information, and is parallel to the vertical axis passing through the average value based on the average value and the standard deviation of the values of the time. A line, a vertical parallel line passing through a value obtained by adding a default positive multiple of the standard deviation to the average value, and a value obtained by subtracting a default positive multiple of the standard deviation from the average value. It is configured so that the vertical parallel lines that pass through can be displayed.
The waveform analysis and display unit further includes, for each of the one or more data sorting reference regions, the first rear measurement data or the measurement data satisfying a predetermined time variation condition among the first front measurement data. With respect to the measurement data consisting of the measurement data satisfying a predetermined time fluctuation condition among the first rear side measurement data, the waveform corresponding to the measurement data intersects the sorting level within the range after the feature point in the sorting range. It is configured to be able to display a distribution map of time values and a time series pattern diagram in which the time values are arranged according to the date and time information, and a vertical axis passing through the average values based on the average value and standard deviation of the time values. Parallel lines, vertical parallel lines passing through a value obtained by adding a default positive multiple of the standard deviation to the average value, and a value obtained by subtracting a default positive multiple of the standard deviation from the average value. It is configured to be able to display vertical parallel lines passing through
The waveform analysis and display unit further includes, for each of the one or more data sorting reference regions, the first rear measurement data or the measurement data satisfying a predetermined time variation condition among the first front measurement data. With respect to the measurement data consisting of the measurement data satisfying a predetermined time fluctuation condition among the first rear side measurement data, the waveform corresponding to the measurement data intersects the sorting level within the range after the feature point in the sorting range. It is configured to be able to display a distribution map of time values and a time series pattern diagram in which the time values are arranged according to the date and time information, and a vertical axis passing through the average values based on the average value and standard deviation of the time values. Parallel lines, vertical parallel lines passing through a value obtained by adding a default positive multiple of the standard deviation to the average value, and a value obtained by subtracting a default positive multiple of the standard deviation from the average value. It is configured to be able to display vertical parallel lines passing through
In the waveform analysis and display unit, the waveform corresponding to the second side measurement data included in the second post-processing measurement data group is further within the sorting range for each of the one or more data sorting reference regions. A data processing apparatus characterized in that a distribution map of time values intersecting a sorting level and a time series pattern diagram in which the time values are arranged according to the date and time information can be displayed.

(10) In the data processing apparatus according to any one of (5) to (9), the condition setting unit is included in the second post-processing measurement data at a plurality of time points within the time range of one cycle. The average value and standard deviation of the measured values for each time point over one or more cycles of the measurement data of one or more cycles are calculated, and the average value waveform for confirmation and the plurality of calculated average values of multiple time points are connected. A positive standard deviation waveform for confirmation, which is a series of values obtained by adding the default positive multiples of the standard deviation of the time point to each mean value of the time point, and the time point of each mean value of multiple time points. A negative standard deviation waveform for confirmation is created by subtracting the default positive multiple of the standard deviation of, and from the start time of the fluctuation to the end of the fluctuation for each set data sorting reference area. Set confirmation feature points based at least in part on one or both of the distribution of measurement data values over one or more cycles at each time point and the time series pattern in which the measurement data values are arranged according to the date and time information. Consists of
The waveform analysis and display unit is a distribution diagram of values of measurement data of one or more cycles included in the measurement data after the second processing at a time point on the vertical axis parallel line passing through an arbitrary time point within the candidate range. A time-series pattern diagram in which the values of the measurement data of the one or more cycles are arranged according to the date and time information can be displayed, and the average is based on the calculated average value and standard deviation of the values over the one or more cycles. A horizontal parallel line passing through a value, a horizontal parallel line passing through a value obtained by adding a predetermined positive multiple of the standard deviation to the average value, and a default positive number of the standard deviation with respect to the average value. A data processing device characterized in that a horizontal axis parallel line passing through a value obtained by subtracting a double can be displayed.

(11) In the data processing apparatus according to any one of (1) to (10), the monitoring waveform creating unit is within a time range of one cycle based on the second post-processing measurement data group. At a plurality of time points, the maximum value and the minimum value for each time point over one or more cycles of the measurement data of one or more cycles are obtained, and the second upper limit waveform is created based on the obtained maximum values at the plurality of time points. A data processing apparatus characterized in that the second lower limit waveform is created based on the obtained minimum values at a plurality of time points.

(12) In the data processing apparatus according to any one of (1) to (11), the measurement data of a plurality of cycles stored in the first storage unit is the same amount for a plurality of different measurement targets. A data processing device comprising measurement data obtained by measuring.

(13) In the data processing apparatus according to any one of (1) to (12), the measurement data of a plurality of cycles stored in the first storage unit is an amount that fluctuates with time according to the operation of the apparatus. A data processing device comprising the measured data to be represented.

(14) In the data processing apparatus according to any one of (1) to (13), the measurement data of a plurality of cycles stored in the first storage unit is measurement data representing the movement of a predetermined part of the human body. A data processing device characterized by including.

(15) In a data monitoring system including a plurality of measurement sites, the data processing device according to any one of (2) to (14), and a communication line connecting the measurement site and the data processing device. There,
Each of the measurement sites includes a measurement device that acquires measurement data of an amount that fluctuates with time so as to repeat one cycle from a start time point to an end point point, and a data monitoring device corresponding to the measurement device. The measurement data acquired at each of the measurement sites is the measurement data obtained by measuring the same amount at all the measurement sites.
The data monitoring device includes an input unit that receives measurement data from the corresponding measuring device, a storage unit that can store measurement data of a plurality of cycles, a monitoring upper limit waveform and a monitoring lower limit value in a time range of one cycle. Including a monitoring unit that monitors measurement data from the corresponding measuring device based on the waveform.
Each of the data monitoring devices included in the plurality of measurement sites transmits measurement data of a plurality of cycles to the data processing device via the communication line.
The data processing device outputs the second upper limit value waveform and the second lower limit value waveform obtained based on the measurement data of the plurality of cycles received from one or more of the data monitoring devices from the output unit to the communication line. Is transmitted to each of the data monitoring devices included in the plurality of measurement sites via
The monitoring unit of the data monitoring device included in the plurality of measurement sites receives the second upper limit value waveform and the second lower limit value waveform received from the data processing device, and measures data acquired after the reception. A data monitoring system characterized in that it is configured to be used as the monitoring upper limit waveform and the monitoring lower limit waveform for monitoring.

(16) In the data monitoring system according to item (15), the data monitoring device has an upper limit value of a time range of one cycle based on measurement data of a plurality of cycles stored in the storage unit of the data monitoring device. The monitoring unit further includes a calculation unit that creates a waveform and a lower limit value waveform, and the monitoring unit uses the upper limit value waveform and the lower limit value waveform created by the calculation unit as the monitoring upper limit value waveform and the monitoring lower limit value waveform, respectively. A data monitoring system characterized by being configured to be usable.

(17) In the data monitoring system according to (15) or (16), an external device connected to the data monitoring device is further included.
When the external device is an alarm device and the data monitoring device determines that the measurement data acquired from the corresponding measuring device is abnormal based on the monitoring upper limit value waveform and the monitoring lower limit value waveform. , A data monitoring system characterized in that the alarm device is configured to generate an alarm.

According to the present invention, in monitoring the measurement data for each cycle, it is possible to reduce the possibility that the measurement data, which is substantially normal even if there is a deviation in the time direction, is determined to be abnormal, and by extension, the operation of the measurement target. It becomes possible to provide a data processing device and a data monitoring system capable of improving the efficiency of the data.
In particular, when a waveform that serves as a reference for monitoring measurement data is generated by performing statistical processing on measurement data of multiple cycles, it shifts in the time direction while maintaining the advantage of being less susceptible to a small number of abnormal data. It is possible to generate an effective waveform that reflects the measurement data that is substantially normal.

It is a block diagram which shows typically the data monitoring system in one Embodiment of this invention. It is a figure which shows the 1st mean value waveform, the 1st positive side standard deviation waveform, and 1st negative side standard deviation waveform calculated for an example of the measurement data of a plurality of cycles in one Embodiment of this invention. It is a figure which shows an example of the candidate range for a data sorting reference area in one Embodiment of this invention. It is a figure which shows an example of the data sorting reference area in one Embodiment of this invention. It is a figure which shows another example of the data sorting reference area in one Embodiment of this invention. It is a figure which shows another example of the data sorting reference area in one Embodiment of this invention. It is a figure which shows another example of the data sorting reference area in one Embodiment of this invention. It is a figure which shows an example of the display in the waveform analysis and the display part in relation to the data sorting reference area shown in FIG. It is a figure which shows an example of the relationship between the sorting range and the sorting level and the measurement data in relation to the data sorting reference area shown in FIG. It is a figure which shows an example of the relationship between the sorting range and the sorting level and the measurement data in relation to the data sorting reference area shown in FIG. It is a figure which shows an example of the relationship between the sorting range and the sorting level and the measurement data in relation to the data sorting reference area shown in FIG. It is a figure which shows an example of the relationship between the sorting range and the sorting level and the measurement data in relation to the data sorting reference area shown in FIG. It is a figure which shows an example of the relationship between the sorting range and the sorting level and the measurement data in relation to the data sorting reference area shown in FIG. It is a figure which shows an example of setting of the front side time threshold range in relation to the data sorting reference area shown in FIG. It is a figure which shows an example of the distribution map of a time value in setting of the front side time threshold value range shown in FIG. It is a figure which shows an example of setting of the posterior time threshold range in relation to the data sorting reference area shown in FIG. It is a figure which shows an example of the distribution map of the time value in setting of the rear time threshold range shown in FIG. It is a figure which shows an example of setting of the two-sided time threshold range in relation to the data sorting reference area shown in FIG. It is a figure which shows an example of the time value distribution map in setting of the two-sided time threshold range shown in FIG. It is a figure which shows an example of the front side time threshold range in relation to the data sorting reference area shown in FIG. It is a figure which shows an example of the posterior time threshold range in relation to the data sorting reference area shown in FIG. In the first example of the data monitoring system in this embodiment, it is a figure which shows an example of the 2nd mean value waveform, the 2nd positive side standard deviation waveform, and the 2nd negative side standard deviation waveform. In the second example of the data monitoring system in this embodiment, it is a figure which shows typically an example of the motion to measure. In the second example of the data monitoring system in the present embodiment, the first mean value waveform, the first positive side standard deviation waveform, and the first negative side standard deviation waveform are set in the first candidate range (I) to the sixth candidate range. It is a figure shown together with (VI). It is a figure which shows an example of the 2nd mean value waveform, 2nd positive side standard deviation waveform, and 2nd negative side standard deviation waveform in the 2nd example of the data monitoring system in this embodiment. In the third example of the data monitoring system in this embodiment, it is a figure which shows an example of the measurement data measured at each measurement site. It is a figure which shows the example of the 2nd upper limit value waveform and the 2nd lower limit value waveform in the 3rd example of the data monitoring system in this embodiment. It is a figure which shows an example of the measurement data in the conventional measurement data monitoring apparatus. (A) and (b) are diagrams showing measurement data which is highly likely to be determined as abnormal due to a time lag in a conventional measurement data monitoring device.

The embodiment of the present invention will be described with reference to the attached figure.
As shown in FIG. 1, the data monitoring system 10 according to the present embodiment includes a plurality of measurement sites 20, a data processing device 50, and a communication line 40 connecting the measurement sites 20 and the data processing device 50. Each measurement site 20 includes a measurement device 21 that acquires measurement data of an amount that fluctuates with time so as to repeat one cycle from the start point to the end point, and a data monitoring device 30 corresponding to the measurement device 21. including. In the present embodiment, the measurement data acquired at each measurement site 20 is the measurement data obtained by measuring the same amount over all the measurement sites 20.

Here, in the present invention, the "same amount" measured at each measurement site 20 means an amount in which the same measurement data should ideally be acquired, and the actual definition thereof is the measurement target and the measurement target. / Or it is appropriately defined according to the data monitoring requirements. Some non-limiting examples of such an actual definition are as follows.
For example, when the measurement target is a device and the data monitoring system 10 is applied to measurement data of an amount (for example, current, mechanical vibration, pressure, etc.) that fluctuates with time according to the operation of the device, a plurality of measurements are performed. When devices of the same model are prepared as measurement targets at each of the sites 20 and each device is operated under the same conditions, the amount of time fluctuation generated in each device is referred to as the "same amount" for each different device. can do. Of course, the definition of "same" in "same model" and "same condition" may have a certain permissible range depending on, for example, the current measurement 20 fields and / or the measurement cycle.
As another example, when the measurement target is a human body and the data monitoring system 10 is applied to measurement data of an amount (for example, acceleration) corresponding to the movement of a predetermined part of the human body, the measurement at each of the plurality of measurement sites 20 is performed. An exercise caused by each individual subject in a predetermined area with the intention of performing a specific predetermined motion (eg, a series of steps in dance; see FIG. 23) common to a plurality of measurement sites 20. Corresponding amounts can be the same amount. In this case, the measurement targets are classified according to conditions such as age, gender, or skill for the target movement, and the amount corresponding to the exercise related only to the measurement targets classified into the same group is set to the same amount (in other words, the same amount). Even if the exercise is intended for the same movement, the amount corresponding to the exercise for individuals belonging to different groups shall be different amounts).

The data monitoring device 30 includes an input unit 31 that receives measurement data from the corresponding measurement device 21 (that is, included in the same measurement site 20), a storage unit 32 that can store measurement data of a plurality of cycles, and a communication line 40. A monitoring unit that monitors measurement data from the corresponding measuring device 21 based on the communication unit 35 that communicates with the data processing device 50 via the data processing device 50 and the monitoring upper limit value waveform and the monitoring lower limit value waveform in the time range of one cycle. It includes 34 and a calculation unit 33 that creates an upper limit value waveform and a lower limit value waveform in a time range of one cycle based on the measurement data of a plurality of cycles stored in its own storage unit 32.

Here, in the present invention, the term "waveform" related to data refers to data that can be expressed as a waveform, not the shape itself as a wave, unless otherwise explicitly stated. Further, in the present specification, the term waveform data is also used with the same meaning.

In the data monitoring system 10, each of the data monitoring devices 30 included in the plurality of measurement sites 20 is configured to transmit measurement data of a plurality of cycles from the communication unit 35 to the data processing device 50 via the communication line 40. To.

The data processing device 50 includes an input unit 51 and an output unit 52, and based on the measurement data received by the input unit 51, the upper limit waveform (second upper limit waveform described later) and the lower limit waveform (second upper limit waveform described later). 2 Lower limit waveform) is created. The measurement data received by the input unit 51 is the measurement data of each of a plurality of cycles transmitted from the data monitoring device 30 included in one or more (preferably all) measurement sites 20. Further, the data processing device 50 is configured to transmit the created upper limit value waveform and lower limit value waveform from the output unit 52 to each of the data monitoring devices 30 included in the plurality of measurement sites 20 via the communication line 40. To.

The monitoring unit 34 of the data monitoring device 30 included in the plurality of measurement sites 20 monitors the upper limit waveform and the lower limit waveform received from the data processing device 50 for the measurement data acquired after the reception. It is configured to be used as a value waveform and a lower limit waveform for monitoring.

In the data monitoring system 10 of the present embodiment, the data monitoring device 30 of each measurement site 20 uses the upper limit value waveform and the lower limit value waveform received from the data processing device 50 as the upper limit value waveform for monitoring and the lower limit value waveform for monitoring. During a predetermined period during which the measurement data is being monitored, the measurement data of a plurality of cycles received from the corresponding measuring device 21 and stored in the storage unit 32 is transmitted again to the data processing device 50, and the data processing device 50 receives each of them. New upper limit value waveforms and lower limit value waveforms are created based on at least a part of the measurement data newly received from the data monitoring device 30 at the measurement site 20 and transmitted to the data monitoring device 30 at each measurement site 20, and each measurement is performed. The monitoring device at the site 20 uses the upper limit value waveform and the lower limit value waveform newly received from the data processing device 50 as the upper limit value waveform for monitoring and the lower limit value waveform for monitoring, and monitors the measurement data acquired thereafter. It may be. Preferably, the data monitoring system 10 repeats such a series of processes during its operation.

Further, the data monitoring device 30 included in each measurement site 20 has a calculation unit 33, and the monitoring unit 34 has a vertical value waveform and a lower limit created by the calculation unit 33 of the data monitoring device 30 if necessary. The measurement data can be monitored by using the value waveform as the upper limit waveform for monitoring and the lower limit waveform for monitoring.

In the data monitoring system 10 of the present embodiment, the data monitoring device 30 includes the communication unit 35 as a component thereof, but in the data monitoring system according to the present invention, the data monitoring device 30 does not include the communication unit 35. In that case, the measurement site 20 including the data monitoring device 30 without the communication unit 35 has a communication function as an external device (not shown) connected to the data monitoring system 10. The data monitoring system 10 may be configured to communicate with the data processing device 50 via the communication line 40 using this device.

Next, the configuration of the data monitoring system 10 in this embodiment will be described in more detail. In the following, as an example for explanation, a case where the data monitoring system 10 is applied to measurement data of an amount that fluctuates with time according to the operation of the device is used. In the first example described below, more specifically, the measurement target at each measurement site 20 is a machine tool that processes a workpiece. In this case, the plurality of different measurement sites 20 are places where machine tools to be measured are installed, for example, in factories existing in a plurality of different regions (for example, different countries such as Japan, the United States, and China). May be good. Each of the machine tools included in all the measurement sites 20 carries out the same machining process on the workpiece to be the same product in the end, and within the range that meets the requirements for monitoring the measurement data. , The above machining process is executed by operating the machine tools of the same model and under the same conditions.

In this example, each machine tool repeatedly processes a plurality of workpieces with one cycle from the start time to the end of the machining process for one workpiece (in other words, the same machining cycle is repeated). Execute). At that time, the amount that fluctuates with time according to the operation of the device is, for example, the load current of the motor included in the machine tool, and the measuring device 21 corresponding to the load current is a current sensor.

In the data monitoring system 10, the input unit 31 of the data monitoring device 30 included in each measurement site 20 receives the measurement data from the corresponding measurement device 21. In this example, the input unit 31 shall sample the measurement data as analog data output from the measuring device 21 at a predetermined rate (for example, 1 kHz) and convert it into digital data consisting of a series of numerical values. The acquisition of is also included in the conversion to such digital data. In this case, each sampling point corresponds to each time point at each sampling interval (for example, 1 ms when the sampling rate is 1 kHz) in a time range over one cycle. In the following description, "time point" and "sampling point" in the time range of one cycle are used as alternative terms. Further, the elapsed time from the start time of one cycle can be expressed by the number of points indicating the number of sampling points, and unless otherwise specified, n points (n is an arbitrary positive). The expression (integer of) means the elapsed time (from the start of one cycle) expressed by the number of such points.
In another example, the data monitoring device 30 may acquire measurement data in synchronization with a pulse signal input at a predetermined timing (usually from an external device (not shown)). In this case, each pulse in the pulse signal is not necessarily input at a fixed time interval, and corresponds to, for example, a specific part of the work piece (for example, determined according to the importance of the machining dimension). In other words, it may be input in correspondence with the feed distance for performing the machining of the portion in the machine tool. In the present invention, the term "point corresponding to elapsed time" includes the case corresponding to each pulse in such pulse synchronization.

The input unit 31 further includes a trigger condition (internal trigger) for starting sampling, and the measurement is performed when the measurement data output from the measuring device 21 satisfies this trigger condition as the start time of one cycle. It may be the one that starts sampling the data. Further, each measurement site 20 includes an external device 22 connected to the input unit 31 of the data monitoring device 30, and the data monitoring device 30 is transmitted from the external device 22 to the input unit 31. The measurement data may be sampled by using the start signal as a trigger condition (external trigger). Alternatively, the monitoring device may start sampling measurement data based on the combination of the internal trigger and the external trigger.

The input unit 31 stores measurement data related to the processing process for one workpiece as waveform data for one cycle in the storage unit 32. In this case, the waveform data is measurement data from the start time to the end time of one cycle time range corresponding to one processing step, and in the case of digital data, it is acquired at each sampling point within the above time range. It is digital data in which numerical values corresponding to the measured values are arranged. While the machine tool repeatedly processes a plurality of workpieces, the storage unit 32 of the data monitoring device 30 stores measurement data for a plurality of cycles corresponding to each processing process as such waveform data. become.

Further, the identification information is added to each of the measurement data for one cycle, and each measurement data may be stored in the storage unit 32 together with the identification information. The measurement data identification information includes identification information of the measurement site 20 (country, factory, building, room / section, etc.), identification information of the measurement target (machinery), identification information of the measurement device 21 (current sensor), and measurement conditions. , One or more of the measurement dates and times are included, but not limited to these.

The data monitoring device 30 transmits the measurement data of a plurality of cycles, preferably together with the identification information, from the communication unit 35 to the data processing device 50 via the communication line 40. At that time, the data monitoring device 30 may transmit the acquired measurement data to the data processing device 50 for each measurement in one cycle. Alternatively, the data monitoring device 30 stores the measurement data of a plurality of cycles stored in the storage unit 32 at an arbitrary appropriate timing (for example, after a predetermined period has elapsed or after acquisition of measurement data for a predetermined number of cycles). It may be collectively transmitted to the data processing device 50.

The present embodiment is not limited to the specific mode of the communication line 40 connecting the data monitoring device 30 and the data processing device 50 of each measurement site 20, and is not limited to, for example, an arbitrary wired and / or wireless communication line. It may include any leased line constructed by a network, LAN, WAN, and / or the Internet.

The monitoring unit 34 of the data monitoring device 30 monitors the measurement data based on the monitoring upper limit waveform and the monitoring lower limit waveform. Specifically, in the monitoring unit 34, in general, is the value at each time point of the currently acquired measurement data included in the range from the value of the monitoring lower limit value waveform at that time point to the value of the monitoring upper limit value waveform? Whether or not it is determined over a time range of one cycle. If there is a time point when the value of the measurement data deviates from this range, the measurement data is determined to be abnormal.

In this example, an external device 23 is connected to the monitoring unit 34, and the external device 23 may be an alarm device that outputs an alarm signal based on a start signal from the monitoring unit 34. The alarm device notifies the personnel of the measurement site 20 (for example, the operator of the machine tool) of the abnormality of the measurement data. Alarm signals include, but are not limited to, acoustic, optical, and electrical signals, for example. When the monitoring unit 34 determines that the currently acquired measurement data is abnormal, the monitoring unit 34 may output an alarm signal to the external device 23 (in this case, the alarm device) at that time. When the personnel at the measurement site 20 notice an abnormality in the measurement data by the alarm signal output from the alarm device, they can take necessary measures for the machine tool and / or the workpiece.

Further, the monitoring unit 34 of the data monitoring device 30 may include one or more monitoring upper limit waveforms and one or more monitoring lower limit waveforms, from these waveforms to one monitoring upper limit waveform. A plurality of sets of monitoring waveform data composed of one monitoring lower limit waveform may be defined. Typically, of these sets of monitoring waveform data, the loosest set is used as the threshold for determining anomalies in the measurement data as described above, and other more stringent sets (generally multiple). The set may be used as a reference for comparing and discriminating the capabilities of each measurement site 20 (eg, machine tool and / or its operation).

As described above, the monitoring upper limit waveform and the monitoring lower limit waveform used by the monitoring unit 34 are usually the upper limit values transmitted from the data processing device 50 to each data monitoring device 30 via the communication line 40. The waveform and the lower limit waveform. However, the monitoring unit 34 uses the upper limit waveform and the lower limit waveform created by the calculation unit 33 based on the measurement data of the plurality of cycles stored in the storage unit 32 as the upper limit waveform for monitoring and the lower limit value for monitoring, respectively. It can also be used as a waveform.

At that time, the calculation unit 33 calculates the mean value and standard deviation of the measurement values of the measurement data of the plurality of cycles stored in the storage unit 32 at a plurality of time points within the time range of one cycle for each time point over the plurality of cycles. Then, the average value waveform obtained by connecting the calculated mean values at multiple time points and the default positive multiple (for example, 3 times, 5 times, etc.) of the standard deviation at that time point with respect to each mean value at multiple time points. ) Is added to the positive standard deviation waveform, and the negative standard deviation waveform is the mean value of each of the multiple time points minus the default positive multiple of the standard deviation at that time point. Is created, and the positive standard deviation waveform and the negative standard deviation waveform may be used as the upper limit value waveform and the lower limit value waveform, respectively.

Instead or in addition, the calculation unit 33 obtains the maximum value and the minimum value of the measured value at each time point over the plurality of cycles of the measurement data of the plurality of cycles stored in the storage unit 32, and obtains the plurality of time points. The upper limit waveform may be created based on the maximum value of, and the lower limit waveform may be created based on the minimum value waveform. In this case, the upper limit waveform may be, for example, a waveform obtained by appropriately increasing or decreasing the maximum value at a plurality of obtained time points and connecting the obtained values, and the lower limit waveform may be a plurality of obtained waveforms. The waveform may be a series of values obtained by appropriately increasing or decreasing the minimum value at a time point.

The data monitoring device 30 generally transmits all the input measurement data to the data processing device 50 and stores it in the storage unit 32. Therefore, the measurement data stored in the storage unit 32 by each data monitoring device 30 and transmitted to the data treatment device also includes the measurement data determined to be abnormal by the monitoring unit 34 during the input.

The data monitoring device 30 as described above is executed on arbitrary dedicated hardware and / or software executed on the hardware, or a general-purpose information processing device such as a personal computer and / or the device. It can be built with software or any combination of these.

Next, the details of the configuration of the data processing device 50 will be described. The data processing device 50 is used to monitor the time-varying amount of measurement data so as to repeat one cycle from the start time to the end time, and the first processing unit 62, the second processing unit 64, and the condition setting unit 71 are used. A monitoring waveform creation unit 66, a first storage unit 61, a second storage unit 63, a third storage unit 65, a fourth storage unit 67, an input unit 51, an output unit 52, and an input / output management unit 53 are included.

The input unit 51 receives measurement data transmitted from the data monitoring device 30 of each measurement site 20 included in the data monitoring system 10 via the communication line 40 (usually together with the identification information added by each data monitoring device 30). accept. The input unit 51 stores the received measurement data in the first storage unit 61. At this time, the input unit 51 further adds identification information to the received measurement data. The identification information added by the input unit 51 of the data processing device 50 includes date and time information representing the date and time when the measurement data was accepted by the input unit 51.

The output unit 52 sends the second upper limit value waveform and the second lower limit value waveform, which will be described later, to the data monitoring device 30 of each measurement site 20 in the data monitoring system 10.

The input / output management unit 53 is configured so that the input unit 51 can set the condition for accepting the measurement data and / or the output unit 52 can set the condition for transmitting the second upper limit value waveform and the second lower limit value waveform. For example, identification information of each data monitoring device 30 (and / or its communication unit 35) in the data monitoring system 10 is registered in advance in the input / output management unit 53, and the data processing device is based on the identification information. The 50 may be guaranteed to communicate data only with the correct device (ie, the data monitoring device 30 in the data monitoring system 10).

The first processing unit 62 creates a first upper limit value waveform and a first lower limit value waveform in a time range of one cycle based on the measurement data of a plurality of cycles stored in the first storage unit 61. The created first upper limit value waveform and first lower limit value waveform are stored in the first storage unit 61.

In addition, the first processing unit 62 is the average value and standard of the measured values of the measurement data of the plurality of cycles stored in the first storage unit 61 at a plurality of time points within the time range of one cycle. The deviation is calculated, and the first mean value waveform, which is a series of the calculated mean values at multiple time points, and the value obtained by adding the default positive multiple of the standard deviation at that time point to each mean value at multiple time points. The first positive standard deviation waveform, which is a series of the above, and the first negative standard deviation waveform, which is the mean value of each of the multiple time points minus the default positive multiple of the standard deviation at that time. create. The created first mean value waveform, first positive side standard deviation waveform, and first negative side standard deviation waveform are stored in the first storage unit 61.

Here, the first processing unit 62 may use the first positive side standard deviation waveform and the first negative side standard deviation waveform as the first upper limit value waveform and the first lower limit value waveform, respectively. Alternatively, the first upper limit value waveform and / or the first lower limit value waveform may be created separately from the first positive side standard deviation waveform and the first negative side standard deviation waveform. For example, the first upper limit value waveform is based on the maximum value of the measured values of the plurality of cycles of the measurement data stored in the first storage unit 61 at each time point over the plurality of cycles, and the obtained maximum values of the plurality of time points. And / or the first lower limit value waveform is the minimum value of the measured values of the plurality of cycles of the measurement data stored in the first storage unit 61 at each time point over the plurality of cycles. It may be created based on the obtained minimum values at a plurality of time points obtained. In this case, the first upper limit waveform may be, for example, a waveform formed by appropriately increasing or decreasing the obtained maximum values at a plurality of time points and connecting the obtained values, and the first lower limit waveform may be obtained. The waveform may be a series of values obtained by appropriately increasing or decreasing the minimum values at a plurality of time points.

In this example, the plurality of measurement sites 20 are factories set up in three different countries (Japan, the United States, and China). The measurement target is a machine tool that processes a predetermined work piece, and the load current of the motor included in the machine tool is acquired as measurement data. The first storage unit 61 of the data processing device 50 stores 18985 cycles of measurement data acquired by operating the machine tool for 20 days at each of the measurement sites 20 in Japan, the United States, and China.

FIG. 2 is a graph in which the horizontal axis represents time in units of the number of sampling points and the vertical axis represents load current (unit: ampere). In this example, the sampling interval is 1 ms, which is represented by the number of sampling points from 1 point to 7800 points, and the time range of one cycle corresponds to 7.8 s. FIG. 2 shows the first average value waveform (hereinafter, also referred to as m waveform) calculated from the measurement data of 18985 cycles stored in the first storage unit 61, and the first with the default positive number set to 1. 1 positive standard deviation waveform (hereinafter, also referred to as + 1σ waveform), 1st negative standard deviation waveform (hereinafter, also referred to as -1σ waveform), 1st positive standard deviation waveform when the above-mentioned default positive number is 3. (Hereinafter, also referred to as + 3σ waveform), first negative standard deviation waveform (hereinafter, also referred to as -3σ waveform), and first positive standard deviation waveform (hereinafter, + 5σ waveform) when the above-mentioned default positive number is 5. (Also also referred to as) and the first negative standard deviation waveform (hereinafter, also referred to as -5σ waveform) are shown.

The condition setting unit 71 has one or more data sorting reference areas 80 (FIGS. 4 to 4) that are a part of the two-dimensional area in the two-dimensional area composed of the time range of one cycle and the entire range of the measured values. 7) is set. Furthermore, the condition setting unit 71, to each of the data sorting criteria area 80, set one of time points of the data reference area 80 as a feature point tp, a single value of the measured values as a sorting level L _C (e.g. , See FIG. 8). Then, the condition setting unit 71, in accordance with the data sorting reference area 80 in the sorting level _{L C,} setting the sorting range (t2 to t3) as a time range including the feature point tp (e.g., see FIG. 8).

Specifically, the condition setting unit 71 first, for each of the one or more candidate ranges set within the time range of one cycle, the fluctuation start time point t0 in which the time variation of the first mean value waveform satisfies a predetermined variation start condition. And, the time variation of the first average value waveform determines the variation end time point t1 that satisfies a predetermined variation end time condition. FIG. 2 shows an example of three candidate ranges of the first candidate range (I), the second candidate range (II), and the third candidate range (III) set within the time range of one cycle. There is. In this example, these candidate ranges are preselected based on the importance of the machining site corresponding to a specific time range in the machining process of the workpiece, the characteristics of the known waveform corresponding to this machining step, and the like. There is.

In FIG. 3, the second candidate range (II) shown in FIG. 2 is enlarged and shown with the fluctuation start time point t0 and the fluctuation end time point t1 specified. In this example, the fluctuation start condition is that the increase in the measured value in the predetermined time width exceeds the predetermined threshold value, and the fluctuation end condition is that the decrease in the measured value in the predetermined time width is below the predetermined threshold value (absolute). (Exceeds as a value). Specifically, in this example, the fluctuation start time t0 is 2593 points, and the fluctuation end time is 3331 points. However, in the present invention, depending on the characteristics of the waveform of the measurement data, the fluctuation start condition is that the decrease in the measured value in the predetermined time width is below the predetermined threshold value, and the increase in the measured value in the predetermined time width is the predetermined threshold value. In some cases, the variable end condition may be greater than the above.

The condition setting unit 71 is based on the fluctuation start time t0 and the fluctuation end time t1, and at least one of the first mean value waveform, the first positive standard deviation waveform, and the first negative standard deviation waveform. The area 80 is set. Specifically, as shown in FIG. 4, the data sorting reference region 80 is the first negative standard at the fluctuation start time t0 in the two-dimensional region composed of the time range of one cycle and the entire range of the measured values. and the lower side E _L connecting the values of the first negative standard deviation waveform in value as the variation end t1 of difference waveform linearly, the value of the first positive standard deviation waveform in fluctuation start time t0 in variation end t1 the value of 1 positive standard deviation waveform may be an area between the upper side E _U connecting straight.

In the example of FIG. 4, the first negative standard deviation waveform defining the lower E _L of the data sorting reference area 80 in this case, -3Shiguma waveform is used, first to determine the upper side E _U of the data sorting reference area 80 A + 3σ waveform is used as the positive standard deviation waveform. The value of the -3σ waveform at the fluctuation start time t0 is 15.719A, and the value of the -3σ waveform at the fluctuation end time t1 is 15.726A. The value of the + 3σ waveform at the start of the fluctuation t0 is 20.263A, and the value of the + 3σ waveform at the end of the fluctuation t1 is 20.267A.

Alternatively, as shown in FIG. 5, the data sorting reference region 80 is the first positive standard deviation waveform at the fluctuation start time t0 in the two-dimensional region composed of the time range of one cycle and the entire range of the measured values. and the lower side E _L connecting the values of the first positive standard deviation waveform in value as the variation end t1 linearly, or may be a region between the upper edge E _U consisting of a first positive standard deviation waveform ..
In the example of FIG. 5, a + 1σ waveform is used as the first positive standard deviation waveform that defines the data sorting reference region 80. The value of the + 1σ waveform at the fluctuation start time t0 is 17.991A, and the value of the + 1σ waveform at the fluctuation end time t1 is 17.996A.

Alternatively, as shown in FIG. 6, the data sorting reference region 80 is a two-dimensional region composed of a time range of one cycle and the entire range of measured values, and the first mean value waveform (m) at the fluctuation start time t0. there the values and the value of the first average value waveform (m) in the change completion time t1 and the lower side E _L connecting linearly, in the region between the upper edge E _U consisting of a first mean value waveform (m) You may.

Alternatively, as shown in FIG. 7, the data sorting reference region 80 is the first negative standard deviation waveform at the fluctuation start time t0 in the two-dimensional region composed of the time range of one cycle and the entire range of the measured values. and the lower side E _L connecting the values of the first negative standard deviation waveform value as the variation end t1 linearly, or may be a region between the upper edge E _U consisting of a first negative standard deviation waveform ..
In the example of FIG. 7, a -1σ waveform is used as the first negative standard deviation waveform that defines the data sorting reference region 80. The value of the -1σ waveform at the fluctuation start time t0 is 17.894A, and the value of the -1σ waveform at the fluctuation end time t1 is 17.900A.

When the condition setting unit 71 arranges the distribution of the measurement data values over a plurality of cycles from the fluctuation start time t0 to the fluctuation end time t1 and the measurement data values according to the date and time information in the data sorting reference area 80. based at least in part on one or both of the series pattern, sets a feature point tp and sortation level L _C. This setting is usually carried out with reference to the functions of the waveform analysis and the display unit 72. Next, this point will be described with reference to FIG.
Here, FIG. 8, the data sorting reference area 80 shown in FIG. 5, corresponding to the example of setting the feature point tp and sortation level L _C. The data sorting reference area 80 is shown as a hatching area in FIG. 8A.

The waveform analysis and display unit 72 is configured to be capable of displaying measurement data with the time direction as the horizontal axis and the measurement value direction of the measurement data as the vertical axis. Further, the waveform analysis and display unit 72 includes a first mean value waveform, a first positive standard deviation waveform, a first negative standard deviation waveform, and a candidate range within the candidate range for each of the one or more candidate ranges. It is configured to be able to display a vertical parallel line that can be moved in the time direction automatically or by input from the operator of the data processing device 50 through any time point in the data processing device 50.

In the example shown in FIG. 8A, the first mean value waveform (m) within the second candidate range (II) is displayed, and the + 1σ waveform and the + 3σ waveform are displayed as the first positive standard deviation waveforms. As the first negative standard deviation waveform, a -1σ waveform and a + 3σ waveform are displayed, and a vertical parallel line Vp at the feature point tp is displayed. Although not shown, the waveform analysis and display unit 72 can generally display an arbitrary number of individual waveform data stored in the first storage unit 61 on top of the elements shown in FIG. 8A. It is composed of.

The waveform analysis and display unit 72 further includes a distribution diagram (FIG. 8 (b)) of the values of the measurement data of the plurality of cycles stored in the first storage unit 61 at the time on the vertical parallel line, and the plurality of cycles. A time-series pattern diagram (FIG. 8 (c)) in which the values of the measurement data are arranged according to the date and time information can be displayed.
In the illustrated example, FIG. 8B shows a distribution diagram of the measured values at the feature point tp as a histogram, and FIG. 8C shows a time-series pattern diagram of the measured values at the feature point tp. It is shown. The horizontal axis (data number) of FIG. 8C is date and time information (and necessary) corresponding to the reception date and time in the input unit 51 of the data processing device 50 for the measurement data stored in the first storage unit 61. If not, the numbers are added in order according to the date and time information of the identification information added at each measurement site 20).

Further, the waveform analysis and display unit 72 calculates the average value of the measurement data of the plurality of cycles stored in the first storage unit 61 based on the mean value and the standard deviation of the values at the time points on the vertical parallel lines over the plurality of cycles. The horizontal parallel lines that pass, the horizontal parallel lines that pass through the mean value plus the default positive multiple of the standard deviation, and the mean value minus the default positive multiple of the standard deviation. The horizontal axis parallel line passing through the value can be displayed.

In the example of FIG. 8, the horizontal axis parallel line h (passing the mean value) passing through the mean value and the standard deviation of the values at the feature point tp of the measurement data of the plurality of cycles stored in the first storage unit 61 over the plurality of cycles. m), horizontal axis parallel line h (+ 1σ) passing through the value obtained by adding the standard deviation to the mean value, horizontal axis parallel line h (+ 3σ) passing through the value obtained by adding 3 times the standard deviation to the mean value, The horizontal axis parallel line h (-1σ) that passes through the value obtained by subtracting the standard deviation from the mean value, and the horizontal axis parallel line h (-3σ) that passes through the current value that is three times the standard deviation with respect to the mean value are shown. Has been done.

Here, in the illustrated example, in FIGS. 8A, 8B, and 8C, the vertical axis (load current axis) is common, and the respective figures are shown side by side in the horizontal axis direction. The horizontal parallel lines h (m), h (+ 1σ), h (+ 3σ), h (-1σ), and h (-3σ) are common to FIGS. 8 (a), (b), and (c), respectively. It is shown as a straight line across each figure. The waveform analysis and display unit 72 may be configured to actually display each data in the manner shown in FIGS. 8A, 8B, and 8C. The present invention is not limited by the details of the waveform analysis and the display mode of the data in the display unit 72, but such a display mode is advantageous in that it is easy to compare and analyze the measured values between the figures. is there.

The waveform analysis and display unit 72 is shown in FIGS. 8 (b) and 8 (c) as the vertical parallel line (not shown) at an arbitrary time point shown in FIG. 8 (a) moves in the time axis direction. The distribution map and the time series pattern diagram are configured to be updated based on the time points on the vertical axis parallel lines, and the condition setting unit 71 is generally a waveform analysis and display unit 72 at an arbitrary time point. While moving the vertical parallel lines in the time axis direction, at least partially based on one or both of the distribution of measured values and the time series pattern as shown in FIGS. 8 (b) and 8 (c) (eg, in which the values in the feature point tp is set to the number of sorting levels L _C is greater than the measured data, a smaller measurement data in consideration of the balance of the number of) characteristic points tp and sortation level L _C. In this example, as a characteristic point tp in the data sorting reference area 80, it is set 2940 points, as the sorting level _{L C,} 19.6A is set.

These settings may include a portion that the condition setting unit 71 automatically executes (in other words, based on a default program) and / or the condition setting unit 71 processes data. It may include a part to be executed based on an input from an operator who operates the device 50 through an operation unit (not shown). At this time, the operator generally operates the data processing device 50 while observing the waveform analysis and the display as shown in FIG. 8 of the display unit 72.

Thus, in the two-dimensional region composed of the time range of one cycle and the entire range of measured values, if the data sorting reference area 80 and the sorting level L _C is set, on the boundary of the data sorting reference area 80 in value of the measured value direction as the time range between two time points t2, t3 taking sorting level _{L C,} sorting range (t2 to t3) is determined. In other words, in the two-dimensional area of the longitudinal axis of abscissa and the measured value of time, as the intersection of the horizontal axis parallel lines corresponding with the outer shape of the data sorting reference area 80 in the sorting level L _C, starting point of the sorting range t2 And the end point t3 is determined. In the example shown in FIG. 8 (a), a horizontal axis parallel line corresponding to the sorting level L _C, the outer shape of the data sorting reference region 80 (hatched portion) (in this case, is the top side + 1 [sigma waveform) are two intersections of the , The start time t2 and the end time t3 of the sorting range. In the case of this embodiment, a sorting level _{L C} 19.6A is the average substantially matching values of the measured values in the feature point tp (2940 points).

Here, when there are a plurality of candidate ranges in relation to the setting of the data sorting reference area 80, usually, for each candidate range, as described above, the data sorting reference area 80, the feature points tp related thereto, and the sorting. in which the level L _C is set (not shown). In the present embodiment, the data of the first candidate range (I) and the third candidate range (III) shown in FIG. 2 are also obtained by the same procedure as the procedure described above in relation to the second candidate range (II). and sort reference region 80, the feature point relates tp and sortation level L _C it is set.

Under the above condition settings, the first processing unit 62 measures the first front side of each of the one or more data sorting reference areas 80 from the measurement data of a plurality of cycles stored in the first storage unit 61. Data, first rear side measurement data, and first both side measurement data are selected, and the selected first front side measurement data, first rear side measurement data, and first side side measurement data are stored in the second storage unit 63. ..

Here, the first front measurement data is the measurement data of the plurality of cycles stored in the first storage unit 61, and as shown in FIGS. 9 and 10, the feature point tp or earlier in the sorting range (t2 to t3). a measurement data intersections and range (t2~tp) waveform corresponding to the measured data in the sorting level L _C is present. Further, the first rear measurement data is the measurement data of a plurality of cycles stored in the first storage unit 61, and as shown in FIGS. 11 and 12, after the feature point tp in the sorting range (t2 to t3). a measurement data crossing the range (tp~t3) in the between the measured data sorting level corresponding waveform L _C exists. The first sides measurement data, as shown in FIG. 13, the intersection of the waveform corresponding to the measured data and the sorting level L _C is, sorting range (t2 to t3)) before the subsequent feature point tp of It is the measurement data existing in both ranges (t2 to tp) and (tp to t3).

The first processing unit 62 further removes the measurement data determined to be abnormal based on the first upper limit value waveform and the first lower limit value waveform from the measurement data stored in the second storage unit 63, and the remaining measurement. A first post-processing measurement data group composed of data is created, and the first post-processing measurement data group is stored in the second storage unit 63.

Then, the second processing unit 64 selects measurement data from the measurement data of a plurality of cycles stored in the first storage unit 61 based on the measurement data group after the first processing, and the second processing unit 64 is composed of the selected measurement data. A post-processing measurement data group is created, and the second post-processing measurement data group is stored in the third storage unit 65.

Specifically, the second processing unit 64 receives the first front side measurement data or the first front side measurement for each of one or more selected data sorting reference areas 80 among the measurement data included in the first post-processing measurement data group. Regarding the measurement data consisting of the measurement data satisfying the predetermined time fluctuation condition among the data and the measurement data satisfying the predetermined time fluctuation condition among the first rear side measurement data, the waveform corresponding to the measurement data is the sorting range (t2 to t3). the average value and the standard deviation of the time value that intersects the sorting level L _C with the feature point tp previous range (t2~tp) of the predetermined integer multiple of the standard deviation relative to the mean value Set the front time threshold range from the subtracted value to the value obtained by adding the default positive multiple of the standard deviation to the average value.

In this example, with respect to the data sorting reference region 80 shown in FIGS. 5 and 8, the first front measurement data is measurement data that satisfies the above-mentioned conditions with reference to FIGS. 9 and 10. The time variation condition in a predetermined time variation satisfying the measurement data of the first front measurement data, as shown in FIG. 9, is to intersect the sorting level L _C while increasing the measurement value, the first time variation condition in a predetermined time variation satisfying the measurement data of the rear-side measurement data, as shown in FIG. 12, it is intersecting the sorting level L _C while reducing the measured value. Therefore, the measurement data used to obtain the front time threshold range is simply the measurement data satisfying the conditions shown in FIGS. 9 and 10, or the measurement data satisfying the conditions shown in FIGS. 9 and 12. is there.

In this example, it is assumed that the front side time threshold range is obtained by using the first front side measurement data (that is, the measurement data satisfying the conditions shown in FIGS. 9 and 10). Specifically, in the second processing unit 64, with respect to the first front measurement data, the waveform corresponding to each measurement data is within the sorting range (t2 to t3) within the range (t2 to tp) before the feature point tp. calculating the average value and the standard deviation of the time value that intersects the sorting level L _C. In this example, the mean value tf is 2769 points and the standard deviation is 78 points (see FIG. 14). Further, in this example, from the value t4 obtained by subtracting 3 times the standard deviation (234 points) from the mean value tf to the value t5 obtained by adding 3 times the standard deviation (234 points) to the mean value tf. The front time threshold range (t4 to t5) is set as the time range of (see FIG. 20).

In connection with this, the waveform analysis and display unit 72 indicates the first front measurement data or the measurement data satisfying a predetermined time fluctuation condition among the first front measurement data for each of the one or more data sorting reference regions 80. When the measurement data of a predetermined time variation satisfying the measurement data of the first rear-side measurement data, the waveform corresponding to the measured data crosses the sorting level L _C within the feature point tp earlier of sorting range It is configured to be able to display a distribution map of time values and a time series pattern diagram in which the time values are arranged according to date and time information. Further, the waveform analysis and display unit 72 added the vertical parallel lines passing through the mean value and the default positive multiple of the standard deviation to the mean value based on the mean value and the standard deviation of the values at this time. The vertical parallel line passing through the value and the vertical parallel line passing through the value obtained by subtracting the default positive multiple of the standard deviation from the mean value can be displayed.

In this example, with respect to the first front measurement data (that is, the measurement data satisfying the conditions shown in FIGS. 9 and 10), the waveform corresponding to the measurement data is in the sorting range (t2 to t3) before the feature point tp (that is, distribution map of time values that intersect the sorting level _{L C} with T2～tp) in (histogram) is shown (see Figure 15). Although not shown, the time-series pattern diagram is a diagram in which the values of this time are arranged and drawn in the order of data numbers as in FIG. 8 (c). Further, in FIG. 14, the vertical axis parallel line Vf passing through the mean value tf, the vertical axis parallel line V5'passing through the value obtained by adding the standard deviation to the mean value tf, and the standard deviation with respect to the mean value tf are shown. The vertical axis parallel line V4'passing through the subtracted value is shown, and FIG. 20 shows the vertical axis parallel line Vf passing through the mean value tf and the vertical axis parallel line passing through the value obtained by adding the standard deviation to the mean value tf. Line V5', vertical parallel line V5 passing through the value t5 obtained by adding 3 times the standard deviation to the mean value tf, vertical parallel line V4' passing through the value obtained by subtracting the standard deviation from the mean value tf, and A vertical parallel line V4 passing through a value t4 obtained by subtracting three times the standard deviation from the mean value tf is shown.

The second processing unit 64 further increases the first rear side measurement data or the first front side measurement for each of one or more selected data sorting reference areas 80 among the measurement data included in the first post-processing measurement data group. Regarding the measurement data consisting of the measurement data satisfying the predetermined time fluctuation condition among the data and the measurement data satisfying the predetermined time fluctuation condition among the first rear measurement data, the waveform corresponding to the measurement data is the feature point tp in the sorting range. the average value and the standard deviation of the time value that intersects the sorting level L _C within the later, the standard for the average value from the value obtained by subtracting the default integer multiple of the standard deviation relative to the mean value Sets the posterior time threshold range up to the sum of the default positive multiples of the deviation.

In this example, with respect to the data sorting reference area 80 shown in FIGS. 5 and 8, the first rear measurement data is measurement data that satisfies the above-mentioned conditions with reference to FIGS. 11 and 12. The time variation condition in a predetermined time variation satisfying the measurement data of the first front measurement data, as shown in FIG. 10, it is to intersect the sorting level L _C while reducing the measurement value, the first time variation condition in a predetermined time variation satisfying the measurement data of the rear-side measurement data, as shown in FIG. 11, it is intersecting the sorting level L _C while increasing the measurement values. Therefore, the measurement data used to obtain the rear time threshold range is simply the measurement data satisfying the conditions shown in FIGS. 11 and 12, or the measurement data satisfying the conditions shown in FIGS. 10 and 11. Is.

In this example, the first posterior measurement data (that is, the measurement data satisfying the conditions shown in FIGS. 11 and 12) is used to determine the posterior time threshold range. Specifically, in the second processing unit 64, with respect to the first rear side measurement data, the waveform corresponding to each measurement data is within the range (tp to t3) after the feature point tp in the sorting range (t2 to t3). in calculating the mean value and the standard deviation of the time value that intersects the sorting level L _C. In this example, the mean tr is 3123 points and the standard deviation is 102 points (see FIG. 16). Further, in this example, from the value t6 obtained by subtracting 3 times the standard deviation (306 points) from the average value tr to the value t7 obtained by adding 3 times the standard deviation (306 points) to the average value tr. The rear time threshold range (t6 to t7) is set as the time range of (see FIG. 21).

In this regard, the waveform analysis and display unit 72 measures the first rear side measurement data or the first front side measurement data that satisfies a predetermined time variation condition for each of the one or more data sorting reference regions 80. for data and measurement data of a predetermined time variation satisfying the measurement data of the first rear-side measurement data, it crosses the sorting level L _C waveform corresponding to the measurement data within the feature point tp after of sorting range It is configured to be able to display a distribution map of time values and a time series pattern diagram in which the time values are arranged according to date and time information. Further, the waveform analysis and display unit 72 added the vertical parallel lines passing through the mean value and the default positive multiple of the standard deviation to the mean value based on the mean value and the standard deviation of the values at this time. The vertical parallel line passing through the value and the vertical parallel line passing through the value obtained by subtracting the default positive multiple of the standard deviation from the mean value can be displayed.

In this example, with respect to the first rear measurement data (that is, the measurement data satisfying the conditions shown in FIGS. 11 and 12), the waveform corresponding to the measurement data is in the sorting range (t2 to t3) after the feature point tp. distribution map of time values that intersect the sorting level _{L C} with (Tp～t3) in (histogram) is shown (see Figure 17). Although not shown, the time-series pattern diagram is a diagram in which the values of this time are arranged and drawn in the order of data numbers as in FIG. 8 (c). Further, in FIG. 16, the vertical axis parallel line Vr passing through the mean value tr, the vertical axis parallel line V7'passing through the value obtained by adding the standard deviation to the mean value tr, and the standard deviation with respect to the mean value tr are shown. The vertical axis parallel line V6'passing through the subtracted value is shown, and FIG. 21 shows the vertical axis parallel line Vr passing through the mean value tr and the vertical axis parallel passing through the value obtained by adding the standard deviation to the mean value tr. Line V7', vertical parallel line V7 passing through the value t7 obtained by adding 3 times the standard deviation to the mean value tr, vertical parallel line V6' passing through the value obtained by subtracting the standard deviation from the mean value tr, and The vertical axis parallel line V6 passing through the value t6 obtained by subtracting 3 times the standard deviation from the mean value tr is shown.

In addition to the display elements described above, waveforms corresponding to a plurality of measurement data are superimposed and displayed in FIGS. 20 and 21. These waveforms correspond to measurement data of the same type as the waveforms displayed in FIGS. 29 (a) and 29 (b), respectively, as a description of the prior art. 20 and 21 also show that these waveforms satisfy the above-mentioned conditions for the front time threshold range (t4 to t5) and the rear time threshold range (t6 to t7), respectively.

The second processing unit 64 further receives one or more data sorting reference areas 80 selected from the data sorting reference areas 80 in which the front time threshold range (t4 to t5) and the rear time threshold range (t6 to t7) are set. to, among the measurement data of a plurality of cycles stored in the first storage unit 61, the front time threshold range (t4 to t5) within a predetermined time variation conditions of the waveform corresponding to the measured data and the sorting level L _C second second front measurement data intersection exists, the rear time threshold range (t6 to t7) within the predetermined time variation satisfying the intersection with the waveform corresponding to the measured data and the sorting level L _C exists which satisfies rear measurement data, and, in both the front time threshold range (t4 to t5) in the rear time threshold range (t6 to t7), there is an intersection of the waveform corresponding to the measured data and the sorting level L _C The second side measurement data is selected, and all the second front side measurement data, the second rear side measurement data, and the second side side measurement data are stored in the third storage unit 65.

The second processing unit 64 further removes the measurement data determined to be abnormal based on the first upper limit value waveform and the first lower limit value waveform from the measurement data stored in the third storage unit 65, and the remaining measurement. A second post-processing measurement data group composed of data is created, and the second post-processing measurement data group is stored in the third storage unit 65.

The second processing unit 64 further receives one or more data sorting reference areas 80 selected from the data sorting reference areas 80 in which the front time threshold range (t4 to t5) and the rear side time threshold range (t6 to t7) are set. to, the average value and the standard deviation of the values of the time waveform corresponding to the second sides measurement data included in the second post-processing measurement data group crosses the sorting level L _C within sorting range, the average value On the other hand, the two-sided time threshold range is set from the value obtained by subtracting the default positive multiple of the standard deviation to the value obtained by adding the default positive multiple of the standard deviation to the mean value.

In this example, the mean value tb is 2968 points and the standard deviation is 325 points (see FIG. 18). Further, in this example, both sides are set as a time range from the value t8 obtained by subtracting the standard deviation (325 points) from the average value tb to the value t9 obtained by adding the standard deviation (325 points) to the average value. The time threshold range (t8 to t9) is set (see FIG. 18).

In relation to this, in the waveform analysis unit and the display unit, for each of the one or more data sorting reference regions 80, the waveform corresponding to the second side measurement data included in the second post-processing measurement data group is the sorting range (t2). -t3) display configured to allow distribution diagram and a time-sequence pattern diagram arranged according to date and time information value of the time value of the sorting level L _C and the time to cross within.
Figure 19, the distribution diagram of a second process after the second sides measured data to the corresponding waveform sorting range (t2 to t3) the sorting level L _C with the time to cross at a value included in the measurement data group in this example An example of (histogram) is shown. Since the distribution shown in FIG. 19 is the distribution related to the second two-sided measurement data, it is a bimodal distribution in which the values are distributed in the first half portion and the second half portion of the sorting range (t2 to t3), respectively. Although not shown, the time-series pattern diagram is a diagram in which the values of this time are arranged and drawn in the order of data numbers as in FIG. 8 (c). Further, in FIGS. 18 and 19, the vertical parallel line Vb passing through the average value tb, the vertical parallel line V9 passing through the value t9 obtained by adding the standard deviation to the average value tb, and the average value tb. A vertical parallel line V8 passing through the value obtained by subtracting the standard deviation is shown.

Then, the second processing section 64, predetermined time variation satisfying the intersection of the second post-processing measurement data group, the waveform corresponding to the measurement data on either side time threshold range (t8 to t9) in a sorting level L _C Is selected, and the selected measurement data is stored in the third storage unit 65.
The time variation condition in this case is, for example, there is an intersection that intersects the sorting level L _C while increasing the measured value is temporally preceding and the sorting level L _C while reducing the measured value to a later point in time There are intersections that intersect.

The monitoring waveform creation unit 66 is based on the second post-processing measurement data group in order to monitor the measurement data acquired after the period in which the plurality of measurement data stored in the first storage unit 61 are acquired. Create one or more second upper limit waveforms and one or more second lower limit waveforms in the time range of one cycle.

For details on the creation of the second upper limit value waveform and the second lower limit value waveform, the monitoring waveform creation unit 66 can arbitrarily set one or more data sorting reference areas 80 among the measurement data included in the second post-processing measurement data group. The second two-sided measurement data related to the combination of the above is selected, a monitoring measurement data group consisting of the selected measurement data is created, the monitoring measurement data group is stored in the fourth storage unit 67, and the fourth It is configured to create a second upper limit value waveform and a second lower limit value waveform based on the monitoring measurement data group stored in the storage unit 67.

In particular, the monitoring waveform creation unit 66 calculates the mean value and standard deviation of the measured values at each time point over one or more cycles of one or more measurement data included in the monitoring measurement data group, and calculates a plurality of time points. The second mean value waveform, which is a series of mean values, and the second positive standard deviation waveform, which is a series of values obtained by adding the default positive multiples of the standard deviation at that time point to the mean value of each time point. , Create a second negative standard deviation waveform that is a series of values obtained by subtracting the default positive multiple of the standard deviation at that time from the mean value of each of multiple time points, and create the second positive standard. The deviation waveform may be the second upper limit value waveform, and the second negative standard deviation waveform may be the second lower limit value waveform.

Here, the monitoring measurement data group is typically a stand-alone monitoring measurement data group, a Japanese-style monitoring measurement data group, or a product-type monitoring measurement data group. Here, the stand-alone monitoring measurement data group is a two-sided time threshold range (t8 to t9) related to one selected data sorting reference region 80 from the second two-sided measurement data included in the second post-processing measurement data group. ) in the consist measurement data of a predetermined time variation satisfying the intersection with the waveform corresponding to the measured data and the sorting level L _C exists.

For example, in this example, if one selected data sorting reference area 80 is the data sorting reference area 80 defined in relation to the second candidate range (II) shown in FIG. 2, the measurement data for independent monitoring is used. The group is sorted from the waveform corresponding to the measurement data within the two-sided time threshold range (t8 to t9) related to the second candidate range (II) among the second two-sided measurement data included in the second post-processing measurement data group. consisting satisfies the measurement data of the predetermined time variation satisfying the intersection of the level L _C exists.

Further, the Japanese-style monitoring measurement data group is at least one range of the two-sided time threshold range related to a plurality of selected data sorting reference regions 80 among the second two-sided measurement data included in the second post-processing measurement data group. within consist measurement data of a predetermined time variation satisfying the intersection with the waveform corresponding to the measured data and the sorting level L _C exists.

For example, in this example, the selected plurality of data sorting reference areas 80 are referred to as the data sorting reference area 80 (hereinafter, also referred to as the first data sorting reference area) defined in relation to the first candidate range (I) shown in FIG. ), The data sorting reference area 80 defined in relation to the second candidate range (II) (hereinafter, also referred to as the second data sorting reference area), and the third candidate range (III). Assuming that the data sorting reference area 80 (hereinafter, also referred to as the third data sorting reference area), the Japanese-style monitoring measurement data group is the first of the second two-sided measurement data included in the second post-processing measurement data group. Within the two-sided time threshold range related to the data sorting reference area, within the two-sided time threshold range (t8 to t9) related to the second data sorting reference area, and within the two-sided time threshold range related to the third data sorting reference area. at least one of which consists of satisfying the measurement data of the predetermined time variation satisfying the intersection with the waveform corresponding to the measured data and the sorting level L _C exists.

Further, the product type monitoring measurement data group is within all the two-sided time threshold ranges related to the plurality of selected data sorting reference regions 80 among the second two-sided measurement data included in the second post-processing measurement data group. predetermined time variation satisfying the intersection with the waveform corresponding to the measured data and the sorting level L _C is made of the measurement data present.

For example, in this example, if the plurality of selected data sorting reference areas 80 are the first data sorting reference area, the second data sorting reference area, and the third data sorting reference area, the product type monitoring measurement data group is Of the second two-sided measurement data included in the second post-processing measurement data group, the two-sided time threshold range related to the first data sorting reference area and the two-sided time threshold range related to the second data sorting reference area (t8 to t9) in, and, provided that all of the sides time threshold range associated with the third data sorting reference area, a predetermined time variation satisfying the intersection with the waveform corresponding to the measured data and the sorting level L _C exists It consists of measurement data that satisfies.

The inclusion relationship as a set of these three types of measurement data groups is that the same data sorting reference area 80 as one data sorting reference area 80 used to determine the stand-alone monitoring measurement data group is a product type. When it is also used to define the measurement data group for monitoring and the measurement data group for Japanese-style monitoring,
(Product type) ⊆ (Independent type) ⊆ (Japanese type)
Is what becomes. However, (product type), (independent type), and (Japanese type) refer to the independent type monitoring measurement data group, the product type monitoring measurement data group, and the Japanese type monitoring measurement data group as a set of measurement data, respectively. Represent. For example, the data sorting reference area 80 used to create these monitoring measurement data groups is appropriately selected according to the importance of the processing process corresponding to each of the plurality of candidate ranges.

The monitoring waveform creation unit 66 sets the second mean value waveform, the second positive standard deviation waveform, and the second positive side standard deviation waveform for each of the stand-alone monitoring measurement data group, the Japanese measurement data group, and the product type monitoring measurement data group. 2 Negative side standard deviation waveforms are created, at least one of the created plurality of second positive side standard deviation waveforms is set as the second upper limit value waveform, and among the plurality of created second negative side standard deviation waveforms. At least one of may be the second lower limit value waveform.

FIG. 22 shows a second mean value waveform created using a stand-alone monitoring measurement data group based on the data sorting reference region 80 defined in relation to the second candidate range (II) shown in FIG. (M), the second positive standard deviation waveform (+ 1Σ) and the second negative standard deviation waveform (-1Σ) when the default positive number is 1, and the second with the above default positive number being 3. The positive standard deviation waveform (+ 3Σ) and the second negative standard deviation waveform (-3Σ) are shown. Each waveform shown in FIG. 22 is shown with reference to, for example, the start time t4 of the front time threshold range (t4 to t5) and the end time t7 of the rear time threshold range (t6 to t7). 20. Compared with each waveform shown in FIG. 21 and the like, the second mean value waveform (M), the second positive side standard deviation waveform (for example, + 1Σ), and the second negative side standard deviation waveform (for example, -1Σ) are , 1st mean waveform (m), 1st positive standard deviation waveform (eg + 1σ), and 1st negative standard deviation waveform (eg -1σ) in the time direction (especially its positive direction). It can be seen that the shape is stretched in both the negative and negative directions.

Of the waveforms shown in FIG. 22, for example, the + 3Σ waveform and the -3Σ waveform may be selected as the second upper limit waveform and the second lower limit waveform, respectively. Alternatively, the set of + 1Σ waveform and -1Σ waveform is set as one set of the second upper limit value waveform and the second lower limit value waveform, and another set of the + 3Σ waveform and the -3Σ waveform, the second upper limit value waveform, and the second lower limit value waveform. As the set, a plurality of sets of the second upper limit value waveform and the second lower limit value waveform may be configured.

Further, the second processing unit 64 sets the front time value range and the rear time threshold range, and the front time threshold range and the rear time threshold range for each of the selected one or more data sorting reference regions 80. A plurality of default positive numbers to be used are used and set for each of a plurality of stages according to the plurality of default positive numbers, and the monitoring waveform creation unit 66 sorts the monitoring measurement data group into one or more data groups. Any combination of the reference area 80 is created for each of a plurality of stages and stored in the fourth storage unit 67, and based on the monitoring measurement data group stored in the fourth storage unit 67, the first is performed for each of the plurality of stages. 2 The upper limit value waveform and the second lower limit value waveform may be created.

For example, three stages of 1, 3, and 5 are used as default positive numbers used when setting the front time threshold range and the rear time threshold range, and a predetermined monitoring measurement data group is used for each of these stages. For each of the three stages of predetermined monitoring measurement data groups, the default positive number for creating the second positive standard deviation waveform and the second negative standard deviation waveform is 0.5. Assuming that there are five types of 1, 3, 4, and 5, a set of five types of the second positive standard deviation waveform and the second negative standard deviation waveform for each of the three stages of predetermined monitoring measurement data groups. Is created. Therefore, if the set of the second positive standard deviation waveform and the second negative standard deviation waveform is the set of the second upper limit value waveform and the second lower limit value waveform, the second upper limit value waveform and the second lower limit value value are used. Fifteen sets of waveforms will be composed. On the other hand, by variously changing the definition of the monitoring measurement data group as described above, the set of the second upper limit value waveform and the second lower limit value waveform can be further increased.

As described above, in the data monitoring system 10, these (1 or more) second upper limit value waveforms and second lower limit value waveforms are transmitted from the output unit 52 to each measurement site 20 via the communication line 40 as described above. It is transmitted to the data monitoring device 30 and used as a monitoring upper limit waveform and a monitoring lower limit waveform for monitoring the measurement data acquired after the reception at each measurement site 20. When a plurality of sets consisting of the second upper limit value waveform and the second lower limit value waveform are created and transmitted to the data monitoring device 30 of each measurement site 20, each data monitoring device 30 has such a plurality of sets. (2) Using the upper limit value waveform and the second lower limit value waveform, a plurality of sets of monitoring waveform data including the upper limit value waveform for monitoring and the lower limit value waveform for monitoring can be determined. As described above, among these plurality of sets of monitoring waveform data, for example, the set with the loosest condition (in other words, the wide allowable range) is used as the threshold value for determining the abnormality of the measurement data, and other more conditions. Strict (generally multiple) pairs, even those used as criteria for comparing and discriminating the capabilities (eg, of machine tools and / or their operations) for each measurement site 20 with respect to the measurement target. Good.

Here, the condition setting unit 71 of the data processing device 50 further sets one or more cycles of the measurement data of one or more cycles included in the measurement data after the second processing at a plurality of time points within the time range of one cycle. The average value and standard deviation of the measured values for each time point over a period of time are calculated, and the average value waveform for confirmation, which is a series of the calculated average values of multiple time points, and the standard value at that time point for each average value of multiple time points. The positive standard deviation waveform for confirmation, which is a series of values obtained by adding the default positive multiples of the deviation, and the value obtained by subtracting the default positive multiple of the standard deviation at that time from the average value of each of the multiple time points. Create a series of negative standard deviation waveforms for confirmation, and distribute and measure the values of the measured data over one or more cycles from the start of the fluctuation to the end of the fluctuation for each set data sorting reference region 80. It is configured to set confirmation feature points based at least in part on one or both of the time series patterns in which data values are arranged according to date and time information.

In this regard, the waveform analysis unit and the display unit are the values of the measurement data of one or more cycles included in the second post-processing measurement data at the time on the vertical parallel line passing through any time point in the candidate range. It is configured to be able to display a time series pattern diagram in which the distribution map and the values of the measurement data of one or more cycles are arranged according to the date and time information, and based on the calculated average value and standard deviation of the values over the one or more cycles. The horizontal axis parallel line passing through the mean value, the horizontal axis parallel line passing through the value obtained by adding the default positive multiple of the standard deviation to the mean value, and the default positive multiple of the standard deviation to the mean value. The horizontal axis parallel line passing through the subtracted value can be displayed.

Further, the monitoring waveform creation unit 66 is based on the second post-processing measurement data group, at a plurality of time points within the time range of one cycle, for each time point of the measurement data of one or more cycles over one or more cycles. The maximum value and the minimum value are obtained, the second upper limit waveform is created based on the obtained maximum values at multiple time points, and the second lower limit waveform is created based on the obtained minimum values at multiple time points. It may be.

The data processing device 50 as described above is executed on arbitrary dedicated hardware and / or software executed on the hardware, or a general-purpose information processing device such as a personal computer and / or the device. It can be built with software or any combination of these. In particular, in the present embodiment, the data processing device 50 includes the first storage unit 61, the second storage unit 63, the third storage unit 65, and the fourth storage unit 67 as the storage means. In the present invention, a configuration that realizes each of the above-mentioned functions in relation to these components can be constructed using any suitable one or more hardware. Similarly, the data processing device 50 in the present embodiment includes a first processing unit 62, a second processing unit 64, an input / output management unit 53, and a waveform creation for monitoring for executing various processes on the stored data. Although it is assumed that the unit 66, the condition setting unit 71, the waveform analysis and the display unit 72 are included, the configuration that realizes each of the above-mentioned functions in relation to these components includes one central processing unit or a central processing unit. It can be constructed using information processing devices and programs running on them, and / or any combination of display devices associated with those devices. Further, in the above description, among various data obtained as a result of the processing executed for the measurement data stored in each storage unit, the data necessary for the subsequent processing is not particularly mentioned. Even in this case, it goes without saying that they are stored in appropriate storage units (the first to fourth storage units 61, 63, 65, 67, or other storage units described above).

Since the data monitoring system 10 in the present embodiment uses the second upper limit value waveform and the second lower limit value waveform created as described above by the data treatment device in order to monitor the measurement data at each measurement site 20. In the monitoring of individual measurement data at each measurement site 20, the possibility of determining measurement data that is substantially normal only with a time lag is reduced, and as a result, the operation of the measurement target (for example, work). It is possible to improve the efficiency of processing the workpiece by the machine).

At that time, since the second upper limit value waveform and the second lower limit value waveform created by the data treatment apparatus are created by aggregating the measurement data at each measurement site 20 and based on the aggregated data. The improvement in efficiency can be achieved while appropriately maintaining the accuracy of the measurement data monitoring work at each measurement site 20.

Further, in the data processing apparatus 50 of this embodiment, the feature point tp and sortation level L _C, be set based at least in part on the distribution and / or chronological pattern of values of the measurement data, these values Since the distribution is often not normally distributed (as compared to, for example, simply setting based on the mean, etc.), it is advantageous to perform a more reasonable setting. At that time, the waveform analysis and display unit 72 of the data processing device 50 is configured to be able to display the corresponding distribution map (histogram) and time series pattern diagram. In particular, these settings are the operations of the data processing device 50. It is even more advantageous if it includes parts that are executed based on input from a person.

Further, since the condition setting unit 71 of the data treatment device in the present embodiment is configured to set the confirmation feature point based on the measurement data after the second processing, the data stored in the first storage unit 61. The feature point tp, which is set based on the above and is used to create the second upper limit value waveform and the second lower limit value waveform, can be compared with the confirmation feature point. As a result, it becomes possible to determine the validity of the second upper limit value waveform and the second lower limit value waveform created by the data processing device 50, and by extension, further improve the accuracy of monitoring the measurement data in the data monitoring system 10. Is possible.

Further, in the data processing device 50 of the present embodiment, the monitoring waveform creation unit 66 has a second upper limit waveform and a second upper limit waveform based on the maximum value and the minimum value for each time point over one or more cycles of the measurement data of one abnormal cycle. Even when the second lower limit waveform is created, the measurement data used to obtain the maximum and minimum values is the measurement data included in the second post-processing measurement data group (thus, the abnormal waveform is removed). is there. Therefore, it is possible to create an appropriate second upper limit waveform and second lower limit waveform based on the above maximum and minimum values without being affected by the abnormal waveform.

Further, in the data monitoring system 10 of the present embodiment, at each of the plurality of measurement sites 20, the common second upper limit value waveform and the second lower limit value waveform created by the data processing device 50 are set to the respective monitoring upper limit values. Since the measurement data is monitored as the waveform and the lower limit waveform for monitoring, the measurement data at each measurement site 20 (for example, factories provided in a plurality of different countries such as Japan, the United States, and China) It is possible to promote the standardization of monitoring work and improve the uniformity of operation of the measurement target. For example, when each measurement site 20 is a factory and the measurement target is a machine tool, this reduces variations in the quality of products manufactured from the workpiece processed by the machine tool between factories. Is possible.

In particular, in the data monitoring system 10 of the present embodiment, since it is possible to use a plurality of monitoring upper limit waveforms and monitoring lower limit waveforms having various stages with respect to the allowable range, the measurement data at each measurement site 20 In the monitoring of, not only the standardization of normal / abnormal judgment is promoted, but also the normal measurement data is leveled more finely by using multiple monitoring upper limit waveforms and monitoring lower limit waveforms with stricter conditions. It becomes possible to divide. This makes it possible to easily and accurately compare and discriminate the positioning of the characteristics of each measurement site 20 regarding the measurement target (for example, the ability of the machine tool and its operation) within the plurality of measurement sites 20. Become.

The present invention is particularly advantageous as a data monitoring system 10 including a plurality of measurement sites 20 as described above, but the data processing device 50 according to the present invention is a data monitoring system 10 having only one measurement site 20. It can also be used, and even in that case, it is possible to improve the efficiency of operation of the measurement target while maintaining the accuracy of monitoring the measurement data.

Next, referring to FIGS. 23 to 25, as a second example of the data monitoring system 10 in the present embodiment, when the data monitoring system 10 is applied to the measurement data representing the movement of the foot in the dance step. An example of is described. The measurement data in this example is an example of measurement data representing the movement of a predetermined part of the human body. In this case, the measurement site 20 is, for example, a dance practice place such as a dance classroom or a home, and for example, one or more individuals belonging to the same practice place are measurement targets at each measurement site 20. In this case, the measuring device 21 is a vibration (acceleration) sensor attached to an individual's foot performing a dance step, preferably a wireless vibration (acceleration) sensor. In this case, the input unit 31 of the data monitoring device 30 has a wireless reception function, and receives the measurement data output from the acceleration sensor using this wireless reception function.

Here, the measurement site 20 (practice place) in this example may have a plurality of measurement sites 20 dispersed in various countries, regions, etc., as in the above-mentioned example. Since this example has the same configuration and function as the first example described above with reference to FIGS. 1 to 22, the description of the overlapping portion is omitted below. Now, the features unique to this example will be mainly described.

FIG. 23 shows an example of a dance step. The symbols (I) to (VI) shown correspond to the order of foot movements performed in this step, and the foot movements starting from (I) and ending at (VI) correspond to one cycle. To do. In this example, the input unit 31 of the data monitoring device 30 samples the measurement data output from the measuring device 21 from 1 point to 2500 points (that is, for 7.5 s) at a sampling interval of 3 ms, and further exercises. In order to extract the characteristics of the sampled digital data more clearly, the effective value calculation shall be performed in real time on the sampled digital data. Such measurement data is stored for a plurality of cycles in the storage unit 32 of the data monitoring device 30 at each measurement site 20.

FIG. 24 shows the first mean value waveform (m) created by the first processing unit 62 of the data processing device 50, the first positive standard deviation waveform (+ 1σ) with the default positive number set to 1, and the first standard deviation waveform (+ 1σ). 1 Negative standard deviation waveform (-1σ) is shown. Further, the time ranges (I) to (VI) shown in FIG. 24 correspond to the footsteps shown in FIG. 23, respectively, and the first candidate range (I) to the sixth candidate range (VI) in this example. ).

In this example, under the above settings, a second upper limit waveform and a second lower limit waveform are created in the data processing device 50 in the same manner as in the first example described above, and these second upper limit waveforms are created. And the second lower limit waveform is used as the upper limit waveform for monitoring and the lower limit waveform for monitoring at each measurement site 20. FIG. 25 shows a second mean waveform (2nd mean value waveform) created by the monitoring waveform creation unit 66 of the data processing device 50 from the stand-alone monitoring measurement data group defined in relation to the fifth candidate range (V). M), and the second positive standard deviation waveform (+ 1Σ) and the second negative standard deviation waveform (-1Σ) with the default positive number set to 1 are displayed. In this example, the second positive standard deviation waveform (+ 1Σ) and the second negative standard deviation waveform (-1Σ) are the second upper limit waveform and the second lower limit waveform, respectively.

The data monitoring system 10 in this example has the same effect as the first example described above in monitoring the measurement data at each measurement site 20 (for example, for the purpose of correcting the dance step). In this example, due to the nature of the measurement target, the number of measurement data acquired at each measurement site 20 is a decimal number as compared with, for example, the first example described above and the third example described later. On the other hand, it is expected that there are a huge number of possible measurement sites 20 in the world. In that sense, the data monitoring system 10 according to the present invention, which aggregates the measurement data from each measurement site 20 by the data processing device 50 and provides a monitoring upper limit waveform and a lower limit waveform common to each measurement site 20. Is also applied advantageously in this example.

Next, referring to FIGS. 26 and 27, as a third example of the data monitoring system 10 in the present embodiment, a case where the data monitoring system 10 is applied to the measurement data in the inspection process of the movement of the robot arm. An example of is described. The measurement data in this example is one of the measurement data different from the first example, which represents an amount that fluctuates with time according to the operation of the device. In this case, the measurement site 20 may be, for example, a manufacturing factory where the robot to be measured is manufactured, or a place where the robot is installed for use.

Since this example has the same configuration and function as the first example described above with reference to FIGS. 1 to 22, the description of the overlapping portion is omitted below. Now, the features unique to this example will be mainly described.

Here, although this example is an example of an amount that fluctuates with time according to the operation of the device as described above, the second item is that the target measurement data is the measurement data representing the movement of the movable element. It has the same characteristics as the examples. As a reflection of this, the measuring device 21 in this example is also a vibration (acceleration) sensor attached to a predetermined position of the arm of the robot, preferably a vibration (wireless) acceleration sensor. In this case, the input unit 31 of the data monitoring device 30 has a wireless reception function, and receives the measurement data output from the acceleration sensor using the wireless reception function. Although the present invention does not depend on the configuration of the vibration (acceleration) sensor, the vibration (acceleration) sensor includes, for example, a thin piezo wire wound with a piezoelectric film attached to a robot arm. May be good.

Further, also in this example, the input unit 31 of the data monitoring device 30 samples the measurement data output from the measuring device 21 and with respect to the sampled digital data in order to extract the characteristics of the motion more clearly. The effective value calculation is executed in real time. In FIG. 26, the measured individual measurement data stored in the first storage unit 61 of the data processing device 50 are the first mean value waveform (m) created by the first processing unit 62 and the first one. It is shown with the candidate range (I) and the second candidate range (II).

In this example, the second upper limit waveform and the second lower limit waveform are created in the data processing device 50 in the same manner as in the first and second examples described above under the above settings, and the second upper limit waveform and the second lower limit waveform are created. The upper limit waveform and the second lower limit waveform are used as the monitoring upper limit waveform and the monitoring lower limit waveform at each measurement site 20. In FIG. 27, in the monitoring waveform creation unit 66 of the data processing device 50, based on the second post-processing measurement data group, the measurement data of one or more cycles at a plurality of time points within the time range of one cycle. The maximum and minimum values for each time point over one or more cycles are calculated, and the second upper limit waveform (MAX) created based on the obtained maximum values at multiple time points is used as the minimum value at multiple time points obtained. Second lower limit waveform (MIN) created based on, at multiple time points within the time range of one cycle of the second post-processing measurement data group, at each time point of measurement data of one or more cycles over one or more cycles A waveform (M') consisting of a series of average values of is shown.

The data monitoring system 10 in this example has the same effect as the first example described above in monitoring the measurement data (for the purpose of inspecting the movement of the robot arm) at each measurement site 20.

10: Data monitoring system, 20: Measurement site, 21: Measurement equipment, 22: External equipment, 23: External equipment, 30: Data monitoring device, 31: Input unit, 32: Storage unit, 33: Calculation unit, 34: Monitoring Unit, 35: Communication unit, 40: Communication line, 50: Data processing device, 51: Input unit, 52: Output unit, 53: Input / output management unit, 61: First storage unit, 62: First processing unit, 63 : 2nd storage unit, 64: 2nd processing unit, 65: 3rd storage unit, 66: monitoring waveform creation unit, 67: 4th storage unit, 71: condition setting unit, 72: waveform analysis and display unit, 80 : Data sorting reference area

(1) A data processing device for monitoring an amount of measurement data that fluctuates with time so as to repeat one cycle from the start time to the end time, and is a first processing unit, a second processing unit, a condition setting unit, and monitoring. It includes a waveform creation unit, a first storage unit, a second storage unit, and a third storage unit.
The first storage unit stores measurement data of a plurality of cycles and stores the measurement data.
The first processing unit creates a first upper limit value waveform and a first lower limit value waveform in a time range of one cycle based on the measurement data of a plurality of cycles stored in the first storage unit.
The condition setting unit sets one or more data sorting reference areas that are a part of the two-dimensional area in a two-dimensional area composed of a time range of one cycle and the entire range of measured values, and the data. In each of the sorting reference areas, one time point in the data sorting reference area is set as a feature point, one value of the measured value is set as a sorting level, and according to the data sorting reference area and the sorting level, A sorting range is set as a time range including the feature points, and the sorting range is set.
The first processing unit further, for each of the one or more data sorting reference regions, is within the range of the sorting range before the feature point from the measurement data of a plurality of cycles stored in the first storage unit. The first front measurement data in which the intersection of the waveform corresponding to the measurement data and the sorting level exists, and the intersection of the waveform corresponding to the measurement data and the sorting level within the range after the feature point in the sorting range. The first rear side measurement data in which the data is present, and the intersection of the waveform corresponding to the measurement data and the sorting level exists on both sides of the sorting range both before and after the feature point. The measurement data is selected, and the selected first front side measurement data, the first rear side measurement data, and the first both side measurement data are stored in the second storage unit.
The first processing unit further removes the measurement data determined to be abnormal based on the first upper limit value waveform and the first lower limit value waveform from the measurement data stored in the second storage unit, and the rest. A first post-processing measurement data group consisting of the measurement data of the above is created, and the first post-processing measurement data group is stored in the second storage unit.
The second processing unit selects measurement data from the measurement data of a plurality of cycles stored in the first storage unit based on the measurement data group after the first processing, and the second processing unit is composed of the selected measurement data. A post-processing measurement data group is created, and the second post-processing measurement data group is stored in the third storage unit.
The monitoring waveform creation unit is based on the second post-processing measurement data group in order to monitor the measurement data acquired after the period in which the plurality of measurement data stored in the first storage unit is acquired. A data processing apparatus characterized in that it is configured to create one or more second upper limit waveforms and one or more second lower limit waveforms in a time range of one cycle.

(3) In the data processing apparatus according to item (2), each of the measurement data received from the input unit is stored in the first storage unit together with date and time information indicating the date and time when the measurement data was received by the input unit. Saved
The first processing unit obtains the mean value and standard deviation of the measured values of the measurement data of the plurality of cycles stored in the first storage unit at each time point over the plurality of cycles at a plurality of time points within the time range of one cycle. The first mean waveform, which is calculated and the calculated mean values of multiple time points, is added to the mean value of each of the multiple time points plus the default positive multiple of the standard deviation of the time point. Create a first positive standard deviation waveform and a first negative standard deviation waveform that is a series of values obtained by subtracting the default positive multiple of the standard deviation at the relevant time point from the mean value of each of the multiple time points. ,
For each of the one or more candidate ranges set within the time range of one cycle, the condition setting unit has a fluctuation start time point in which the time fluctuation of the first mean value waveform satisfies a predetermined fluctuation start condition, and the first. The time variation of the mean value waveform determines the fluctuation end time point that satisfies a predetermined fluctuation end time condition, and the fluctuation start time point and the fluctuation end time point, the first mean value waveform, the first positive standard deviation waveform, and The data sorting reference region is set based on at least one of the first negative standard deviation waveforms, and the data sorting reference region spans a plurality of cycles at each time point from the fluctuation start time to the fluctuation end time. The feature points and the sorting level are set based at least partially based on the distribution of the values of the measurement data and one or both of the time series patterns in which the values of the measurement data are arranged according to the date and time information.
The data sorting reference region is a two-dimensional region composed of a time range of one cycle and the entire range of measured values, the value of the first negative standard deviation waveform at the start of the fluctuation and the value of the standard deviation waveform at the end of the fluctuation. The lower side that linearly connects the values of the first negative standard deviation waveform, the value of the first positive standard deviation waveform at the start of the fluctuation, and the value of the first positive standard deviation waveform at the end of the fluctuation are linearly connected. The region between the upper side connected in a shape, or the value of the first positive standard deviation waveform at the start of the fluctuation and the value of the first positive standard deviation waveform at the end of the fluctuation are linearly connected. lower and the region between the upper side and consisting of a first positive standard deviation waveform or, the value of the first average value waveform in the value and the variation end point of the first average value waveform in the variation start point The region between the lower side linearly connecting the above and the upper side composed of the first average value waveform, or the value of the first negative standard deviation waveform at the start of the fluctuation and the first at the end of the fluctuation. Includes a region between the lower side that linearly connects the values with the negative standard deviation waveform and the upper side that consists of the first negative standard deviation waveform.
The sorting range is a two-dimensional region composed of a time range of one cycle and the entire range of measured values, and at two time points in which the value in the measured value direction takes the sorting level on the boundary of the data sorting reference region. A data processing device characterized in that it is configured to be a time range between.

(5) (3) or the data process unit according to any one of (4), said data processing apparatus further includes a fourth storage unit,
The second processing unit receives the first front measurement data or the first front measurement for each of one or more selected data sorting reference regions among the measurement data included in the first post-processing measurement data group. Regarding the measurement data consisting of the measurement data satisfying the predetermined time fluctuation condition among the data and the measurement data satisfying the predetermined time fluctuation condition among the first rear side measurement data, the waveform corresponding to the measurement data is the said in the sorting range. The average value and standard deviation of the time values that intersect the sorting level within the range before the feature point are calculated, and the average value is obtained by subtracting the default positive multiple of the standard deviation from the average value. On the other hand, the front time threshold range up to the value obtained by adding the default positive multiple of the standard deviation is set.
Of the measurement data included in the first post-processing measurement data group, for each of the selected one or more data sorting reference regions, a predetermined one of the first rear measurement data or the first front measurement data. Regarding the measurement data consisting of the measurement data satisfying the time fluctuation condition and the measurement data satisfying the predetermined time fluctuation condition among the first rear side measurement data, the waveform corresponding to the measurement data is the range after the feature point in the sorting range. The average value and standard deviation of the value of the time when the sorting level intersects with the sorting level are calculated, and the standard deviation with respect to the average value is obtained by subtracting the predetermined positive multiple of the standard deviation from the average value. Set the rear time threshold range up to the value obtained by adding the default positive multiples of
For each of the one or more data sorting reference areas selected from the data sorting reference area in which the front time threshold range and the rear time threshold range are set, the measurement data of a plurality of cycles stored in the first storage unit. Among them, the second front measurement data in which an intersection satisfying a predetermined time fluctuation condition between the waveform corresponding to the measurement data and the sorting level exists within the front time threshold range, and the measurement data within the rear time threshold range. Second rear measurement data in which there is an intersection satisfying a predetermined time fluctuation condition between the waveform corresponding to the above and the sorting level, and measurement data in both the front time threshold range and the rear time threshold range. The second-sided measurement data in which the intersection of the waveform corresponding to the above and the sorting level exists is selected, and all the second front-side measurement data, the second rear-side measurement data, and the second-sided measurement data are subjected to the first 3 Save in the storage
The second processing unit further removes the measurement data determined to be abnormal based on the first upper limit value waveform and the first lower limit value waveform from the measurement data stored in the third storage unit, and the rest. A second post-processing measurement data group consisting of the measurement data of the above is created, and the second post-processing measurement data group is stored in the third storage unit.
The second processing unit further measures after the second processing for each one or more data sorting reference regions selected from the data sorting reference regions in which the front time threshold range and the rear time threshold range are set. The average value and standard deviation of the value of the time when the waveform corresponding to the second two-sided measurement data included in the data group intersects the sorting level within the sorting range is calculated, and the standard deviation is calculated with respect to the average value. A two-sided time threshold range is set from a value obtained by subtracting a predetermined positive number multiple to a value obtained by adding a predetermined positive multiple of the standard deviation to the average value, and from the second post-processing measurement data group, the said Measurement data having an intersection that satisfies a predetermined time fluctuation condition between the waveform corresponding to the measurement data and the sorting level within the two-sided time threshold range is selected, and the selected measurement data is stored in the third storage unit.
The monitoring waveform creation unit selects the second two-sided measurement data related to any combination of one or more of the data sorting reference regions from the measurement data included in the second post-processing measurement data group. A monitoring measurement data group consisting of selected measurement data is created, the monitoring measurement data group is stored in the fourth storage unit, and based on the monitoring measurement data group stored in the fourth storage unit. , The data processing apparatus characterized in that the second upper limit value waveform and the second lower limit value waveform are created.

(7) In the data processing apparatus according to item (6), the monitoring measurement data group is one selected data sorting from the second side measurement data included in the second post-processing measurement data group. A stand-alone monitoring measurement data group consisting of measurement data in which an intersection that satisfies a predetermined time variation condition between the waveform corresponding to the measurement data and the sorting level exists within the two-sided time threshold range related to the reference region, and a measurement data group for independent monitoring. Among the second two-sided measurement data included in the second post-processing measurement data group, the measurement data corresponds to at least one range of the two-sided time threshold range related to the plurality of selected data sorting reference regions. A Japanese-style monitoring measurement data group consisting of measurement data in which an intersection satisfying a predetermined time variation condition between the waveform and the sorting level exists, and the second-sided measurement data included in the second post-processing measurement data group. From the measurement data in which, within all the two-sided time threshold ranges related to the plurality of selected data sorting reference regions, there is an intersection that satisfies a predetermined time variation condition between the waveform corresponding to the measurement data and the sorting level. It consists of a group of measurement data for product type monitoring.
The monitoring waveform creation unit has, for each of the stand-alone monitoring measurement data group, the Japanese-style monitoring measurement data group, and the product-type monitoring measurement data group, the second average value waveform and the second positive. A plurality of created side standard deviation waveforms and the second negative standard deviation waveform are created, and at least one of the created plurality of the second positive standard deviation waveforms is set as the second upper limit value waveform. A data processing apparatus, wherein at least one of the second negative standard deviation waveforms is used as the second lower limit value waveform.

The data processing apparatus according to any one of (8) to (5) (7), the second processing unit, for each of said data sorting reference area of one or more selected, the front time threshold value range And the rear side time threshold range, a plurality of the predetermined positive numbers used when setting the front side time threshold value range and the rear side time threshold value range, according to the plurality of the predetermined positive numbers. Set for each stage of
The monitoring waveform creation unit creates the monitoring measurement data group for any combination of one or more of the data sorting reference regions at each of the plurality of stages and stores the data in the fourth storage unit. (4) Based on the monitoring measurement data group stored in the storage unit, the second upper limit value waveform and the second lower limit value waveform are created for each of the plurality of steps. Data processing device.

(10 ) In the data processing apparatus described in item ( 9 ) , the condition setting unit is the measurement data of one or more cycles included in the second post-processing measurement data at a plurality of time points within the time range of one cycle. The average value and standard deviation of the measured values for each time point over one or more cycles are calculated, and the average value waveform for confirmation, which is a series of the calculated average values of multiple time points, and the average value of each of the multiple time points are used. On the other hand, the positive standard deviation waveform for confirmation, which is a series of values obtained by adding the default positive multiples of the standard deviation at that time point, and the default positive number of the standard deviation at that time point for each average value of multiple time points. A confirmation negative standard deviation waveform is created by subtracting multiples, and for each set data sorting reference region, one or more cycles of each time point from the start time of the change to the end time of the change are extended. It is configured to set confirmation feature points based on at least partly the distribution of the values of the measurement data and one or both of the time series patterns in which the values of the measurement data are arranged according to the date and time information.
The waveform analysis and display unit is a distribution diagram of values of measurement data of one or more cycles included in the measurement data after the second processing at a time point on the vertical axis parallel line passing through an arbitrary time point within the candidate range. A time-series pattern diagram in which the values of the measurement data of the one or more cycles are arranged according to the date and time information can be displayed, and the average is based on the calculated average value and standard deviation of the values over the one or more cycles. A horizontal parallel line passing through a value, a horizontal parallel line passing through a value obtained by adding a predetermined positive multiple of the standard deviation to the average value, and a default positive number of the standard deviation with respect to the average value. A data processing device characterized in that a horizontal axis parallel line passing through a value obtained by subtracting a double can be displayed.

(15) A plurality of measurement sites, the data processing apparatus according to any one of (11) to (14) quoting (2) to (10) and (2) , the measurement site and the data processing. A data monitoring system that includes a communication line that connects to the device.
Each of the measurement sites includes a measurement device that acquires measurement data of an amount that fluctuates with time so as to repeat one cycle from a start time point to an end point point, and a data monitoring device corresponding to the measurement device. The measurement data acquired at each of the measurement sites is the measurement data obtained by measuring the same amount at all the measurement sites.
The data monitoring device includes an input unit that receives measurement data from the corresponding measuring device, a storage unit that can store measurement data of a plurality of cycles, a monitoring upper limit waveform and a monitoring lower limit value in a time range of one cycle. Including a monitoring unit that monitors measurement data from the corresponding measuring device based on the waveform.
Each of the data monitoring devices included in the plurality of measurement sites transmits measurement data of a plurality of cycles to the data processing device via the communication line.
The data processing device outputs the second upper limit value waveform and the second lower limit value waveform obtained based on the measurement data of the plurality of cycles received from one or more of the data monitoring devices from the output unit to the communication line. Is transmitted to each of the data monitoring devices included in the plurality of measurement sites via
The monitoring unit of the data monitoring device included in the plurality of measurement sites receives the second upper limit value waveform and the second lower limit value waveform received from the data processing device, and measures data acquired after the reception. A data monitoring system characterized in that it is configured to be used as the monitoring upper limit waveform and the monitoring lower limit waveform for monitoring.

Claims (17)

A data processing device for monitoring the amount of measurement data that fluctuates with time so as to repeat one cycle from the start time to the end time, and is a first processing unit, a second processing unit, a condition setting unit, and a waveform creation for monitoring. A unit, a first storage unit, a second storage unit, and a third storage unit are included.
The first storage unit stores measurement data of a plurality of cycles and stores the measurement data.
The first processing unit creates a first upper limit value waveform and a first lower limit value waveform in a time range of one cycle based on the measurement data of a plurality of cycles stored in the first storage unit.
The condition setting unit sets one or more data sorting reference areas that are a part of the two-dimensional area in a two-dimensional area composed of a time range of one cycle and the entire range of measured values, and the data. In each of the sorting reference areas, one time point in the data reference area is set as a feature point, one value of the measured value is set as a sorting level, and the data sorting reference area and the sorting level are set according to the above. Set the sorting range as the time range including the feature points,
The first processing unit further, for each of the one or more data sorting reference regions, is within the range of the sorting range before the feature point from the measurement data of a plurality of cycles stored in the first storage unit. The first front measurement data in which the intersection of the waveform corresponding to the measurement data and the sorting level exists, and the intersection of the waveform corresponding to the measurement data and the sorting level within the range after the feature point in the sorting range. The first rear side measurement data in which the data is present, and the intersection of the waveform corresponding to the measurement data and the sorting level exists on both sides of the sorting range both before and after the feature point. The measurement data is selected, and the selected first front side measurement data, the first rear side measurement data, and the first both side measurement data are stored in the second storage unit.
The first processing unit further removes the measurement data determined to be abnormal based on the first upper limit value waveform and the first lower limit value waveform from the measurement data stored in the second storage unit, and the rest. A first post-processing measurement data group consisting of the measurement data of the above is created, and the first post-processing measurement data group is stored in the second storage unit.
The second processing unit selects measurement data from the measurement data of a plurality of cycles stored in the first storage unit based on the measurement data group after the first processing, and the second processing unit is composed of the selected measurement data. A post-processing measurement data group is created, and the second post-processing measurement data group is stored in the third storage unit.
The monitoring waveform creation unit is based on the second post-processing measurement data group in order to monitor the measurement data acquired after the period in which the plurality of measurement data stored in the first storage unit is acquired. A data processing apparatus characterized in that it is configured to create one or more second upper limit waveforms and one or more second lower limit waveforms in a time range of one cycle. 1. The first aspect of the present invention is characterized in that it further includes an input unit that receives measurement data stored in the first storage unit from the outside, and an output unit that sends out the second upper limit value waveform and the second lower limit value waveform to the outside. The data processing apparatus described in 1. Each of the measurement data received from the input unit is stored in the first storage unit together with date and time information indicating the date and time when the measurement data was received by the input unit.
The first processing unit obtains the mean value and standard deviation of the measured values of the measurement data of the plurality of cycles stored in the first storage unit at each time point over the plurality of cycles at a plurality of time points within the time range of one cycle. The first mean waveform, which is calculated and the calculated mean values of multiple time points, is added to the mean value of each of the multiple time points plus the default positive multiple of the standard deviation of the time point. Create a first positive standard deviation waveform and a first negative standard deviation waveform that is a series of values obtained by subtracting the default positive multiple of the standard deviation at the relevant time point from the mean value of each of the multiple time points. ,
For each of the one or more candidate ranges set within the time range of one cycle, the condition setting unit has a fluctuation start time point in which the time fluctuation of the first mean value waveform satisfies a predetermined fluctuation start condition, and the first. The time variation of the mean value waveform determines the fluctuation end time point that satisfies a predetermined fluctuation end time condition, and the fluctuation start time point and the fluctuation end time point, the first mean value waveform, the first positive standard deviation waveform, and The data sorting reference region is set based on at least one of the first negative standard deviation waveforms, and the data sorting reference region spans a plurality of cycles at each time point from the fluctuation start time to the fluctuation end time. The feature points and the sorting level are set based at least partially based on the distribution of the values of the measurement data and one or both of the time series patterns in which the values of the measurement data are arranged according to the date and time information.
The data sorting reference region is a two-dimensional region composed of a time range of one cycle and the entire range of measured values, the value of the first negative standard deviation waveform at the start of the fluctuation and the value of the standard deviation waveform at the end of the fluctuation. The lower side that linearly connects the values of the first negative standard deviation waveform, the value of the first positive standard deviation waveform at the start of the fluctuation, and the value of the first positive standard deviation waveform at the end of the fluctuation are linearly connected. The region between the upper side connected in a shape, or the value of the first positive standard deviation waveform at the start of the fluctuation and the value of the first positive standard deviation waveform at the end of the fluctuation are linearly connected. The region between the lower side and the upper side consisting of the first positive standard deviation waveform, or the value of the first average value waveform at the start of the fluctuation and the value of the first average value waveform at the end of the fluctuation. The region between the lower side linearly connecting the two and the upper side composed of the first average value waveform, or the value of the first negative standard deviation waveform at the start of the fluctuation and the first negative at the end of the fluctuation. Includes a region between the lower side that linearly connects the values with the side standard deviation waveform and the upper side that consists of the first negative standard deviation waveform.
The sorting range is a two-dimensional region composed of a time range of one cycle and the entire range of measured values, and at two time points in which the value in the measured value direction takes the sorting level on the boundary of the data sorting reference region. The data processing apparatus according to claim 2, wherein the data processing apparatus is configured such that the time range is between. The third processing unit is characterized in that the first positive side standard deviation waveform is the first upper limit value waveform and the first negative side standard deviation waveform is the first lower limit value waveform. The data processing device described. The data monitoring device further includes a fourth storage unit.
The second processing unit receives the first front measurement data or the first front measurement for each of one or more selected data sorting reference regions among the measurement data included in the first post-processing measurement data group. Regarding the measurement data consisting of the measurement data satisfying the predetermined time fluctuation condition among the data and the measurement data satisfying the predetermined time fluctuation condition among the first rear side measurement data, the waveform corresponding to the measurement data is the said in the sorting range. The average value and standard deviation of the time values that intersect the sorting level within the range before the feature point are calculated, and the average value is obtained by subtracting the default positive multiple of the standard deviation from the average value. On the other hand, the front time threshold range up to the value obtained by adding the default positive multiple of the standard deviation is set.
Of the measurement data included in the first post-processing measurement data group, for each of the selected one or more data sorting reference regions, a predetermined one of the first rear measurement data or the first front measurement data. Regarding the measurement data consisting of the measurement data satisfying the time fluctuation condition and the measurement data satisfying the predetermined time fluctuation condition among the first rear side measurement data, the waveform corresponding to the measurement data is the range after the feature point in the sorting range. The average value and standard deviation of the value of the time when the sorting level intersects with the sorting level are calculated, and the standard deviation with respect to the average value is obtained by subtracting the predetermined positive multiple of the standard deviation from the average value. Set the rear time threshold range up to the value obtained by adding the default positive multiples of
For each of the one or more data sorting reference areas selected from the data sorting reference area in which the front time threshold range and the rear time threshold range are set, the measurement data of a plurality of cycles stored in the first storage unit. Among them, the second front measurement data in which an intersection satisfying a predetermined time fluctuation condition between the waveform corresponding to the measurement data and the sorting level exists within the front time threshold range, and the measurement data within the rear time threshold range. Second rear measurement data in which there is an intersection satisfying a predetermined time fluctuation condition between the waveform corresponding to the above and the sorting level, and measurement data in both the front time threshold range and the rear time threshold range. The second-sided measurement data in which the intersection of the waveform corresponding to the above and the sorting level exists is selected, and all the second front-side measurement data, the second rear-side measurement data, and the second-sided measurement data are subjected to the first 3 Save in the storage
The second processing unit further removes the measurement data determined to be abnormal based on the first upper limit value waveform and the first lower limit value waveform from the measurement data stored in the third storage unit, and the rest. A second post-processing measurement data group consisting of the measurement data of the above is created, and the second post-processing measurement data group is stored in the third storage unit.
The second processing unit further measures after the second processing for each one or more data sorting reference regions selected from the data sorting reference regions in which the front time threshold range and the rear time threshold range are set. The average value and standard deviation of the value of the time when the waveform corresponding to the second two-sided measurement data included in the data group intersects the sorting level within the sorting range is calculated, and the standard deviation is calculated with respect to the average value. A two-sided time threshold range is set from a value obtained by subtracting a predetermined positive number multiple to a value obtained by adding a predetermined positive multiple of the standard deviation to the average value, and from the second post-processing measurement data group, the said Measurement data having an intersection that satisfies a predetermined time fluctuation condition between the waveform corresponding to the measurement data and the sorting level within the two-sided time threshold range is selected, and the selected measurement data is stored in the third storage unit.
The monitoring waveform creation unit selects the second two-sided measurement data related to any combination of one or more of the data sorting reference regions from the measurement data included in the second post-processing measurement data group. A monitoring measurement data group consisting of selected measurement data is created, the monitoring measurement data group is stored in the fourth storage unit, and based on the monitoring measurement data group stored in the fourth storage unit. The data processing apparatus according to any one of claims 2 to 4, wherein the second upper limit value waveform and the second lower limit value waveform are created. The monitoring waveform creation unit calculates the average value and standard deviation of the measured values at each time point over one or more cycles of one or more measurement data included in the monitoring measurement data group, and at a plurality of calculated time points. A second average value waveform in which the average values are connected, and a second positive standard deviation waveform in which the values obtained by adding the default positive multiples of the standard deviation at the relevant time points to the average values at each of a plurality of time points are connected. A second negative standard deviation waveform is created by subtracting a predetermined positive multiple of the standard deviation at the relevant time point from the average value of each of the multiple time points, and the second positive standard deviation waveform is used. The data processing apparatus according to claim 5, wherein the second upper limit value waveform is used, and the second negative standard deviation waveform is used as the second lower limit value waveform. The monitoring measurement data group includes measurement data within the two-sided time threshold range related to one selected data sorting reference region among the second two-sided measurement data included in the second post-processing measurement data group. A stand-alone monitoring measurement data group consisting of measurement data having an intersection satisfying a predetermined time variation condition between the waveform corresponding to the above and the sorting level, and the second both sides included in the second post-processing measurement data group. Of the measurement data, an intersection that satisfies a predetermined time variation condition between the waveform corresponding to the measurement data and the sorting level within at least one range of the two-sided time threshold range related to the plurality of selected data sorting reference regions. Of the Japanese-style monitoring measurement data group consisting of measurement data in which is present and the second-sided measurement data included in the second post-processing measurement data group, all related to a plurality of selected data sorting reference regions. It consists of a product-type monitoring measurement data group consisting of measurement data in which an intersection that satisfies a predetermined time fluctuation condition between the waveform corresponding to the measurement data and the sorting level exists within the two-sided time threshold range.
The monitoring waveform creation unit has, for each of the stand-alone monitoring measurement data group, the sum-type measurement data group, and the product-type monitoring measurement data group, the second mean value waveform and the second positive standard. The deviation waveform and the second negative standard deviation waveform are created, and at least one of the created plurality of the second positive standard deviation waveforms is set as the second upper limit value waveform, and the plurality of created above. The data processing apparatus according to claim 6, wherein at least one of the second negative standard deviation waveforms is the second lower limit value waveform. When the second processing unit sets the front time value range and the rear time threshold range, and the front time threshold range and the rear time threshold range for each of one or more selected data sorting reference regions. Using a plurality of the default positive numbers used in the above, set for each of a plurality of steps according to the plurality of the default positive numbers.
The monitoring waveform creation unit creates the monitoring measurement data group for any combination of one or more of the data sorting reference regions at each of the plurality of stages and stores the data in the fourth storage unit. (4) Based on the monitoring measurement data group stored in the storage unit, the second upper limit value waveform and the second lower limit value waveform are created for each of the plurality of steps. The data processing apparatus according to any one of claims 5 to 7. It further includes a waveform analysis and display unit configured to display measurement data with the time direction as the horizontal axis and the measurement value direction of the measurement data as the vertical axis.
The waveform analysis and display unit includes the first mean value waveform, the first positive standard deviation waveform, and the first negative standard deviation waveform in the candidate range for each of the one or more candidate ranges. It is configured to be able to display a vertical parallel line that can move in the time direction automatically or by input from the operator, passing through any time point within the candidate range.
The waveform analysis and display unit further displays a distribution diagram of the values of the measurement data of the plurality of cycles stored in the first storage unit at the time on the vertical axis parallel line and the values of the measurement data of the plurality of cycles as the date and time information. The time-series pattern diagram arranged according to the above is configured to be displayable, and based on the calculated mean value and standard deviation of the values over the plurality of cycles, the horizontal axis parallel line passing through the mean value, the standard with respect to the mean value. The horizontal axis parallel line passing through the value obtained by adding the default positive multiple of the deviation and the horizontal axis parallel line passing through the value obtained by subtracting the default positive multiple of the standard deviation from the mean value can be displayed. Being done
The waveform analysis and display unit further includes measurement data satisfying a predetermined time variation condition among the first front measurement data or the first front measurement data with respect to each of the one or more data sorting reference regions. Regarding the measurement data consisting of the measurement data satisfying the predetermined time fluctuation condition among the first rear side measurement data, the time when the waveform corresponding to the measurement data intersects the sorting level within the range before the feature point in the sorting range. It is configured to be able to display a distribution map of the values of and a time series pattern diagram in which the values of the time are arranged according to the date and time information, and is parallel to the vertical axis passing through the average value based on the average value and the standard deviation of the values of the time. A line, a vertical parallel line passing through a value obtained by adding a default positive multiple of the standard deviation to the average value, and a value obtained by subtracting a default positive multiple of the standard deviation from the average value. It is configured so that the vertical parallel lines that pass through can be displayed.
The waveform analysis and display unit further includes, for each of the one or more data sorting reference regions, the first rear measurement data or the measurement data satisfying a predetermined time variation condition among the first front measurement data. With respect to the measurement data consisting of the measurement data satisfying a predetermined time fluctuation condition among the first rear side measurement data, the waveform corresponding to the measurement data intersects the sorting level within the range after the feature point in the sorting range. It is configured to be able to display a distribution map of time values and a time series pattern diagram in which the time values are arranged according to the date and time information, and a vertical axis passing through the average values based on the average value and standard deviation of the time values. Parallel lines, vertical parallel lines passing through a value obtained by adding a default positive multiple of the standard deviation to the average value, and a value obtained by subtracting a default positive multiple of the standard deviation from the average value. It is configured to be able to display vertical parallel lines passing through
The waveform analysis and display unit further includes, for each of the one or more data sorting reference regions, the first rear measurement data or the measurement data satisfying a predetermined time variation condition among the first front measurement data. With respect to the measurement data consisting of the measurement data satisfying a predetermined time fluctuation condition among the first rear side measurement data, the waveform corresponding to the measurement data intersects the sorting level within the range after the feature point in the sorting range. It is configured to be able to display a distribution map of time values and a time series pattern diagram in which the time values are arranged according to the date and time information, and a vertical axis passing through the average values based on the average value and standard deviation of the time values. Parallel lines, vertical parallel lines passing through a value obtained by adding a default positive multiple of the standard deviation to the average value, and a value obtained by subtracting a default positive multiple of the standard deviation from the average value. It is configured to be able to display vertical parallel lines passing through
In the waveform analysis and display unit, the waveform corresponding to the second side measurement data included in the second post-processing measurement data group is further within the sorting range for each of the one or more data sorting reference regions. Any one of claims 5 to 8, wherein a distribution map of time values that intersect the sorting level and a time series pattern diagram in which the time values are arranged according to the date and time information can be displayed. The data processing apparatus described in. The condition setting unit is an average value of the measured values of one or more cycles of the measurement data included in the second post-processing measurement data at a plurality of time points within the time range of one cycle at each time point over one or more cycles. And the standard deviation is calculated, and the confirmation average value waveform that connects the calculated average values at multiple time points and the default positive multiple of the standard deviation at that time point are added to each average value at multiple time points. The positive standard deviation waveform for confirmation, which is a series of the values, and the negative standard deviation waveform for confirmation, which is the average value of each of the multiple time points minus the default positive multiple of the standard deviation at that time point. And, for each of the set data sorting reference areas, the distribution of the measurement data values over one or more cycles from the fluctuation start time to the fluctuation end time and the value of the measurement data are set to the date and time. Configured to set confirmation feature points, at least in part, based on one or both of the informed time series patterns.
The waveform analysis and display unit is a distribution map of measurement data values of one or more cycles included in the second post-processing measurement data at a time point on the vertical axis parallel line passing through an arbitrary time point within the candidate range. A time-series pattern diagram in which the values of the measurement data of the one or more cycles are arranged according to the date and time information can be displayed, and the average is based on the calculated average value and standard deviation of the values over the one or more cycles. A horizontal parallel line passing through a value, a horizontal parallel line passing through a value obtained by adding a predetermined positive multiple of the standard deviation to the average value, and a default positive number of the standard deviation with respect to the average value. The data processing apparatus according to any one of claims 5 to 9, wherein a horizontal axis parallel line passing through a value obtained by subtracting a double is displayable. Based on the second post-processing measurement data group, the monitoring waveform creation unit is the maximum of the measurement data of one or more cycles at each time point over one or more cycles at a plurality of time points within the time range of one cycle. The value and the minimum value are obtained, the second upper limit waveform is created based on the obtained maximum values at a plurality of time points, and the second lower limit waveform is created based on the obtained minimum values at the plurality of time points. The data processing apparatus according to any one of claims 1 to 10, wherein the data processing apparatus is configured as described above. Any of claims 1 to 11, wherein the measurement data of the plurality of cycles stored in the first storage unit includes measurement data obtained by measuring the same amount with respect to a plurality of different measurement targets. The data processing apparatus according to item 1. The invention according to any one of claims 1 to 12, wherein the measurement data of the plurality of cycles stored in the first storage unit includes measurement data representing an amount that fluctuates with time according to the operation of the device. Data processing device. The data processing apparatus according to any one of claims 1 to 13, wherein the measurement data of the plurality of cycles stored in the first storage unit includes measurement data representing the movement of a predetermined part of the human body. .. A data monitoring system including a plurality of measurement sites, the data processing device according to any one of claims 2 to 14, and a communication line connecting the measurement site and the data processing device.
Each of the measurement sites includes a measurement device that acquires measurement data of an amount that fluctuates with time so as to repeat one cycle from a start time point to an end point point, and a data monitoring device corresponding to the measurement device. The measurement data acquired at each of the measurement sites is the measurement data obtained by measuring the same amount at all the measurement sites.
The data monitoring device includes an input unit that receives measurement data from the corresponding measuring device, a storage unit that can store measurement data of a plurality of cycles, a monitoring upper limit waveform and a monitoring lower limit value in a time range of one cycle. Including a monitoring unit that monitors measurement data from the corresponding measuring device based on the waveform.
Each of the data monitoring devices included in the plurality of measurement sites transmits measurement data of a plurality of cycles to the data processing device via the communication line.
The data processing device outputs the second upper limit value waveform and the second lower limit value waveform obtained based on the measurement data of the plurality of cycles received from one or more of the data monitoring devices from the output unit to the communication line. Is transmitted to each of the data monitoring devices included in the plurality of measurement sites via
The monitoring unit of the data monitoring device included in the plurality of measurement sites receives the second upper limit value waveform and the second lower limit value waveform received from the data processing device, and measures data acquired after the reception. A data monitoring system characterized in that it is configured to be used as the monitoring upper limit waveform and the monitoring lower limit waveform for monitoring. The data monitoring device further includes a calculation unit that creates an upper limit waveform and a lower limit waveform in a time range of one cycle based on the measurement data of a plurality of cycles stored in the storage unit of the data monitoring device. 15. The monitoring unit is characterized in that the upper limit waveform and the lower limit waveform created by the calculation unit can be used as the monitoring upper limit waveform and the monitoring lower limit waveform, respectively. Described data monitoring system. Further including an external device connected to the data monitoring device,
When the external device is an alarm device and the data monitoring device determines that the measurement data acquired from the corresponding measuring device is abnormal based on the monitoring upper limit value waveform and the monitoring lower limit value waveform. The data monitoring system according to claim 15 or 16, wherein the alarm device is configured to generate an alarm.