CN109643115B

CN109643115B - Data processing apparatus, data processing method, and program

Info

Publication number: CN109643115B
Application number: CN201880003289.9A
Authority: CN
Inventors: 丸地康平; 菊池元太; 服部阳平
Original assignee: Toshiba Corp
Current assignee: Toshiba Corp
Priority date: 2017-07-18
Filing date: 2018-02-20
Publication date: 2021-10-22
Anticipated expiration: 2038-02-20
Also published as: TWI710872B; CN109643115A; JP2019021056A; TW201908899A; WO2019016993A1; JP6829158B2

Abstract

A data processing apparatus as one aspect of the present invention includes an index calculator, an exclusion condition calculator, and a refiner. The index calculator calculates index data containing an index representing a state of the monitoring target based on the measurement data on the monitoring target. The exclusion condition calculator calculates an exclusion condition for removing any index related to the interference from the index data based on the measurement data. The refiner removes the index related to the disturbance from the index data based on the exclusion condition, thereby refining the index data.

Description

Data processing apparatus, data processing method, and program

Technical Field

Embodiments relate to a data processing apparatus, a data processing method, and a program.

Background

In order to continuously operate the system safely and stably, it is essential to check the reliability of the system and perform maintenance as needed. However, the higher the inspection frequency, the higher the maintenance cost becomes. Therefore, the next inspection plan is made based on the elapsed time from the last maintenance, the running time of the system, and the like. A maintenance method of making an inspection plan based on time in this manner is called TBM (time-based maintenance).

In recent years, with the progress of radio technology, the cost reduction of sensors, and the like, it has become mainstream to grasp the system state in real time by remote monitoring. If the current state of the system can be grasped, the check date can be postponed, unnecessary check items can be omitted, and the like. A maintenance method of making an inspection plan based on the current state of the system in this way is called CBM (condition-based maintenance). By shifting from the TBM to the CBM, maintenance costs are expected to be reduced.

However, various disturbances are liable to enter data measured in order to grasp the current state of the system during the operation of the system. Therefore, in order to grasp the current state of the system with high accuracy, unnecessary interference needs to be removed from the measurement data.

Drawings

Fig. 1 is a block diagram showing an example of a data processing apparatus according to a first embodiment;

FIG. 2 is a diagram depicting metric data;

fig. 3 is a diagram showing an example of an exclusion condition;

FIG. 4 is a diagram showing the processing results produced by the rejector and the deviation determiner;

FIG. 5 is a diagram showing an example of output generated by an output device;

fig. 6 is a diagram showing an example of a schematic flowchart of the entire process performed by the data processing apparatus according to the first embodiment;

fig. 7 is a block diagram showing an example of a schematic configuration of a data processing apparatus according to the second embodiment;

fig. 8 is a diagram describing a modification to the exclusion condition;

fig. 9 is a diagram showing an example of a flowchart of the exclusion condition update process;

fig. 10 is a block diagram showing an example of a schematic configuration of a data processing apparatus according to a third embodiment;

fig. 11 is a diagram showing an example of a result regarding counted sample counts;

fig. 12 is a diagram showing another example of a result regarding counted sample counts;

fig. 13 is a block diagram showing an example of a schematic configuration of a data processing apparatus according to a fourth embodiment;

FIG. 14 is a diagram depicting evaluation inputs with respect to state determination;

fig. 15 is a diagram showing an example of a history of correct answers to the state determination;

fig. 16 is a diagram showing an example of a flowchart of state determination; and

fig. 17 is a block diagram showing an example of a hardware configuration according to an embodiment of the present invention.

Detailed Description

One embodiment of the invention determines any interference contained in the measurement data about the monitored target.

A description is given below of embodiments of the present invention with reference to the accompanying drawings. The present invention is not limited to these examples.

(first embodiment)

Fig. 1 is a block diagram showing an example of a data processing apparatus according to a first embodiment. The data processing apparatus according to the first embodiment includes a measurement data acquirer 1, a memory 2, a data processor 3, and an output device 4. The data processor 3 includes an index calculator 31, an exclusion condition calculator 32, an excluder (refiner) 33, and a deviation determiner 34.

The data processing apparatus according to the present embodiment calculates index data containing an index representing a state of a monitoring target based on measurement data on the monitoring target. The index is used to detect any abnormality of the monitoring target, determine whether the monitoring target needs to be checked, and the like.

It is conceivable that the index to be calculated indicates a state that is difficult to measure with a sensor or the like. For example, the index may be one (performance index) indicating the performance state of the monitoring target.

The monitoring target is not particularly limited, and may be one device or a system composed of a plurality of devices. Further, the monitoring target may be a living body such as a human or an animal.

The measurement data refers to time-series data including measurement results generated by a sensor or the like. The measurement results include measurement time, measurement values, and the like. As the sensor, a known sensor can be used.

The measurement items of the measurement data may be items measurable by well-known sensors or the like. The measurement point may be the entire monitoring target or a specific portion of the monitoring target. Examples of possible measurement items include temperature, humidity, flow rate, current, voltage, pressure, and position. Further, when the monitoring target is a movable body such as a vehicle, it is conceivable that the speed, acceleration, and the like of the movable body will also be included in the measurement items.

It is assumed that the measurements for obtaining measurement data can be made at any time. Thus, the measurement data may include both data measured while the monitoring target is running and data measured while the monitoring target is stopped.

Incidentally, the measurement result produced by the monitoring target itself may also be included in the measurement data. In addition, the settings input to the monitoring target may also be included in the measurement data. For example, data indicating a period of time during which the monitoring target is powered on or off may also be included. For example, if the monitoring target has a power saving mode that reduces power consumption, data indicating a period during which the monitoring target is in the power saving mode may also be included in the measurement data. For example, if the monitoring target is an air conditioner, data indicating a state such as "cooling", "heating", or "humidification" may be included in the measurement data. If the monitoring target is a vehicle, data indicating conditions (such as "running", "accelerating", "decelerating", and "temporarily stopping") may be included in the measurement data.

The measurement data may also include a state of the monitoring target determined by the monitoring target itself or another external device based on the measurement data. For example, if the current flowing through the built-in motor of the monitor target is measured and the monitor target determines that the motor is abnormal based on the value of the measured current, the monitor target may add a value indicating any abnormality of the motor or the monitor target to the measurement data.

The index included in the index data is represented as a combination of at least an index value and a time corresponding to the index value. An indicator value is calculated from one or more measured values of one or more measured items. For example, the index value may be calculated from measurement values of a plurality of currents within a predetermined period. For example, five current values may be used to calculate a single index value. Alternatively, the index value may be calculated based on one measured value of the current and one measured value of the engine temperature in the same period. Incidentally, if the absolute value of the difference between the measurement times is equal to or smaller than a predetermined value, the measurement times can be considered to belong to the same period.

The time corresponding to the index value may be the same as the measurement time corresponding to the measurement value used to calculate the index value. If the index value is calculated from a plurality of measurement values different in measurement time, the time corresponding to the index value may be calculated based on a statistical value of the plurality of measurement times. For example, an average value, a median value, or the like of a plurality of measurement times may be used as the time corresponding to the index value.

Fig. 2 is a diagram describing index data. The graph indicated by the circle represents the index. The ordinate represents the index value, and the abscissa represents the time corresponding to the index value. The broken lines in fig. 2 respectively represent the upper limit and the lower limit of the index value expected to be measured when the monitoring target is normal. That is, when the index value falls within the range between the upper limit and the lower limit, the monitoring target is regarded as normal. The range of the index value in which the monitoring target is regarded as normal is referred to as an allowable range. In fig. 2, the indices within the allowable range are represented by white circles, and the indices outside the allowable range are represented by black circles. Indicators outside of the allowable range are referred to as deviation scenarios.

The deviation indicates that the monitored target deviates from a normal state. That is, the monitoring target is likely to be abnormal. However, not all deviation cases indicate abnormality of the monitoring target.

As described above, the measurement data includes the measurement values acquired while the monitoring target is operating. When the monitoring target is running, the measurement data is more likely to contain interference than when the monitoring target is stopped, and the accuracy of the index tends to decrease. For example, if the monitoring target is a vehicle and the current flowing through a specific area is being measured, when the vehicle is driven, it is likely that the measurement data contains disturbance and the measurement value contains an abnormal value.

Therefore, when any disturbance enters the measurement data, the calculated index may show an abnormal value, resulting in a deviation situation. Since deviations may be caused by disturbances in this way, it is necessary to determine whether a given deviation situation is caused by a disturbance or by an abnormality of the monitored target.

Therefore, the data processing device removes any index relating to the disturbance from the calculated index data, thereby refining the index data. Then, the use of the refined index data makes it possible to grasp the state of the monitoring target with high accuracy.

The internal configuration of the data processor will be described. The measurement data acquirer 1 acquires measurement data. The measurement data acquirer 1 may acquire measurement data directly from a sensor or the like, or may acquire measurement data indirectly via an external device. The measurement data obtainer 1 may manipulate the measurement data to calculate measurement data to be processed by the data processor 3. For example, after removing unnecessary measurement items, the measurement data acquirer 1 may calculate a single measurement data by combining a plurality of measurement data.

The memory 2 stores data for various processes of the data processor 3. The data is stored in the memory 2 in advance. Further, data input in the data processor 3, data calculated in various processes of the data processor 3, and other data may be stored, and there is no particular limitation on the data to be stored. Incidentally, the memory 2 may be divided according to the stored data.

The data processor 3 processes the measurement data and calculates indicator data. Details will be described together with the internal configuration.

The output device 4 outputs data related to the data processor 3. For example, the output device 4 outputs an exclusion condition described later, refined index data, and a determination result based on the refined index data. Further, the output device 4 may output data used in processing of various components and processing results produced by the various components.

Incidentally, the data output by the output device 4 is not particularly limited, and the data stored in the memory 2 may be output. Further, the output scheme of the output device 4 is not particularly limited. An image, voice, or the like may be output to a display or the like, and an electronic file containing the processing result may be saved in an external memory.

The internal configuration of the data processor 3 will be described. The index calculator 31 calculates index data based on the measurement data. The index calculator 31 may calculate the index value based on one or more measurement values using a predetermined calculation formula. Alternatively, the measured value itself may be used as the index value. As the calculation formula, a well-known calculation formula can be used.

The exclusion condition calculator 32 calculates a condition for removing the index data related to the interference from the index data. This condition is called an exclusion condition. The index in the index data is divided into an index marked as excluded and an index marked as to be checked for deviation based on the exclusion condition.

The exclusion condition is used to determine whether to measure the first measurement data or the second measurement data in a situation where interference is likely to enter. The first measurement data is used to calculate a given index. The second measurement data is measured during the same time period as the first measurement data. As a situation where interference is liable to enter, a predetermined operation state may be defined in advance according to the monitoring target.

For example, when the index value is calculated based on the measured current value, if the exclusion condition of the speed measured in the same period of time as the measured current value is used, the index of the monitoring target during high-speed travel in which interference is easy to enter can be removed. The measurement items for calculating the index and the measurement items for calculating the exclusion condition may be established in advance.

The predetermined operating state may simply be a state in which the monitoring target is operating, or an operating state in which the output value or the like is equal to or higher than a predetermined value. For example, when the monitoring target is a generator or the like, the predetermined operating state may be a state in which electric power equal to or higher than a predetermined value is generated. Alternatively, the predetermined operation state may be a state in which the power consumption of the monitoring target is equal to or higher than a predetermined value. Alternatively, when the monitoring target is a vehicle, the predetermined running state may be a state in which the vehicle runs at a speed equal to or higher than a predetermined value. Alternatively, when the internal temperature of the monitoring target is equal to or higher than a predetermined value, the monitoring target may be considered to be operating under a high load. In this way, it becomes clear whether the first measurement data or the second measurement data has been measured under the predetermined operation state, and it becomes possible to determine that the index calculated from the first measurement data is likely to have been affected by the disturbance.

Fig. 3 is a diagram showing an example of the exclusion condition. The exclusion condition in fig. 3 uses a decision tree. For example, assume that the measurement data contains measurements on three sensors A, B and C. The index is created based on at least any one of the three sensors A, B and C, and the exclusion condition is also created based on at least any one of the three sensors A, B and C.

With the exclusion condition in fig. 3, first, index values are classified based on the measurement values on the sensor a. If the measured value on sensor A is 3 or more, the index is determined to be a deviation. If the measured value on sensor A is less than 3, the index values are further classified based on the measured value on sensor B. If the measured value on the sensor B is less than 5.2, the index is determined to be deviated, and if the measured value on the sensor B is 5.2 or more, an index value is determined to be excluded.

Under the above exclusion condition, for example, when the index is calculated using all the measurement values on the sensors A, B and C, the index calculated when the measurement value on the sensor a is 2 and the measurement value on the sensor B is 6 is marked as exclusion. In addition, in the case where the index is calculated based on only the measurement value on the sensor C, when the measurement value on the sensor a at 13 o 'clock is 2 and the measurement value on the sensor B is 6, the index calculated from the measurement value on the sensor C at 13 o' clock is also marked as excluded.

Incidentally, the measurement items of each sensor may be the same or different. For example, both sensor a and sensor B may measure current at the same site. Alternatively, sensor B may measure current at a different location than sensor a. Alternatively, sensor a may measure current, while sensor B may measure voltage.

The exclusion condition may be calculated using machine learning based on the determination result produced by the deviation determiner 34. With regard to machine learning, well-known techniques may be used. Machine learning techniques include, for example, methods using models for classifying metrics into groups containing metrics marked to check for deviations and groups containing metrics marked to exclude.

Further, there is a technique for calculating an exclusion condition by quantifying the degree of deviation of the index and performing sparse regression. It is conceivable to calculate the degree of deviation on the basis of the difference from the reference value or the nearest limit value. The exclusion condition may be calculated based on statistical values of indexes such as an average value and a median value.

The rejector 33 (refiner) removes the index data related to the disturbance from the index data based on the rejection condition, thereby refining the index data.

The deviation determiner 34 determines whether each index in the refined index data deviates from a predetermined allowable range. Incidentally, within a predetermined allowable range, only either one of the upper limit and the lower limit may be specified. That is, when the index is equal to or less than the upper limit, or equal to or greater than the lower limit, it may be determined that the index is allowable. Further, the allowable range may be set within a predetermined value above and below the reference value. For example, if the reference value of the current is 10A, the allowable range may be set to 0.5A above and below the reference value. In this case, it is permissible to determine the measured current value falling within the range of 9.5A to 10.5A.

The allowable range may be changed according to the state of the monitoring target. For example, it is conceivable that the allowable range of the index related to the measurement data changes between when the monitoring target moves and when the monitoring target stops. The state of the monitoring target may be determined based on the measurement data.

Fig. 4 is a diagram showing the processing results produced by the ejector 33 and the deviation determiner 34. Fig. 4 is also an example of an output from the output device 4. The triangle pattern and the square pattern indicate the index excluded by the excluder 33 among the indexes shown in fig. 2. In fig. 4, it is assumed that the eliminator 33 has performed refinement using an exclusion condition made up of a plurality of conditions such as that shown in fig. 3. The conditions constituting the exclusion condition are referred to as sub-conditions. The index indicated by the triangle is an index excluded under the first sub-condition, and the index indicated by the square is an index excluded under the second sub-condition. As shown in fig. 4, indices within the tolerance range may even be marked as excluded.

In fig. 4, the indicators marked to check for deviations are indicated by circles. Among the indexes indicated by circles, indexes outside the allowable range are indicated by black circles. In this way, the deviation determiner 34 determines whether each index marked to be checked for deviation deviates from the allowable range.

Incidentally, as shown in fig. 4, by changing the shape, color, or the like of the figure, the output device 4 can display whether each index is within the allowable range, out of the allowable range, or has been excluded. Further, the deviation rate may be output. The deviation rate is calculated by division using the number of indexes marked to be checked for deviation as a denominator and the number of indexes of deviation as a numerator.

Further, the output device 4 may output the degree of deviation marked as an index to be checked for deviation within a predetermined period such as one day or one week. Fig. 5 is a diagram showing an example of the output generated by the output device 4. FIG. 5 is a box-and-while diagram showing the distribution of index data on a day basis. The output device 4 can change the display form in this manner.

Next, the flow of processing performed by the components of the data processing apparatus will be described. Fig. 6 is a diagram showing an example of a schematic flowchart of the entire process performed by the data processing apparatus according to the first embodiment.

The measurement data acquirer 1 acquires measurement data (S101). The index calculator 31 calculates index data based on the measurement data (S102). Incidentally, the index value may be calculated each time the measurement value is acquired. Alternatively, the index value may be calculated at a time when a predetermined number of measurement values are acquired.

The exclusion condition calculator 32 calculates an exclusion condition based on the history of the past deviation determination and the measurement data related to the past deviation determination (S103). Incidentally, when there is no history of past deviation determination, a predetermined condition (initial condition) is used as the exclusion condition.

Based on the exclusion conditions calculated by the exclusion condition calculator 32, the excluder 33 excludes any index marked as exclusion from the index data received from the index calculator 31 (S104). Each index in the index data refined because the index marked as excluded has been excluded by the exclusion condition calculator 32 is checked for deviation by the deviation determiner 34 (S105). The deviation determiner 34 updates the deviation determination history (S106). Therefore, the exclusion condition is updated in the next process. Then, the output device 4 displays the processing result and the like (S107), thereby completing the current flow.

Incidentally, this flowchart is merely an example, and the processing order and the like are not limited as long as a required processing result is available. For example, the process of S103 may be performed before the processes of S101 and S102. Further, the result of each process may be sequentially stored in the memory 2, and each component may acquire the process result by referring to the memory 2.

As described above, the present embodiment can remove unnecessary interference from measurement data. Therefore, the present embodiment makes it possible to grasp the current state of the monitoring target with high accuracy from the measurement data acquired during the operation of the monitoring target even containing many disturbances. As a result, the present embodiment makes it possible to reduce the frequency of inspection of the monitoring target, omit unnecessary inspection, and suppress maintenance costs.

Incidentally, the data processing apparatus may be composed of a plurality of apparatuses capable of exchanging data via communication or electric signals. For example, the data processing apparatus may be divided into a first apparatus provided with an excluder 33 or the like and configured to create refined index data, and a second apparatus provided with a deviation determiner 34 and configured to determine a deviation.

(second embodiment)

Fig. 7 is a block diagram showing an example of a schematic configuration of a data processing apparatus according to the second embodiment. The second embodiment differs from the first embodiment in that there is also an input device 5, the input device 5 being configured to receive input from a user. Description of points similar to those of the first embodiment will be omitted.

Once the output device 4 outputs the exclusion condition calculated by the exclusion condition calculator 32, users such as monitoring personnel and administrators of monitoring targets can determine whether the output exclusion condition is appropriate. There may be situations where it is desirable to mitigate or enforce the exclusion conditions.

Therefore, according to the present embodiment, the input device 5 receives an input related to the exclusion condition. For example, parameters required for creating the exclusion condition, a modification to the calculated exclusion condition, and the like are input. The exclusion condition creator creates or modifies an exclusion condition based on an input value received by the input device 5.

Fig. 8 is a diagram describing a modification to the exclusion condition. The GUI (graphical user interface) shown in fig. 8 is output by the output device 4 to receive the modification from the user.

On the right side of fig. 8, the exclusion conditions calculated using the decision tree are shown in a tree structure. The calculation conditions for the exclusion conditions are shown on the left side of fig. 8. The learning data items used for calculation, the applied technique, and the applied condition are displayed as calculation conditions. The applied technique is an already applied machine learning technique. The settings of the applied conditions vary with the selected applied technology. In the case of decision trees, because machine learning techniques are applied as a classification problem, it is necessary to select groups for use in classification. In the example of fig. 8, the indices are classified into a set of indices within the allowable range and a set of indices outside the allowable range. The indices may be classified into two groups according to the distance from the reference value.

The GUI in fig. 8 is adapted to display a change in content and serves as an input interface for modifying the exclusion condition. That is, when changes in the GUI are input into the input device 5, the exclusion condition calculator 32 recreates the exclusion conditions based on the changes.

For example, it is conceivable to adjust the threshold value of the exclusion condition. In the example of fig. 8, after changing the threshold in the tree structure shown on the right, when the user presses the "modify model" button displayed at the bottom of the screen, the exclusion condition with the modified threshold is re-created.

Incidentally, it is also preferable that the setting of the decision tree parameter be changed. Conceivable parameters include tree building algorithms and pruning scopes.

In the case where a model with a desired accuracy is not created even if the exclusion condition parameters are adjusted, or in the like case, it is conceivable to change or reconstruct the model. After determining the items required for the exclusion condition calculation on the left side of fig. 8, when the user presses a "rebuild model" button displayed at the bottom of the screen, a new exclusion condition is created (the exclusion condition is recreated).

Incidentally, if the exclusion condition has no problem, the user can press an "application model" button displayed at the bottom of the screen. After the exclusion condition calculator 32 calculates the exclusion condition, the output device 4 may be configured to output the GUI in fig. 8, and the processes of the excluder 33 and the deviation determiner 34 may be configured to be executed only when the "application model" button is pressed. In this way, the processing can be set to continue only after confirmation by the user to prevent an unsatisfactory processing result from being output.

Next, a flow of excluding the condition update will be described. Fig. 9 is a diagram showing an example of a flowchart of the exclusion condition update process. It is assumed that the present flow is executed between S103 and S104 in the overall process shown in fig. 6. Further, the present flow may be executed after the process of S107, and the process of S104 may be executed again after the present flow is ended.

The output device 4 outputs the exclusion condition using a GUI such as that shown in fig. 8 (S201). If the output exclusion condition is not approved (no in S202), the exclusion condition modified on the GUI is acquired by the input device 5 (S203). The modified exclusion condition is passed to the exclusion condition calculator 32, and the exclusion condition calculator 32 then updates the exclusion condition (S204). The updated exclusion condition is output again by the output device 4 (S201). In this way, the processing of S201 to S204 is repeated until the exclusion condition is approved. Then, when the exclusion condition is approved (yes in S202), the present flow is ended.

As described above, according to the present embodiment, the exclusion condition calculator 32 creates or modifies an exclusion condition based on an input to the input device 5. This makes it possible to create an exclusion condition reflecting the idea and experience of the user and to ensure the effectiveness of the interference cancellation method.

(third embodiment)

Fig. 10 is a block diagram showing an example of a schematic configuration of a data processing apparatus according to the third embodiment. The third embodiment differs from the above-described embodiments in that the data processor 3 further comprises a counter 35. Although in fig. 10, the counter 35 is added to the second embodiment, the counter 35 may be added to other embodiments. Descriptions of points similar to those of the above-described embodiment will be omitted.

The counter 35 counts the number of indexes included in the refined index data. The number of indices contained in the refined index data will be referred to as a sample count. That is, it can be said that the counter 35 counts the number of valid samples in the index data. The index data is considered reliable if the sample count in the index data is equal to or greater than a predetermined threshold. For example, in the case of the processing result shown in fig. 4, the sample count is 11, which is equal to the number of indexes marked to check for a deviation.

Assume that the threshold for the sample has been established in advance. Further, by classifying the sample count according to the state of the monitoring target, the sample count can be counted for a predetermined period.

Fig. 11 is a diagram showing an example of a result of sample counting with respect to counting. It is conceivable that the output device 4 outputs a table such as in fig. 11. In fig. 11, the sample count is counted by classifying the sample count into three operation modes and five periods. The operation mode indicates the type of the state of the monitoring target. The period can be set as desired. The length of the period need not be fixed.

For example, when the monitored target is a vehicle, in one conceivable case, the vehicle has three operating modes: acceleration, deceleration, and temporary stop. Then, if the period from 0 o 'clock to 5 o' clock is divided into intervals of one hour, and the sample count is counted by classifying the sample count according to the operation mode and period, a counting result such as in fig. 11 is produced.

The numbers in parentheses in fig. 11 indicate the threshold values of the sample counts. The threshold may vary with duration and mode of operation. In the example of fig. 11, it is assumed that the periods 4 and 5 are longer than the periods 1 to 3. Therefore, the threshold values of the periods 4 and 5 are set to be larger than the threshold values of the periods 1 to 3. In this way, a threshold value of the sample count may also be output.

In the example of fig. 11, the sample count in run mode 3 in period 4 is less than its threshold. Therefore, it is conceivable that the index data in the operation mode 3 in the period 4 has low reliability.

Fig. 12 is a diagram showing another example of the result of sample counting with respect to counting. Although the counting results are displayed in a table form in fig. 11, the counting results may be displayed in a bar graph as shown in fig. 12. It is possible to allow the display format of the sample count result to be specified via the input device 5 shown in the second embodiment.

After the excluder 33 excludes the index marked as exclusion based on the exclusion condition, the counting process performed by the counter 35 may be performed. In other words, in the flow shown in fig. 6, the counting process may be performed after the process of S104. The flowchart according to the present embodiment is omitted.

As described above, according to the present embodiment, the number of valid samples in the index data is counted. This allows the sample count in the refined index data to be verified, thereby ensuring the reliability of the output index data.

(fourth embodiment)

Fig. 13 is a block diagram showing an example of a schematic configuration of a data processing apparatus according to the fourth embodiment. The fourth embodiment differs from the above-described embodiments in that the data processor 3 further includes a state determination condition calculator 36 and a state determiner 37. Although these components are added to the third embodiment in fig. 13, these components may be added to other embodiments. Descriptions of points similar to those of the above-described embodiment will be omitted.

According to the present embodiment, the data processing apparatus itself performs the status determination and determines the status of the monitoring target. The state determination result regarding the monitoring target may be used in various applications. For example, in the case where the portion is determined to be normal based on the determination, the next examination may be omitted. Further, if it is determined in the determination that the monitoring target is abnormal, the output device 4 may output an alarm in the form of an image or sound. In this case, it can be said that the data processing device is both the state determination device and the abnormality detection device.

Since the measurement data is acquired during the operation of the monitoring target, if the state of the monitoring target is immediately determined by acquiring the measurement data in real time each time the measurement is performed, it is possible to detect an abnormality in real time.

The state determination condition calculator 36 calculates a state determination condition used in the state determination. The state determination condition is calculated using a learning model for calculating the state determination condition. This model will be referred to as a learning model for state determination condition calculation. When an evaluation (correct answer) of state determination is input by a user or the like, a learning model for state determination condition calculation is updated by learning the evaluation. This improves the accuracy of the state determination condition. As for the learning method, a known technique can be used in the same manner as when the exclusion condition calculator 32 calculates the exclusion condition. The evaluation of the above-mentioned state determination is obtained via the input device 5.

Fig. 14 is a diagram describing evaluation input regarding state determination. The GUI shown in fig. 14 is an interface for receiving an evaluation of the state determination output from the output device 4. The rate of deviation and sample count are shown sorted by duration and mode of operation. In the example of fig. 14, there are two buttons available: "inspection labor-saving available" and "inspection labor-saving unavailable". These buttons are used to specify whether checking of items for which the check box has been marked can be omitted.

The gray areas indicate that the deviation rate or sample count was not identified as normal. In the example of fig. 14, the deviation rate in the operation mode 1 in the period 1 is high, and the sample count in the operation mode 3 in the period 4 is low. Therefore, it may be determined that the two checks cannot be omitted.

Fig. 15 is a diagram showing an example of a history of correct answers to the state determination. Incidentally, items other than those shown in fig. 15 may be included in the correct answer. It is assumed that such a correct answer history is accumulated in the memory 2. Then, by using the correct answer history as learning data, machine learning is performed as a classification question for classifying the refined index data, and thus the state determination condition is calculated. Known techniques may be used for machine learning. Incidentally, the state determination condition may be calculated for each operation mode by learning with respect to each operation mode.

Therefore, the state determination condition calculator 36 repeats learning based on the state of the monitoring target determined by the state determiner and the data on the validity of the state, thereby updating the learning model for the state determination condition calculation. Since the state determination condition is calculated from the updated learning model for state determination condition calculation, the accuracy of the state determination condition is improved. Incidentally, if the user supplies the state determination condition via the input device 5, the state determination condition calculator 36 may be omitted.

Based on the state determination condition, the state determiner 37 determines the state of the monitoring target according to the determination result produced by the deviation determiner 34. Incidentally, the state of the monitoring target can be classified into two types: normal and abnormal. Alternatively, the states of the monitoring targets may be classified into three or more types according to the degree of difference from the reference value.

Based on the determined state of the monitoring target, the state determiner 37 may determine whether the monitoring target needs to be checked. For example, when it is determined that the state is normal, the state determiner 37 may determine that the predetermined check item may be omitted. In addition, for example, when the deviation rate is equal to or higher than 5% but lower than 10%, the state determiner 37 may determine the state in which close attention to the monitoring target is required, and may omit the check related to the first check item instead of the second check item.

The state determiner 37 may not perform the state determination if the sample count counted by the counter 35 is less than the threshold, or may add a warning that the determination result has low reliability. For example, as in the example of fig. 14, when the sample count does not satisfy the condition, the state determiner 37 may determine that the check cannot be omitted because the state cannot be accurately determined. Incidentally, the state determination may be made with respect to each operation mode. If there is an inspection item corresponding to each operation mode, in the example of fig. 11, the inspection items related to the operation modes 1 and 2 may be omitted while prohibiting the omission of the inspection item related to the operation mode 3.

The state determination result generated by the state determiner 37 is output from the output device 4. It is assumed that the output information includes the state of the monitoring target, whether or not the inspection can be omitted and the reason therefor, and the inspection item that can be omitted. For example, the output device 4 may display a message on a monitor connected to the data processing apparatus indicating that the check cannot be omitted because the deviation rate is above the threshold and the sample count in run mode 3 in period 4 is insufficient.

Next, the flow of state determination will be described. Fig. 16 is a diagram showing an example of a flowchart of state determination. It is assumed that this flow is executed after S105 in the overall process shown in fig. 6.

Based on the state determination condition, the state determiner 37 determines the state from the index data (S301). The determination result may be output together with or separately from the processing result of the above-described embodiment. Then, the output device 4 outputs a GUI for correct answer input (S302). When the user operates the GUI for correct answer input, the input device 5 receives the correct answer of the state determination (S303). Then, the state determination condition calculator 36 updates the state determination condition based on the index data and the correct answer history (S304), thereby completing the present flow. Therefore, the updated state determination condition is used in the next process, thereby improving the accuracy of the state determination.

As described above, the data processing apparatus according to the present embodiment also determines the state of the monitoring target based on the index data. This makes it possible to automatically determine the state of the monitoring target, omit inspection, and the like. Further, the state determination condition may also be automatically created by learning.

Each process in the above-described embodiments may be realized by software (program). Therefore, the above-described embodiments can be implemented using, for example, a general-purpose computer apparatus as basic hardware and causing a processor installed in the computer apparatus to execute the program.

Fig. 17 is a block diagram showing an example of a hardware configuration according to an embodiment of the present invention. The data processing apparatus comprises a processor 61, a main memory 62, a secondary memory 63, a network interface 64 and a device interface 65, which are interconnected via a bus 66 and implemented as the computer apparatus 6. Furthermore, the data processing device may also include an input device and an output device.

The data processing apparatus according to the present embodiment can be realized by installing a program executed by each apparatus in advance on the computer apparatus 6, or by installing a program in a storage medium such as a CD-ROM, or distributed on the computer apparatus 6 through a network as needed.

Incidentally, in fig. 17, the computer apparatus includes each of all the components, but may include a plurality of units of each component. Further, although a single computer device is shown in fig. 17, the software may be installed on a plurality of computer devices. Each of the plurality of computer devices may perform processing of a different portion of the software to produce a processing result. That is, the data processing apparatus may be configured as a system.

The processor 61 is an electronic circuit including a control device and an arithmetic device of a computer. The processor 61 executes arithmetic processing based on data or programs input from various devices within the computer apparatus 6, and outputs calculation results and control signals to various apparatuses. Specifically, the processor 61 executes an OS (operating system) or an application program of the computer apparatus 6, and controls various devices of the computer apparatus 6.

The processor 61 is not particularly limited as long as the above-described processing can be performed. The processor 61 may be, for example, a general purpose processor, a Central Processing Unit (CPU), a microprocessor, a Digital Signal Processor (DSP), a controller, a microcontroller, or a state machine. Further, the processor 61 may be incorporated into an application specific integrated circuit, a Field Programmable Gate Array (FPGA), or a Programmable Logic Device (PLD). Further, the processor 61 may be composed of a plurality of processing devices. For example, the processor 61 may be a combination of a DSP and a microprocessor, or one or more microprocessors configured to cooperate with a DSP core.

The main memory 62 is a storage device configured to store commands executed by the processor 61 and various data, and information stored in the main memory 62 is directly read by the processor 61. The secondary memory 63 is a storage device other than the primary memory 62. Incidentally, the term "storage device" denotes any electronic component capable of storing electronic information. Volatile memories such as RAM, DRAM, and SRAM for temporarily holding information are mainly used as the main memory 62, but in the embodiment of the present invention, the main memory 62 is not limited to these volatile memories. The storage devices used as the primary memory 62 and the secondary memory 63 may be volatile or non-volatile memories. Non-volatile memory includes programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable PROM (eeprom), non-volatile random access memory (NVRAM), flash memory, and MRAM. In addition, a magnetic or optical data memory may be used as the auxiliary memory 63. Available data storage includes magnetic disks such as hard disks, optical disks such as DVDs, flash memory such as USB memory, and magnetic tapes.

Incidentally, if the processor 61 reads and/or writes information directly or indirectly from/to the main memory 62 or the auxiliary memory 63, it can be said that the storage device is in electrical communication with the processor. Incidentally, the main memory 62 may be integrated into the processor. Also, the main memory 62 may be said to be in electrical communication with the processor.

The network interface 64 is an interface for connecting to a communication network by radio or wire. Any network interface 64 compatible with existing communication standards may be used. The output result or the like can be transmitted to the external device 8 connected online via the communication network 7 through the network interface 64.

The device interface 65 is a USB or other interface for connecting to the external apparatus 8 configured to record the output result or the like. The external device 8 may be an external storage medium or a memory such as a database. The external storage medium may be any recording medium such as HDD, CD-R, CD-RW, DVD-RAM, DVD-R, or SAN (storage area network). Alternatively, the external device 8 may be an output device. For example, the external device 8 may be a display device configured to display an image, a device configured to output voice, or the like. Examples of the external device 8 include an LCD (liquid crystal display), a CRT (cathode ray tube), a PDP (plasma display panel), and a speaker, but the external device 8 is not limited to these.

Further, all or part of the computer device 6, that is, all or part of the data processing device may be constituted by a dedicated electronic circuit (i.e., hardware), such as a semiconductor integrated circuit equipped with the processor 61 or the like. The dedicated hardware may be constructed in combination with a storage device such as a RAM or ROM.

Incidentally, although a single computer device is shown in fig. 17, the software may be installed on a plurality of computer devices. Each of the plurality of computer devices may perform processing of a different portion of the software to produce a processing result.

While certain embodiments have been described, these embodiments have been presented by way of example only, and are not intended to limit the scope of the invention. Indeed, the novel embodiments described herein may be embodied in a variety of other forms; furthermore, various omissions, substitutions and changes in the form of the embodiments described herein may be made without departing from the spirit of the inventions. The accompanying claims and their equivalents are intended to cover such forms or modifications as would fall within the scope and spirit of the inventions.

Claims

1. A data processing apparatus comprising:

an index calculator configured to calculate index data containing an index representing a state of a monitoring target based on measurement data on the monitoring target;

an exclusion condition calculator configured to calculate an exclusion condition for removing any index relating to interference from the index data based on the measurement data; and

a refiner configured to refine the index data by removing an index related to interference from the index data based on the exclusion condition, wherein:

the index represents a performance state of the monitoring target; and

the exclusion condition is used to determine whether the indicator is related to the disturbance by determining whether second measurement data, which is measured in the same period of time as the first measurement data used to calculate the indicator, represents a predetermined operating state of the monitoring target.

2. The data processing apparatus of claim 1, further comprising: a deviation determiner configured to determine whether an index included in index data refined by the refiner deviates from an allowable range.

3. The data processing apparatus of claim 2, further comprising: a state determiner configured to determine a state of the monitoring target according to a determination result generated by the deviation determiner based on a state determination condition.

4. The data processing apparatus according to claim 3, wherein the state determiner determines whether the monitoring target needs to be checked based on the determined state of the monitoring target.

5. The data processing apparatus according to claim 3 or 4, further comprising: a state determination condition calculator configured to:

updating a learning model for calculating the state determination condition based on the state of the monitoring target determined by the state determiner and data on validity of the determined state of the monitoring target, and

the state determination condition is calculated based on a learning model for calculating the state determination condition.

6. The data processing apparatus according to any one of claims 1 to 4, further comprising:

an output device configured to output at least the exclusion condition; and

an input device configured to receive input of data regarding the exclusion condition,

wherein the exclusion condition calculator recreates the exclusion condition based on the data regarding the exclusion condition received by the input device.

7. The data processing apparatus according to any one of claims 1 to 4, further comprising: a counter configured to count the number of indices contained in the index data refined by the refiner.

8. A method of data processing, comprising:

calculating index data including an index representing a state of a monitoring target based on measurement data about the monitoring target;

calculating an exclusion condition for removing any indicators related to interference from the indicator data based on the measurement data; and

removing an indicator related to interference from the indicator data based on the exclusion condition, thereby refining the indicator data, wherein:

the index represents a performance state of the monitoring target; and

9. A program configured to cause a computer to:

the index represents a performance state of the monitoring target; and