JP2019021056A - Data processing apparatus, data processing method, and program - Google Patents

Abstract

To determine a disturbance included in a measured data related to a monitoring object.SOLUTION: A data processing apparatus according to one embodiment of the present invention has an index calculating unit, a removal condition calculating unit, and a refining unit. The index calculating unit calculates, based on measured data related to a monitoring object, index data including an index indicating a state of the monitoring object. The removal condition calculating unit, based on the measured data, calculates a removal condition for use in removing an index related to a disturbance from the index data. The refining unit removes the index related to the disturbance from the index data on the basis of the removal condition to thereby refine the index data.SELECTED DRAWING: Figure 1

Description

  Embodiments described herein relate generally to a data processing device, a data processing method, and a program.

  In order to keep the system operating safely and stably, it is essential to check the soundness of the system and to maintain it as necessary. However, the maintenance cost increases as the inspection frequency increases. Therefore, the next inspection plan is created based on the elapsed time from the previous maintenance, the operation time of the system, and the like. Such a maintenance method for creating an inspection plan based on time is called TBM (Time Based Maintenance).

  In recent years, it is becoming mainstream to grasp the state of the system in real time by remote monitoring due to the development of wireless technology and the price reduction of sensors. If the current state of the system can be grasped, the inspection period can be extended, and unnecessary inspection items can be omitted. Such a maintenance method for creating an inspection plan based on the current state of the system is called CBM (Condition Based Maintenance). The transition from TBM to CBM is expected to reduce maintenance costs.

  However, various disturbances are likely to occur in the data measured during the operation of the system in order to grasp the current state of the system. Therefore, in order to grasp the current state of the system with high accuracy, it is necessary to remove unnecessary disturbance from the measurement data.

Japanese Patent No. 5306902

  One embodiment of the present invention determines a disturbance included in measurement data for a monitoring target.

  A data processing apparatus as one aspect of the present invention includes an index calculation unit, an exclusion condition calculation unit, and a purification unit. The index calculation unit calculates index data including an index indicating a state of the monitoring target based on measurement data related to the monitoring target. The exclusion condition calculation unit calculates an exclusion condition for removing an index related to disturbance from the index data based on the measurement data. The refinement unit refines the index data by removing the index related to the disturbance from the index data based on the exclusion condition.

1 is a block diagram showing an example of a data processing device according to a first embodiment. The figure explaining index data. The figure which shows an example of exclusion conditions. The figure which shows the processing result of an exclusion part and a deviation determination part. The figure which shows an example of the output by an output part. The figure which shows an example of the schematic flowchart of the whole process of the data processor which concerns on 1st Embodiment. The block diagram which shows an example of schematic structure of the data processor which concerns on 2nd Embodiment. The figure explaining correction of exclusion conditions. The figure which shows an example of the flowchart of the update process of exclusion conditions. The block diagram which shows an example of schematic structure of the data processor which concerns on 3rd Embodiment. The figure which shows an example of the result of the total number of samples. The figure which shows another example of the result of the total number of samples. The block diagram which shows an example of schematic structure of the data processor which concerns on 4th Embodiment. The figure explaining the input of evaluation with respect to state determination. The figure which shows an example of the log | history of the correct answer of a state determination. The figure which shows an example of the flowchart of a state determination. The block diagram which shows an example of the hardware constitutions in one Embodiment of this invention.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(First embodiment)
FIG. 1 is a block diagram illustrating an example of a data processing apparatus according to the first embodiment. The data processing apparatus according to the first embodiment includes a measurement data acquisition unit 1, a storage unit 2, a data processing unit 3, and an output unit 4. The data processing unit 3 includes an index calculation unit 31, an exclusion condition calculation unit 32, an exclusion unit (purification unit) 33, and a departure determination unit 34.

  The data processing apparatus according to the present embodiment calculates index data including an index indicating the state of the monitoring target based on the measurement data related to the monitoring target. Based on the index, detection of abnormality of the monitoring target, determination of necessity of inspection for the monitoring target, and the like are performed.

  The calculated index is considered to indicate a state in which measurement with a sensor or the like is difficult. For example, it may be an index (performance index) indicating the performance status of the monitoring target.

  The monitoring target is not particularly limited, and may be a device or a system composed of a plurality of devices. Moreover, living bodies, such as a person and an animal, may be sufficient.

  Measurement data means time-series data including measurement results of sensors and the like. The measurement result includes the measurement time and the measurement value. A known sensor may be used.

  The measurement item of the measurement data may be an item that can be measured by a known sensor or the like. The measurement location may be the entire monitoring target or a specific location of the monitoring target. For example, the measurement item may be temperature, humidity, flow rate, current, voltage, pressure, position, and the like. In addition, when the monitoring target is a moving body such as a vehicle, the speed, acceleration, and the like of the moving body are assumed to be measurement items.

  It is assumed that the measurement data can be arbitrarily measured. Therefore, it is assumed that the measurement data can include data measured when the monitoring target is operating and data measured when the monitoring target is stopped.

  In addition, you may include the measurement result by monitoring object itself in measurement data. Moreover, the setting input to the monitoring target may be included in the measurement data. For example, data indicating ON and OFF time zones of the power supply to be monitored may be included. Further, for example, when the monitoring target has a power saving mode for suppressing power consumption, data indicating a time zone in which the monitoring target is in the power saving mode may be included in the measurement data. For example, when the monitoring target is an air conditioner, data indicating a state such as during cooling, during heating, or during humidification may be included in the measurement data. Further, when the monitoring target is a vehicle, the measurement data may include data indicating a state such as running, accelerating, decelerating, or temporarily stopping.

  The state of the monitoring target determined by the monitoring target itself or another external device based on the measurement data may also be included in the measurement data. For example, when the current flowing through the motor built in the monitoring target is measured and the monitoring target determines that the motor is abnormal based on the measured current value, the monitoring target measures the value indicating the motor or monitoring target abnormality It may be added to the data.

  The index included in the index data is represented by a combination including at least the index value and the time corresponding to the index value. The index value is calculated from one or more measurement values related to one or more measurement items. For example, the index value may be calculated from measured values of a plurality of currents during a predetermined period. For example, one index value may be calculated using five current values. Alternatively, the index value may be calculated based on one measured value of current and one measured value of engine temperature in the same time zone. If the absolute value of the difference in measurement time is within a predetermined value, it may be regarded as the same time zone.

  The time corresponding to the index value may be the same as the measurement time corresponding to the measurement value used for calculating the index value. When index values are calculated from a plurality of measurement values at different measurement times, the index values may be calculated based on statistical values at a plurality of measurement times. For example, the time corresponding to the index value may be an average value or median value of a plurality of measurement times.

  FIG. 2 is a diagram for explaining the index data. A plot indicated by a circle indicates an index. The vertical axis indicates the index value, and the horizontal axis indicates the time corresponding to the index value. Each dotted line in FIG. 2 indicates an upper limit value and a lower limit value of measurement values that are expected to be measured when the monitoring target is normal. That is, when the index value exists between the upper limit value and the lower limit value, the monitoring target is expected to be normal. The index value range in which the monitoring target is expected to be normal is described as the allowable range. In FIG. 2, indices within the allowable range are indicated by white circles, and indices outside the allowable range are indicated by black circles. Indices outside the allowable range are described as deviation cases.

  The deviation case indicates that the monitoring target is far from the normal state. That is, there is a high possibility that the monitoring target is abnormal. However, not all deviation cases indicate abnormalities to be monitored.

  As described above, the measurement data includes measurement values when the monitoring target is operating. When the monitoring target is operating, there is a higher possibility that the measurement data includes disturbance and the accuracy of the index is likely to be lower than when the monitoring target is stopped. For example, when the monitoring target is a vehicle and the current flowing through a specific part is being measured, when the vehicle is in operation, the measurement data may include disturbances and the measurement values may include abnormal values Becomes higher.

  Therefore, when a disturbance enters the measurement data, the calculated index may show an abnormal value, which may be a deviation case. As described above, since the deviation case includes a disturbance, it is necessary to determine whether the deviation case is caused by the disturbance or the monitoring target abnormality.

  Therefore, the data processing device refines the index data by removing the index related to the disturbance from the calculated index data. Then, by using the refined index data, it is possible to grasp the state of the monitoring target with high accuracy.

  The internal configuration of the data processing apparatus will be described. The measurement data acquisition unit 1 acquires measurement data. The measurement data acquisition unit 1 may acquire directly from a sensor or the like, or may acquire it indirectly via an external device. Moreover, the measurement data acquisition part 1 may calculate the measurement data which the data process part 3 processes by processing measurement data. For example, one measurement data may be calculated by combining a plurality of measurement data after excluding unnecessary measurement items.

  The storage unit 2 stores data used for each process of the data processing unit 3. It is assumed that the data is stored in the storage unit 2 in advance. Further, data input to the data processing unit 3, data calculated in each process of the data processing unit 3, and the like may be stored, and the stored data is not particularly limited. In addition, the memory | storage part 2 may be divided for every data memorize | stored.

  The data processing unit 3 processes the measurement data and calculates index data. Details will be described together with the internal configuration.

  The output unit 4 outputs data related to the data processing unit 3. For example, a determination result based on an exclusion condition described later, refined index data, and refined index data is output. Moreover, the data used for the process of each part and the process result of each part may be output.

  Note that the data output by the output unit 4 is not particularly limited, and the data stored in the storage unit 2 may be output. The output method of the output unit 4 is not particularly limited. Images, sounds, and the like may be output to a display or the like, and the processing result may be saved in an external storage as an electronic file.

  The internal configuration of the data processing unit 3 will be described. The index calculation unit 31 calculates index data based on the measurement data. The index calculation unit 31 may calculate an 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 known formula may be used.

  The exclusion condition calculation unit 32 calculates a condition for removing the index data related to the disturbance from the index data. This condition is described as an exclusion condition. According to the exclusion condition, each index in the index data is divided into an index to be excluded and an index to be deviated.

  The exclusion condition is whether the first measurement data used for calculating the index or the second measurement data measured in the same time zone as the first measurement data is measured under a situation in which disturbance is likely to occur. Judgment. What is necessary is just to predetermine the predetermined operation state according to the monitoring object under the condition where disturbance is easy to enter.

  For example, when the index value is calculated based on the measured current value, if the exclusion condition for the speed measured in the same time zone as the measured current value is used, the index at the time of high-speed movement of the monitoring target subject to disturbance can be obtained. Can be removed. The measurement item used for calculating the index and the measurement item used for calculating the exclusion condition may be determined in advance.

  The predetermined operating state may be a state in which the monitoring target is simply operating, or an operating state in which an output value or the like is equal to or greater than a predetermined value. For example, when the monitoring target is a generator or the like, it may be in a state of generating electric power of a predetermined value or more. Alternatively, the monitored power consumption may be a predetermined value or more. Alternatively, when the monitoring target is a vehicle, the vehicle may be traveling 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 regarded as operating at a high load. In this way, it is determined whether the first measurement data or the second measurement data is measured in a predetermined operating state, and the index calculated from the first measurement data is highly likely to be an index related to disturbance. Can be determined.

  FIG. 3 is a diagram illustrating an example of the exclusion condition. The exclusion condition in FIG. 3 is a condition using a decision tree. For example, it is assumed that the measurement data includes measurement values from three sensors a, b, and c. The index is created based on at least one of these three sensors a, b, and c, but the exclusion condition is also created based on at least one of these three sensors a, b, and c.

  In the exclusion condition of FIG. 3, first, classification is performed based on the measured value by the sensor a. When the measured value of the sensor a is 3 or more, the index is determined as a departure determination target. When the measured value of the sensor a is less than 3, the classification is further performed based on the measured value of the sensor b. The index is determined as a departure determination target when the measured value of the sensor b is less than 5.2, and is determined as an exclusion target when the measured value of the sensor b is 5.2 or more.

  In the above exclusion condition, for example, when the index is calculated using all measured values of the sensors a, b and c, the calculated value is 2 when the measured value of the sensor a is 6 and the measured value of the sensor b is 6. Indicators are excluded. Further, for example, when the index is calculated based only on the measured value of sensor c, when the measured value of sensor a at 2 is 2 and the measured value of sensor b is 6, the measured value of sensor c at 13:00 The index calculated by is also excluded.

  Note that the measurement items of each sensor may be the same or different. For example, both sensor a and sensor b may measure the current at the same location. Alternatively, the sensor b may measure a current at a location different from the sensor a. Alternatively, the sensor a may measure current and the sensor b may measure voltage.

  The exclusion condition can be calculated by using machine learning based on the determination result of the deviation determination unit 34. Machine learning may use a known method. As a machine learning method, for example, there is a method of using a model that classifies into a group including a deviation determination target index and a group including an exclusion target index.

  There is also a method for calculating an exclusion condition by quantifying the degree of deviation of an index and performing sparse regression. The degree of deviation may be calculated based on a reference value or a difference from the nearest limit value. Moreover, it may be calculated based on statistical values such as an average value and a median value of the index.

  The exclusion unit 33 (purification unit) refines the index data by removing the index data related to the disturbance from the index data based on the exclusion condition.

  The deviation determination unit 34 determines whether each index of the refined index data deviates from a predetermined allowable range. Note that only one of the upper limit value and the lower limit value may be defined as the allowable range. That is, if the index is equal to or lower than the upper limit value or higher than the lower limit value, it may be determined that the index is acceptable. Further, it may be determined to be within a predetermined value before and after the reference value. For example, when the reference value of current is 10 A, 0.5 A before and after the reference value may be determined as the allowable range. In this case, if the measured value of the current is included in the range of 9.5A to 10.5A, it is determined to be acceptable.

  The allowable range may change 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 when the monitoring target is moving is different from the allowable range of the index related to the measurement data when the monitoring target is stopped. The state of the monitoring target can be determined based on the measurement data.

  FIG. 4 is a diagram illustrating processing results of the exclusion unit 33 and the departure determination unit 34. FIG. 4 is also an example of output from the output unit 4. Among the indexes shown in FIG. 2, the indexes excluded by the exclusion unit 33 are indicated by triangular and square plots. In FIG. 4, it is assumed that the exclusion unit 33 performs purification using an exclusion condition including a plurality of conditions as illustrated in FIG. 3. A condition that constitutes an exclusion condition is referred to as a sub condition. An index indicated by a triangle is an index excluded by the first sub condition, and an index indicated by a square is an index excluded by the second sub condition. As shown in FIG. 4, even an index within an allowable range can be excluded.

  In FIG. 4, the deviation determination target index is indicated by a circle. Among the indicators indicated by circles, indicators that are outside the allowable range are indicated by black circles. In this way, the departure determination unit 34 determines whether each indicator for departure determination is out of the allowable range.

  As shown in FIG. 4, the output unit 4 displays whether each index is within or outside the allowable range, or is an excluded index by changing the shape and color of the plot. Also good. Moreover, you may output a deviation rate. The departure rate is calculated by division using the number of departure determination target indices as the denominator and the number of departure indices as the numerator.

  Further, the output unit 4 may output the degree of deviation of the deviation determination target index in a predetermined period such as one day or one week. FIG. 5 is a diagram illustrating an example of output from the output unit 4. FIG. 5 is a box-and-whisker diagram showing the distribution of index data for each day. The output unit 4 may switch the display form in this way.

  Next, the flow of processing by each component will be described. FIG. 6 is a diagram illustrating an example of a schematic flowchart of overall processing of the data processing apparatus according to the first embodiment.

  The measurement data acquisition unit 1 acquires measurement data (S101). The index calculation unit 31 calculates index data based on the measurement data (S102). The index value may be calculated every time one measurement value is acquired, or may be calculated collectively after acquiring a predetermined number of measurement values.

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

  Based on the exclusion condition calculated by the exclusion condition calculation unit 32, the exclusion unit 33 excludes the index to be excluded from the index data from the index calculation unit 31 (S104). Further, the deviation determination unit 34 determines the deviation of each index of the index data refined by the exclusion condition calculation unit 32 excluding the exclusion target (S105). Further, the index deviation determination unit 34 updates the deviation determination history (S106). As a result, the exclusion condition is updated in the next processing. Then, the output unit 4 displays the processing result and the like (S106), and this flow ends.

  Note that this flowchart is an example, and the order of processing is not limited as long as a required processing result can be obtained. For example, the process of S103 may be performed before the processes of S101 and S102. In addition, the processing result of each process may be sequentially stored in the storage unit 2, and each component may acquire the processing result with reference to the storage unit 2.

  As described above, according to the present embodiment, unnecessary disturbance can be removed from the measurement data. Therefore, according to the present embodiment, it is possible to grasp the current state of the monitoring target with high accuracy even from measurement data acquired during operation of the monitoring target and including many disturbances. Therefore, as a result, it is possible to take measures such as increasing the inspection frequency of the monitoring target and omitting unnecessary inspections, thereby reducing maintenance costs.

  Note that the data processing device may be composed of a plurality of devices that can exchange data by communication or electrical signals. For example, it may be divided into a first device that includes the exclusion unit 33 and the like and creates refined index data, and a second device that includes the departure determination unit 34 and performs departure determination.

(Second Embodiment)
FIG. 7 is a block diagram illustrating an example of a schematic configuration of a data processing device according to the second embodiment. The second embodiment is different from the first embodiment in that it further includes an input unit 5 that receives an input from the user. The description of the same points as in the first embodiment will be omitted.

  When the output unit 4 outputs the exclusion condition calculated by the exclusion condition calculation unit 32, a user such as a monitor or a monitoring target administrator can determine whether or not the output exclusion condition is appropriate. In some cases, the exclusion condition may be relaxed or strengthened.

  Therefore, in the present embodiment, the input unit 5 receives an input related to the exclusion condition. For example, input of parameters necessary for creation of exclusion conditions, correction of calculated exclusion conditions, and the like are input. The exclusion condition creation unit creates and corrects the exclusion condition based on the input value received by the input unit 5.

  FIG. 8 is a diagram for explaining exclusion condition correction. The GUI (Graphical User Interface) shown in FIG. 8 is output by the output unit 4 in order to accept corrections from the user.

  On the right side of FIG. 8, the exclusion condition calculated by the decision tree is displayed in a tree structure. On the left side of FIG. 8, calculation conditions for exclusion conditions are displayed. As calculation conditions, a period of learning data used for calculation, an application method, and application conditions are displayed. The applied method means an applied machine learning method. The application conditions are set differently depending on the selected application method. In the case of a decision tree, a machine learning method is applied as a classification problem, so it is necessary to select a group to be classified. In the example of FIG. 8, the permissible range and the permissible range are classified. The distance from the reference value may be divided into two groups on the boundary.

  The GUI of FIG. 8 can change the display content and functions as an input interface for correcting the exclusion condition. That is, a change in GUI is input to the input unit 5, and the exclusion condition calculation unit 32 recreates the exclusion condition based on the change.

  For example, it is conceivable to adjust the threshold value of the exclusion condition. In the example of FIG. 8, after changing the threshold value in the tree structure shown on the right side, pressing the “Modify Model” button displayed on the lower side of the screen, the exclusion condition whose threshold value has been corrected is Recreated.

  It is desirable that the decision tree parameters can be set and changed. As parameters, a tree construction algorithm, a range for pruning, and the like are assumed.

  In addition, when a model having a desired accuracy is not created even after adjusting the parameters of the exclusion condition, it is conceivable to change or rebuild the model. After determining the necessary items for calculating the exclusion condition on the left side of FIG. 8, a new exclusion condition is recreated by pressing the “model rebuild” button displayed at the bottom of the screen.

  If there is no problem with the exclusion condition, the “model determination” button displayed at the bottom of the screen may be pressed. After the exclusion condition calculation unit 32 calculates the exclusion condition, the output unit 4 outputs the GUI shown in FIG. 8 and the processes of the exclusion unit 33 and the deviation determination unit 34 until the “model determination” button is pressed. May not be done. In this way, it may be possible to prevent an unsatisfactory processing result from being output as the processing proceeds after receiving confirmation from the user.

  Next, the flow of updating the exclusion condition will be described. FIG. 9 is a diagram illustrating an example of a flowchart of the exclusion condition update process. It is assumed that this flow is performed between S103 and S104 of the overall processing shown in FIG. Further, this flow may be performed after the processing of S107, and the processing of S104 may be performed again after the end of this flow.

  The output unit 4 outputs an exclusion condition using the GUI as shown in FIG. 8 (S201). When the output exclusion condition is not approved (NO in S202), the input unit 5 acquires the exclusion condition corrected by the GUI (S203). The modified exclusion condition is transferred to the exclusion condition calculation unit 32, and the exclusion condition calculation unit 32 updates the exclusion condition (S204). The updated exclusion condition is output again by the output unit 4 (S201). In this way, the processing from S201 to S204 is repeated until the exclusion condition is approved. If the exclusion condition is approved (YES in S202), this flow ends.

  As described above, according to the present embodiment, the exclusion condition creates or corrects the exclusion condition based on the input to the input unit 5. Thereby, it is possible to create an exclusion condition reflecting the user's thoughts and experiences, and to ensure the validity of the disturbance removal method.

(Third embodiment)
FIG. 10 is a block diagram illustrating an example of a schematic configuration of a data processing device according to the third embodiment. The third embodiment is different from the previous embodiments in that the data processing unit 3 further includes a counting unit 35. In FIG. 10, the counting unit 35 is added to the second embodiment, but the counting unit 35 may be added to other embodiments. Description of the same points as in the previous embodiments is omitted.

  The totaling unit 35 totals the number of indices included in the refined index data. The number of the index is described as the number of samples. That is, it can be said that the totaling unit 35 totals the number of effective samples of the index data. If the number of samples of index data is equal to or greater than a predetermined threshold, the index data is considered to be reliable. For example, in the case of the processing result shown in FIG. 4, the number of samples is 11 which is the number of indicators for deviation determination.

  It is assumed that the threshold value for the sample is predetermined. Further, the number of samples may be totaled for each predetermined period and for each state to be monitored.

  FIG. 11 is a diagram illustrating an example of the result of the total number of samples. It is conceivable that the output unit 4 outputs a table as shown in FIG. In FIG. 11, the number of samples is tabulated for each of three operation modes and five periods. The operation mode indicates the type of state to be monitored. The period may be arbitrarily determined. The length of each period may not be constant.

  For example, when the monitoring target is a vehicle, there may be three situations where the vehicle operation mode includes acceleration, deceleration, and temporary stop. Then, when the hours from 0:00 to 5:00 are divided every hour and the number of samples is totaled for each operation mode and period, a totaling result as shown in FIG. 11 is created.

  The numbers in parentheses in FIG. 11 indicate threshold values for the number of samples. The threshold value may be different for each period and operation mode. In the example of FIG. 11, the periods 4 and 5 are longer than the periods 1 to 3. Therefore, the number of threshold values in the periods 4 and 5 is set to be larger than the number of threshold values in the periods 1 to 3. In this way, a threshold value for the number of samples may also be output.

  In the example of FIG. 11, the number of samples in the operation mode 3 in the period 4 is smaller than the corresponding threshold value. Therefore, it is considered that the reliability of the index data related to the period 4 and the operation mode 3 is low.

  FIG. 12 is a diagram illustrating another example of the result of the total number of samples. In FIG. 11, the tabulation results are shown in a table format, but may be shown as a bar graph as shown in FIG. You may enable it to designate the display format of the result of a sample number via the input part 5 shown in 2nd Embodiment.

  The aggregation process by the aggregation unit 35 may be performed after the exclusion unit 33 excludes the index to be excluded based on the exclusion condition. In other words, it may be performed after the processing of S104 in the flow shown in FIG. The flowchart of this embodiment is omitted.

  As described above, according to this embodiment, the number of effective samples of index data is aggregated. Thereby, the number of samples of the refined index data can be confirmed, and the reliability of the output index data is guaranteed.

(Fourth embodiment)
FIG. 13 is a block diagram illustrating an example of a schematic configuration of a data processing device according to the fourth embodiment. The fourth embodiment is different from the previous embodiments in that it further includes a state determination condition calculation unit 36 and a state determination unit 37. In FIG. 13, these components are added to the third embodiment, but may be added to other embodiments. Description of the same points as in the previous embodiments is omitted.

  In this embodiment, the data processing apparatus itself makes a state determination to determine the state of the monitoring target. The determination result of the state of the monitoring target can be used for various purposes. For example, the next inspection may be omitted for a portion determined to be normal based on the determination. Moreover, when it determines with it being abnormal based on the said determination, the output part 4 may output an alert with an image, a sound, etc. In this case, it can be said that the data processing device is also a state determination device and an abnormality detection device.

  In addition, since the measurement data is acquired even while the monitoring target is in operation, an abnormality can be detected in real time if the measurement data is acquired in real time for each measurement and the state of the monitoring target is immediately determined.

  The state determination condition calculation unit 36 calculates a state determination condition that is a condition for performing state determination. The state determination condition is calculated using a learning model for calculating the state determination condition. The model is referred to as a state determination condition calculation learning model. The state determination condition calculation learning model is updated by learning the evaluation (correct answer) of the state determination input from the user or the like. Thereby, the accuracy of the state determination condition is improved. As the learning method, a known method may be used as in the case where the exclusion condition calculation unit 32 calculates the exclusion condition. The evaluation for the previous state determination is acquired via the input unit 5.

  FIG. 14 is a diagram for explaining an evaluation input for the state determination. The GUI shown in FIG. 14 is an interface for accepting an evaluation for the state determination output by the output unit 4. For each period and mode of operation, the deviation rate and number of samples are shown. In the example of FIG. 14, there are two types of buttons, “can save labor for inspection” and “cannot save labor for inspection”. .

  The grayed area indicates that the deviation rate or sample size was not recognized as normal. In the example of FIG. 14, the deviation rate is high in the operation mode 1 of the period 1, and the number of samples is small in the operation mode 3 of the period 4. For this reason, it is considered that these two places cannot be omitted.

  FIG. 15 is a diagram illustrating an example of a correct answer history of state determination. Note that items other than those shown in FIG. 15 may be included in the correct answer. It is assumed that such correct answer history is stored in the storage unit 2. The state determination condition is calculated by performing machine learning as a classification problem for classifying the refined index data using the correct answer history as learning data. Machine learning may use a known method. Note that learning may be performed for each operation mode, and the state determination condition may be calculated for each operation mode.

  In this way, the state determination condition updates the state determination condition calculation learning model by repeating learning based on the state of the monitoring target determined by the state determination unit and the data relating to the validity of the state. By calculating the state determination condition from the updated state determination condition calculation learning model, the accuracy of the state determination condition is increased. When the state determination condition is given from the user via the input unit 5, the state determination condition calculation unit 36 may be omitted.

  The state determination unit 37 determines the state of the monitoring target from the determination result of the departure determination unit 34 based on the state determination condition. Note that the status of the monitoring target may be classified into two types, normal and abnormal, or may be classified into three or more types based on the degree of deviation from the reference value.

  Moreover, the state determination part 37 may determine the necessity of the test | inspection with respect to the monitoring object based on the determined state of the monitoring object. For example, when the state is determined to be normal, the state determination unit 37 may determine that a predetermined inspection item may be omitted. Further, for example, when the deviation rate is 5% or more and less than 10%, the state determination unit 37 determines that the state to be monitored is caution, and the inspection for the first inspection item can be omitted, but the second inspection item It may be determined that the inspection for cannot be omitted.

  In addition, when the number of samples counted by the counting unit 35 is less than the threshold, the state determination unit 37 may add a warning that the state determination is not performed or the determination result is low in reliability. For example, as in the example of FIG. 11, when the number of samples does not satisfy the condition, it may be determined that the state cannot be accurately determined and the inspection cannot be omitted. The state determination may be performed for each operation mode. When there are inspection items corresponding to each operation mode, the inspection items according to operation modes 1 and 2 can be omitted in the example of FIG. 11, but the inspection items according to operation mode 3 may not be omitted.

  The state determination result of the state determination unit 37 is output by the output unit 4. It is assumed that the information to be output includes the status of the monitoring target, whether inspection can be omitted, the reason why inspection can be omitted, and inspection items that can be omitted. For example, the output unit 4 displays a message on the monitor connected to the data processing device that the inspection rate cannot be omitted because the deviation rate is higher than the threshold and the number of samples in the operation mode 3 in the period 4 is insufficient. May be.

  Next, the flow of state determination will be described. FIG. 16 is a diagram illustrating an example of a state determination flowchart. It is assumed that this flow is performed after S105 of the entire process shown in FIG.

  The state determination unit 37 determines the state from the index data based on the state determination condition (S301). The determination result may be output together with the processing result of the previous embodiments, or may be output individually. Then, the output unit 4 outputs a correct input GUI (S302). When the user operates the correct input GUI, the input unit 5 receives the correct state determination (S303). Then, the state determination condition calculation unit 36 updates the state determination condition based on the index data and the correct answer history (S304), and this flow ends. As a result, the updated state determination condition is used in the next process, and the state determination accuracy is improved.

  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. Thereby, it is possible to automate determinations such as the state of the monitoring target and the omission of inspection. Also, the state determination condition can be automatically created by learning.

  Each process in the embodiment described above may be realized by a dedicated circuit, or may be realized by using software (program). When software (program) is used, the embodiment described above uses, for example, a general-purpose computer device as basic hardware, and a processor such as a central processing unit (CPU) installed in the computer device. It can be realized by executing the program.

  FIG. 17 is a block diagram illustrating an example of a hardware configuration according to an embodiment of the present invention. The data processing device includes a processor 61, a main storage device 62, an auxiliary storage device 63, a device interface 64, and a network interface 65, and these can be realized as a computer device 6 connected via a bus 66. . Further, the data processing device may further include an input device 67 and an output device 68.

  The data processing apparatus according to the present embodiment may be realized by previously installing a program to be executed by each apparatus in the computer apparatus 6, or storing the program in a storage medium such as a CD-ROM or a network. It may be realized by distributing it via the computer and installing it appropriately in the computer device 6.

  In FIG. 17, the computer device includes one component, but may include a plurality of the same components. In FIG. 17, one computer apparatus is shown, but software may be installed in a plurality of computer apparatuses. A processing result may be generated by each of the plurality of computer devices executing a part of processing different in software. That is, the data processing apparatus may be configured as a system.

  The processor 61 is an electronic circuit including a computer control device and an arithmetic device. The processor 61 performs arithmetic processing based on data or a program input from each device having an internal configuration of the computer device 6, and outputs a calculation result and a control signal to each device. Specifically, the processor 61 executes an OS (Operating System), an application, and the like of the computer device 6 and controls each device configuring the computer device 6.

  The processor 61 is not particularly limited as long as the above 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, a state machine, or the like. Further, the processor 61 may be incorporated in an application specific integrated circuit, a field programmable gate array (FPGA), or a programmable logic circuit (PLD). Further, the processor 61 may be composed of a plurality of processing devices. For example, it may be a combination of a DSP and a microprocessor, or one or more microprocessors that cooperate with the DSP core.

  The main storage device 62 is a storage device that stores instructions executed by the processor 61 and various data. Information stored in the main storage device 62 is directly read out by the processor 61. The auxiliary storage device 63 is a storage device other than the main storage device 62. Note that the storage device means any electronic component capable of storing electronic information. As the main storage device 62, a volatile memory used to store temporary information such as RAM, DRAM, SRAM, etc. is mainly used. In the embodiment of the present invention, the main storage device 62 is used as the volatile memory. It is not limited. A storage device used as the main storage device 62 and the auxiliary storage device 63 may be a volatile memory or a non-volatile memory. 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, MRAM, etc. . Further, magnetic or optical data storage may be used as the auxiliary storage device 63. As data storage, a magnetic disk such as a hard disk, an optical disk such as a DVD, a flash memory such as a USB, a magnetic tape, or the like may be used.

  Note that if the processor 61 reads or writes information to or from the main storage device 62 or the auxiliary storage device 63 directly or indirectly, the storage device is in electrical communication with the processor. Can do. The main storage device 62 may be integrated with the processor. Again, it can be said that the main storage device 62 is in electrical communication with the processor.

  The network interface 64 is an interface for connecting to a communication network by wireless or wired. The network interface 64 may be one that conforms to existing communication standards. An output result or the like may be transmitted by the network interface 64 to the external device 8 that is communicatively connected via the communication network 7.

  The device interface 65 is an interface such as a USB connected to the external device 8 that records output results and the like. The external device 8 may be an external storage medium or a storage such as a database. The external storage medium may be an arbitrary recording medium such as an HDD, a CD-R, a CD-RW, a DVD-RAM, a DVD-R, or a SAN (Storage area network). Alternatively, the external device 8 may be an output device. For example, a display device for displaying an image or a device for outputting sound or the like may be used. Examples include, but are not limited to, a liquid crystal display (LCD), a cathode ray tube (CRT), a plasma display panel (PDP), and a speaker.

  Further, a part or all of the computer apparatus 6, that is, a part or all of the data processing apparatus may be configured by a dedicated electronic circuit (that is, hardware) such as a semiconductor integrated circuit on which the processor 61 is mounted. Good. The dedicated hardware may be configured in combination with a storage device such as a RAM or a ROM.

  In FIG. 17, one computer apparatus is shown, but software may be installed in a plurality of computer apparatuses. A processing result may be generated by each of the plurality of computer devices executing a part of processing different in software.

  Although one embodiment of the present invention has been described above, these embodiment are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the scope of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the invention described in the claims and the equivalents thereof.

1 Measurement Data Acquisition Unit 2 Storage Unit 3 Data Processing Unit 31 Index Calculation Unit 32 Exclusion Condition Calculation Unit 33 Exclusion Unit (Purification Unit)
34 Deviation determination unit 35 Totalization unit 36 State determination condition calculation unit 37 State determination unit 4 Output unit 5 Input unit 6 Computer device 61 Processor 62 Main storage device 63 Auxiliary storage device 64 Network interface 65 Device interface 66 Bus 7 Communication network 8 External device

Claims (10)

An index calculation unit that calculates index data including an index indicating a state of the monitoring target based on measurement data related to the monitoring target;
Based on the measurement data, an exclusion condition calculation unit that calculates an exclusion condition for removing an index related to disturbance from the index data;
Based on the exclusion condition, a purification unit that refines the index data by removing the index related to the disturbance from the index data;
A data processing apparatus comprising: The index indicates the performance status of the monitoring target;
The exclusion condition is determined by determining whether second measurement data measured in the same time zone as the first measurement data used for calculating the index indicates a predetermined operating state of the monitoring target. The data processing apparatus according to claim 1, wherein it is determined whether or not the index is related to disturbance. The data processing apparatus according to claim 1, further comprising a deviation determination unit that determines whether an index included in the index data refined by the purification unit deviates from an allowable range. The data processing device according to claim 3, further comprising: a state determination unit that determines a state of the monitoring target from a determination result of the deviation determination unit based on a state determination condition. The data processing device according to claim 4, wherein the state determination unit determines whether or not the inspection of the monitoring target is necessary based on the determined state of the monitoring target. The state determination condition calculation learning model is updated based on the state of the monitoring target determined by the state determination unit and data relating to the validity of the determined state of the monitoring target, and the state determination condition The data processing device according to claim 4, further comprising: a state determination condition calculation unit that calculates the state determination condition based on a calculation learning model. An output unit that outputs at least the exclusion condition;
An input unit for receiving input of data relating to the exclusion condition;
Further comprising
The data processing apparatus according to claim 1, wherein the exclusion condition calculation unit recreates the exclusion condition based on data related to the exclusion condition received by the input unit. The data processing device according to any one of claims 1 to 7, further comprising: a totaling unit that counts the number of the indexes included in the index data purified by the purification unit. Calculating index data including an index indicating a state of the monitoring target based on measurement data related to the monitoring target;
Calculating an exclusion condition for removing an index related to disturbance from the index data based on the measurement data;
Refining the index data by removing the index related to the disturbance from the index data based on the exclusion condition;
A data processing method comprising: Calculating index data including an index indicating a state of the monitoring target based on measurement data related to the monitoring target;
Calculating an exclusion condition for removing an index related to disturbance from the index data based on the measurement data;
Refining the index data by removing the index related to the disturbance from the index data based on the exclusion condition;
A program that causes a computer to execute.