JP7020876B2 - Decision device, correction device, decision system, decision method and computer program - Google Patents

Description

Embodiments of the present invention relate to a determination device, a correction device , a determination system, a determination method, and a computer program.

Various facilities such as industrial and social infrastructure suffer various economic losses such as functional deterioration due to aging deterioration and sudden failure, energy consumption loss, efficiency decrease, and cost increase due to deterioration of customer service quality. In order to avoid such problems, a diagnostic system is useful for maintaining the function and performance of the equipment at the same time as the maintenance and inspection of the equipment. The target of diagnosis is not only the hardware of the equipment, but also the control loop related to the equipment and sensors that monitor, detect, and measure the equipment status. In the conventional abnormality diagnosis, the abnormality diagnosis is made by evaluating the steady deviation between the data that is the reference of the abnormality diagnosis and the data to be diagnosed with respect to the indicated value of the sensor calibrator.
However, since inspections may be performed on different days for each sensor, in a group consisting of multiple sensors, in order to set the period of data that is the basis for abnormality diagnosis, each sensor belonging to the group must be inspected. It is necessary to confirm the inspection date before setting. Therefore, it may take time and effort to do it manually.

Japanese Unexamined Patent Publication No. 2016-16658 Japanese Unexamined Patent Publication No. 2016-177676 Japanese Patent No. 5025776 Japanese Patent No. 5337909

An object to be solved by the present invention is to provide a determination device, a correction device , a determination system, a determination method, and a computer program that can reduce the time and effort of the user.

The determination device of the embodiment has a collection unit and a determination unit. The collection unit collects data from multiple devices installed in the equipment. Based on the inspection date of each device, the determination unit determines the extraction start position of the reference data used as the reference for the abnormality diagnosis process of the unit composed of the plurality of devices from the collected data. The determination unit further determines a period for evaluating the reliability level of the reference data for each device.

The figure which shows the system configuration of the diagnostic system of embodiment. The schematic block diagram which shows the functional structure of the diagnostic apparatus in 1st Embodiment. The figure which shows the specific example of the inspection data table. The figure which shows the specific example of the diagnostic unit definition table. The figure which shows the specific example of the diagnosis group definition table. The figure which shows the specific example of the diagnostic table. The figure which shows the specific example of time series data. The figure which shows the specific example of the time series data for diagnosis. A flowchart showing the flow of abnormality diagnosis processing performed by the diagnostic device. A flowchart showing the flow of evaluation processing of the data reliability level performed by the diagnostic device. The figure which shows the specific example of the determination method of each period and the evaluation process of a data reliability level. The figure which shows the specific example of the determination method of each period and the evaluation process of a data reliability level. The figure which shows an example of the display screen displayed on the display part. The figure for demonstrating the flow of processing in 2nd Embodiment. The figure for demonstrating the flow of processing in 3rd Embodiment. The figure which shows the example of the sensor error estimation estimated using the error growth model. The figure which shows another example of the display screen displayed on the display part.

Hereinafter, the determination device, the correction device , the determination system, the determination method, and the computer program of the embodiment will be described with reference to the drawings.

FIG. 1 is a diagram showing a system configuration of the diagnostic system 100 of the embodiment. The diagnostic system 100 in the present embodiment is used as an abnormality diagnostic system for equipment provided in a building such as a building or a factory. The diagnostic system 100 includes a first sensor group 20, a second sensor group 30, a first control system system 40, a first information system system 50, a second control system system 60, a second information system system 70, and a data center 80. .. The first control system system 40, the first information system system 50, the second control system system 60, the second information system system 70, and the data center 80 are communicably connected to each other via the network 90. The network 90 is, for example, the Internet.

The first sensor group 20 is composed of a plurality of sensors. The sensor is, for example, a temperature sensor, a humidity sensor, a flow rate sensor, or the like. The sensor periodically outputs the measured value to the first control system system 40. The first sensor group 20 is installed in an individual building or factory.
The second sensor group 30 is composed of a plurality of sensors. The sensor periodically outputs the measured value to the second control system system 60. The second sensor group 30 is installed in an individual building or factory geographically separated from the place where the first sensor group 20 is installed.

The first control system system 40 is a system provided in the same place where the first sensor group 20 is installed. The first control system system 40 includes one or more control devices 41, a monitoring device 42, and a router 43.
The control device 41 is configured by using an information processing device such as a personal computer. The control device 41 acquires measured values from each sensor constituting the first sensor group 20. The control device 41 transmits the acquired measured value to the monitoring device 42.

The monitoring device 42 is configured by using an information processing device such as a personal computer. The monitoring device 42 monitors each control device 41 and collects the measured values transmitted from the control device 41. The monitoring device 42 transmits the collected measured values to the data center 80 via the router 43 and the network 90.

The monitoring device 42 passively or actively transmits the measured value to the data center 80. When the monitoring device 42 passively transmits the measured value, the monitoring device 42 transmits the measured value to the data center 80 in response to a request from the data center 80. Further, when the monitoring device 42 actively transmits the measured value, the monitoring device 42 periodically (for example, every 1 minute, 10 minutes, 30 minutes, 1 hour) transmits the measured value to the data center 80.
The router 43 relays communication between the first control system 40 and other systems and between the first control system 40 and the data center 80.

The first information system system 50 is a system that performs maintenance management of the first sensor group 20 and the first control system system 40. The first information system system 50 may be provided in the same place as the first control system system 40, or may be provided in a different place from the first control system system 40. The first information system system 50 includes a client device 51, an equipment maintenance device 52, and a router 53.

The client device 51 is configured by using an information processing device such as a smartphone, a mobile phone, a tablet terminal, a notebook computer, or a personal computer. The client device 51 is, for example, a terminal carried by an operator who inspects the first sensor group 20. The client device 51 transmits the inspection result of the first sensor group 20 to the equipment maintenance device 52. For example, the inspection result includes the date of inspection, the sensor ID, the sensor name, the measured value of the sensor actually measured by the operator, the measured value measured by the sensor, and the bias value.

The sensor ID is identification information for identifying the sensor. The sensor ID may be, for example, the MAC address of the sensor, or may be information to which identification information for identifying the location where the sensor is installed is added to the MAC address of the sensor. The sensor name represents the name of the sensor. The bias represents the amount of deviation between the measured value of the sensor and the measured value. When the measured value of the sensor and the measured value are input, the client device 51 calculates the bias by subtracting the measured value and the measured value of the input sensor.

Here, when the bias value is equal to or greater than the threshold value, the operator calibrates the sensor having the bias value equal to or greater than the threshold value. In this case, the operator re-inspects the sensor that has been calibrated. Then, the worker inputs the inspection result to the client device 51. At this time, the operator inputs information indicating that the inspection result is after calibration. The client device 51 transmits the inspection result after calibration to the equipment maintenance device 52.

The equipment maintenance device 52 is configured by using an information processing device such as a personal computer. The equipment maintenance device 52 manages the operating state of each device provided in the first control system system 40 and the operating state of the first sensor group 20. Further, the equipment maintenance device 52 acquires the inspection result transmitted from the client device 51. The equipment maintenance device 52 holds the acquired inspection result. When the equipment maintenance device 52 obtains the inspection result before calibration and the inspection result after calibration as the inspection result of the same sensor on the same inspection date, the inspection result before calibration and the inspection after calibration are obtained. Holds the result in association with it.

The equipment maintenance device 52 passively or actively transmits the inspection result to the data center 80. When the equipment maintenance device 52 passively transmits the inspection result, the equipment maintenance device 52 transmits the inspection result to the data center 80 in response to a request from the data center 80. Further, when the equipment maintenance device 52 actively transmits the inspection result, the equipment maintenance device 52 periodically (for example, every 1 minute, 10 minutes, 30 minutes, 1 hour) transmits the inspection result to the data center 80.
The router 53 relays communication between the first information system system 50 and other systems, and between the first information system system 50 and the data center 80.

The second control system system 60 is a system provided in the same place where the second sensor group 30 is installed. The second control system system 60 includes one or more control devices 61, a monitoring device 62, and a router 63.
The control device 61 is configured by using an information processing device such as a personal computer. The control device 61 acquires measured values from each sensor constituting the second sensor group 30. The control device 61 transmits the acquired measured value to the monitoring device 62.

The monitoring device 62 is configured by using an information processing device such as a personal computer. The monitoring device 62 monitors each control device 61 and collects the measured values transmitted from the control device 61. The monitoring device 62 transmits the collected measured values to the data center 80 via the router 63 and the network 90.

The monitoring device 62 passively or actively transmits the measured value to the data center 80. When the monitoring device 62 passively transmits the measured value, the monitoring device 62 transmits the measured value to the data center 80 in response to a request from the data center 80. Further, when the monitoring device 62 actively transmits the measured value, the monitoring device 62 periodically (for example, every 1 minute, 10 minutes, 30 minutes, 1 hour) transmits the measured value to the data center 80.
The router 63 relays communication between the second control system 60 and other systems and between the second control system 60 and the data center 80.

The second information system system 70 is a system that performs maintenance and management of the second sensor group 30 and the second control system system 60. The second information system system 70 may be installed in the same place as the second control system system 60, or may be installed in a place different from the second control system system 60. The second information system system 70 includes a client device 71, an equipment maintenance device 72, and a router 73.

The client device 71 is configured by using an information processing device such as a smartphone, a mobile phone, a tablet terminal, a notebook computer, or a personal computer. The client device 71 is, for example, a terminal carried by an operator who inspects the second sensor group 30. The client device 71 transmits the inspection result of the second sensor group 30 to the equipment maintenance device 72.

The equipment maintenance device 72 is configured by using an information processing device such as a personal computer. The equipment maintenance device 72 manages the operating state of each device provided in the second control system system 60 and the operating state of the second sensor group 30. Further, the equipment maintenance device 72 acquires the inspection result transmitted from the client device 71. The equipment maintenance device 72 holds the acquired inspection result. When the equipment maintenance device 72 obtains the inspection result before calibration and the inspection result after calibration as the inspection result of the same sensor on the same inspection date, the inspection result before calibration and the inspection after calibration are obtained. Holds the result in association with it.

The equipment maintenance device 72 passively or actively transmits the inspection result to the data center 80. When the equipment maintenance device 72 passively transmits the inspection result, the equipment maintenance device 72 transmits the inspection result to the data center 80 in response to a request from the data center 80. Further, when the equipment maintenance device 72 actively transmits the inspection result, the equipment maintenance device 72 periodically (for example, every 1 minute, 10 minutes, 30 minutes, 1 hour) transmits the inspection result to the data center 80.
The router 73 relays communication between the second information system 70 and other systems and between the second information system 70 and the data center 80.

The data center 80 holds information transmitted from each system (first control system system 40, first information system system 50, second control system system 60, and second information system system 70). The data center 80 includes a router 81 and a diagnostic device 82.
The router 81 relays communication between the diagnostic device 82 and other systems.

The diagnostic device 82 performs abnormality diagnosis processing for each sensor group based on the information transmitted from each system. More specifically, the diagnostic device 82 performs abnormality diagnosis processing for each diagnosis unit and each diagnosis group preset in the sensor group. Here, the diagnostic unit is a group composed of a plurality of sensors. The same sensor may belong to a plurality of diagnostic units. The diagnostic group is a group composed of a plurality of diagnostic units. The same diagnostic unit may belong to a plurality of diagnostic groups.
Hereinafter, a plurality of embodiments will be described in detail as an example.

(First Embodiment)
FIG. 2 is a schematic block diagram showing a functional configuration of the diagnostic apparatus 82 according to the first embodiment.
The diagnostic device 82 includes a CPU (Central Processing Unit), a memory, an auxiliary storage device, and the like connected by a bus, and executes a diagnostic program. By executing the diagnostic program, the diagnostic device 82 has an inspection data acquisition unit 821, a related information storage unit 822, a period determination unit 823, a time series data acquisition unit 824, a time series data storage unit 825, and a time series data creation unit 826 for diagnosis. It functions as a device including a time-series data storage unit for diagnosis 827, a diagnosis unit 828, a diagnosis result storage unit 829, a reliability evaluation unit 830, a reliability storage unit 831, a display control unit 832, and a display unit 833. In addition, all or a part of each function of the diagnostic apparatus 82 may be realized by using hardware such as ASIC (Application Specific Integrated Circuit), PLD (Programmable Logic Device), and FPGA (Field Programmable Gate Array). The diagnostic program may also be recorded on a computer-readable recording medium. The computer-readable recording medium is, for example, a flexible disk, a magneto-optical disk, a portable medium such as a ROM or a CD-ROM, or a storage device such as a hard disk built in a computer system. Further, the diagnostic program may be transmitted and received via a telecommunication line.

The inspection data acquisition unit 821 acquires the inspection result of each sensor from the equipment maintenance device 52 and the equipment maintenance device 72. For example, the inspection data acquisition unit 821 may acquire the inspection result by requesting the equipment maintenance device 52 and the equipment maintenance device 72 for the inspection result at a predetermined cycle. Further, for example, the inspection data acquisition unit 821 may acquire the inspection results transmitted from the equipment maintenance device 52 and the equipment maintenance device 72. The inspection data acquisition unit 821 stores the acquired inspection result in the related information storage unit 822.

The related information storage unit 822 is configured by using a storage device such as a magnetic hard disk device or a semiconductor storage device. The related information storage unit 822 stores the inspection data table 200, the diagnostic unit definition table 300, the diagnostic group definition table 400, and the diagnostic table 500. The inspection data table 200, the diagnostic unit definition table 300, the diagnostic group definition table 400, and the diagnostic table 500 will be described in detail with reference to FIGS. 3 to 6.

FIG. 3 is a diagram showing a specific example of the inspection data table 200. The inspection data table 200 has a plurality of records (hereinafter referred to as "inspection data records") representing information regarding inspection results. The inspection data record has each value of inspection date, sensor ID, sensor name, measured value before calibration, measured value before calibration, bias before calibration, measured value after calibration, measured value after calibration, and bias after calibration. The measured value after calibration, the measured value after calibration, and the bias value after calibration do not have to be registered in all the inspection data records. For example, each value of the measured value after calibration, the measured value after calibration, and the bias after calibration is registered when the value of the bias before calibration is a threshold value (for example, ± 5) or more.

The inspection date represents the date on which the inspection of the sensor was performed. The sensor ID represents identification information for identifying the sensor inspected. The sensor ID may be, for example, the MAC address of the sensor, or may be information to which identification information for identifying a base where the sensor is installed is added to the MAC address of the sensor. The sensor name represents the name of the sensor.

The measured value before calibration represents the value measured by the worker before the sensor is calibrated. The pre-calibration measurement value represents the value measured by the sensor before the sensor is calibrated. The pre-calibration bias represents the amount of deviation between the pre-calibration measured value and the pre-calibration measured value. The pre-calibration bias may be a value calculated based on the measured value before calibration, or may be a value calculated based on the measured value before calibration. When the pre-calibration measured value is used as a reference, the pre-calibration bias is the value obtained by subtracting the pre-calibration measured value from the pre-calibrated measured value. When the pre-calibration measurement value is used as a reference, the pre-calibration bias is the value obtained by subtracting the pre-calibration measurement value from the pre-calibration measurement value. In this embodiment, a case where the measured value before calibration is used as a reference will be described as an example.

The measured value after calibration represents the value measured by the worker after the sensor is calibrated. The measured value after calibration represents the value measured by the sensor after the sensor has been calibrated. The post-calibration bias represents the amount of deviation between the measured value after calibration and the measured value after calibration. The post-calibration bias may be a value based on the measured value after calibration or a value based on the measured value after calibration. In the case of a value based on the measured value after calibration, the value obtained by subtracting the measured value after calibration from the measured value after calibration is the bias after calibration. In the case of a value based on the measured value after calibration, the value obtained by subtracting the measured value after calibration from the measured value after calibration is the bias after calibration.

In the example shown in FIG. 3, a plurality of inspection data records are registered in the inspection data table 200. In the inspection data record 201 in FIG. 3, the inspection date value is "2017/2/2", the sensor ID value is "01120-A1-4", the sensor name value is "PEP-4", and the measured value before calibration. The value of is "21.7", the value of the measured value before calibration is "21.7", the value of bias before calibration is "0.0", the value of the measured value after calibration is "-", and the value of the measured value after calibration is "-". The value is "-" and the value of the bias after calibration is "-". As described above, since the sensor "PEP-4" registered in the inspection data record 201 does not need to be calibrated because the value of the bias before calibration is less than the threshold value, the data after calibration is not registered.

On the other hand, in the inspection data record 202 in FIG. 3, the value of the inspection date is "2017/2/3", the value of the sensor ID is "01120-A1-5", the value of the sensor name is "PEP-5", and the calibration is performed. The value of the pre-measured value is "21.3", the value of the pre-calibration measured value is "22.1", the value of the pre-calibration bias is "-0.8", and the value of the post-calibration measured value is "21.8". , The value of the measured value after calibration is "22.0", and the value of the bias after calibration is "-0.2". As described above, the sensor "PEP-5" registered in the inspection data record 202 is calibrated because the value of the pre-calibration bias is equal to or higher than the threshold value. Therefore, the data after calibration is registered in the inspection data record 202.

FIG. 4 is a diagram showing a specific example of the diagnostic unit definition table 300. The diagnostic unit definition table 300 has a plurality of records (hereinafter referred to as “diagnosis unit records”) representing information regarding the diagnostic unit. The diagnostic unit record has each value of the diagnostic unit ID and the belonging sensor ID.
The diagnostic unit ID represents identification information for identifying the diagnostic unit. The sensor ID represents identification information for identifying the sensor inspected.

In the example shown in FIG. 4, a plurality of diagnostic unit records are registered in the diagnostic unit definition table 300. In the diagnostic unit record 301 in FIG. 4, the value of the diagnostic unit ID is "A1", the value of the belonging sensor ID is "01120-A1-4", "01120-A1-2", "01120-A1-3", " 01120-A1-4 "," 01120-A1-5 "," 01120-A1-6 "," ... ". As described above, the diagnostic unit identified by the diagnostic unit ID "A1" includes at least the sensor IDs "01120-A1-4", "01120-A1-2", "01120-A1-3", and "01120-A1". It is indicated that the sensor identified by "-4", "01120-A1-5", and "01120-A1-6" belongs to it.

FIG. 5 is a diagram showing a specific example of the diagnostic group definition table 400. The diagnostic group definition table 400 has a plurality of records (hereinafter referred to as “diagnosis group records”) representing information regarding the diagnostic group. The diagnosis group record has each value of a diagnosis group ID, a affiliation diagnosis unit ID, a diagnosis date, a first number of days, a second number of days, and a data reliability level determination number of days.

The diagnostic group ID represents identification information for identifying the diagnostic group. The affiliation diagnostic unit ID represents identification information for identifying a diagnostic unit belonging to a diagnostic group. The diagnosis date represents the date on which the abnormality diagnosis process is performed. The diagnosis date is registered in advance by the worker. The first number of days represents the number of days to be included in determining the period (hereinafter referred to as "reference data period") for extracting the reference measurement value (hereinafter referred to as "reference data") used for the abnormality diagnosis process. The second number of days is the number of days included to determine the period (hereinafter referred to as "diagnosis data period") for extracting the measured value (hereinafter referred to as "diagnosis data") to be the target of abnormality diagnosis used in the abnormality diagnosis process. Represents. The number of days for determining the data reliability level represents the number of days as a reference for determining the high reliability of the measured value used for the abnormality diagnosis process.

In the example shown in FIG. 5, a plurality of diagnostic group records are registered in the diagnostic group definition table 400. In the diagnosis group record 401 in FIG. 5, the value of the diagnosis group ID is "S1", the value of the belonging diagnosis unit ID is "A1", "...", the value of the diagnosis date is "20XX / YY / XX", and the first. The value of the number of days of 1 is "31 days", the value of the second number of days is "7 days", and the value of the data reliability level determination days is "31 days". As described above, at least the diagnostic unit identified by the diagnostic unit ID "A1" belongs to the diagnostic group identified by the diagnostic group ID "S1", and the diagnostic group identified by the diagnostic group ID "S1" is defined. The day when the diagnostic process is executed is "20XX / YY / XX", the number of days included to determine the reference data period is "31 days", and the number of days included to determine the diagnostic data period is "7 days". It is indicated that the number of days as a reference for determining the high reliability of the measured value used for the abnormality diagnosis process is "31 days".

FIG. 6 is a diagram showing a specific example of the diagnostic table 500. The diagnosis table 500 has a plurality of records (hereinafter referred to as "abnormality diagnosis processing records") representing information regarding the abnormality diagnosis processing. The abnormality diagnosis processing record has each value of the diagnosis group ID, the diagnosis date, and the diagnosis completion.
Completion of diagnosis indicates whether or not the abnormality diagnosis process has been performed. In the present embodiment, when it indicates that the abnormality diagnosis process has been performed, "completed" is registered in the item of diagnosis completion, and when it indicates that the abnormality diagnosis process has not been performed, the item of diagnosis completion. It is assumed that "not yet" is registered in.

In the example shown in FIG. 6, a plurality of abnormality diagnosis processing records are registered in the diagnosis table 500. In the abnormality diagnosis processing record 501 in FIG. 6, the value of the diagnosis group ID is "S1", the value of the diagnosis date is "20XX / YY / ZZ", and the value of the diagnosis completion is "not yet". As described above, the diagnosis group identified by the diagnosis group ID "S1" indicates that the abnormality diagnosis process has not been completed.

Further, in the abnormality diagnosis processing record 502 in FIG. 6, the value of the diagnosis group ID is "S2", the value of the diagnosis date is "20XX / YY / ZZ", and the value of the diagnosis completion is "finished". As described above, the diagnosis group identified by the diagnosis group ID “S2” indicates that the abnormality diagnosis process has been completed.

Returning to FIG. 2, the description of the diagnostic apparatus 82 will be continued.
The period determination unit 823 determines the extraction start position of the measured value used for the abnormality diagnosis process of the diagnostic unit based on the inspection date of each sensor constituting the diagnostic unit. Then, the period determination unit 823 determines the period for specifying the range for extracting the measured value with the determined extraction start position as a reference. Specifically, the period determination unit 823 determines the reference data period and the diagnostic data period as the period for specifying the range for extracting the measured value. Hereinafter, the reference data period and the diagnostic data period will be described.

The reference data period is set with reference to the first reference date. The first reference date is the latest date of inspection of all sensors belonging to the diagnostic unit. The reference data is a measured value that serves as a reference used in the abnormality diagnosis process as described above. Therefore, it is desirable that the reference data is a more accurate measured value, for example, a measured value acquired immediately after the inspection.

Therefore, the period determination unit 823 determines the period from a certain day after the first reference date to the day when the first number of days (for example, several days) has elapsed as the reference data period. For example, the period determination unit 823 determines the period from the day following the first reference date to the day when 31 days have passed as the reference data period. The period determination unit 823 acquires the first number of days from the diagnosis group definition table 400. In the present embodiment, "a certain day after the first reference date" will be described as "the day after the first reference date". The period determination unit 823 notifies the diagnostic time series data creation unit 826 of the determined reference data period. The period determination unit 823 determines the reference data period for each diagnostic unit.

A method for determining the reference data period will be described with a specific example. First, the sensors belonging to the diagnostic unit are referred to as sensor 1, sensor 2, and sensor 3. Here, it is assumed that the sensor 1 is inspected on January 1, the sensor 2 is inspected on January 2, and the sensor 3 is inspected on January 3. In this case, the period determination unit 823 determines January 3 as the first reference date. Then, the period determination unit 823 is based on the period from January 4th to February 2nd, which is the period from January 4th, which is the day after January 3rd, to February 2nd, which is the day when 31 days have passed. Determined as the data period.

The diagnostic data period is set with respect to the second reference date. The second reference date is the date on which the abnormality diagnosis process is performed for the diagnosis group. The diagnostic data is a measured value that is the target of abnormality diagnosis as described above. Therefore, it is desirable that the diagnostic data is, for example, a measured value acquired before the abnormality diagnosis process.

Therefore, the period determination unit 823 determines the period from a certain day before the second reference date to the day retroactive to the second number of days (for example, several days) as the diagnostic data period. For example, the period determination unit 823 determines the period from the day before the second reference date to the day retroactive by 7 days as the diagnostic data period. The period determination unit 823 acquires the second number of days from the diagnosis group definition table 400. In the present embodiment, "a certain day before the second reference date" will be described as "the day before the second reference date". The period determination unit 823 notifies the diagnostic time series data creation unit 826 of the determined diagnostic data period. Further, the period determination unit 823 determines the diagnostic data period for each diagnostic group.

A method for determining the diagnostic data period will be described with a specific example. First, it is assumed that the abnormality diagnosis process for a certain diagnosis group is performed on June 13. In this case, the period determination unit 823 determines June 13 as the second reference date. Then, the period determination unit 823 determines the period from June 6 to June 12, which is the period from June 12, which is the day before June 13, to June 6, which is the day when it goes back seven days. Determined as a period.

Further, the period determination unit 823 determines a period for evaluating the data reliability level of the reference data used for the abnormality diagnosis process (hereinafter referred to as “evaluation period”) for each diagnosis unit based on the inspection data table 200. Specifically, the period determination unit 823 determines the number of days from the oldest date to the latest inspection date (inspection date) of the sensor belonging to the diagnostic unit as the evaluation period. The period determination unit 823 performs this processing for each diagnostic unit belonging to the diagnostic group in which the abnormality diagnosis processing is performed. The period determination unit 823 notifies the reliability evaluation unit 830 of the determined evaluation period.

The data reliability level represents an index of high reliability of the reference data used for the abnormality diagnosis process. The data reliability level is represented by, for example, "high" and "low". The data reliability level may be represented by three or more levels (for example, “high”, “medium”, and “low”). When the data reliability level is "high", it means that the reference data used for the abnormality diagnosis process has high reliability.

Here, high reliability means that the bias of the inspection data used as the reference data may be small (the bias is close to 0). That is, by using the reference data having the data reliability level of "high" for the abnormality diagnosis processing, the abnormality diagnosis becomes possible more accurately.
On the other hand, when the data reliability level is "low", it means that the reference data used for the abnormality diagnosis processing has low reliability. Here, low reliability means that the bias of the inspection data used as the reference data may not be small (the bias is farther from 0).

The time-series data acquisition unit 824 acquires the measured values of each sensor as time-series data from the monitoring device 42 and the monitoring device 62. For example, the time-series data acquisition unit 824 may acquire the measured value of each sensor by requesting the monitoring device 42 and the monitoring device 62 for the measured value of each sensor at a predetermined cycle. Further, for example, the time-series data acquisition unit 824 may acquire the measured values of each sensor transmitted from the monitoring device 42 and the monitoring device 62.

The time-series data storage unit 825 is configured by using a storage device such as a magnetic hard disk device or a semiconductor storage device. The time-series data storage unit 825 stores the time-series data acquired by the time-series data acquisition unit 824. FIG. 7 shows a specific example of the time-series data stored in the time-series data storage unit 825.

FIG. 7 is a diagram showing a specific example of the time series data 600. The time-series data 600 is configured by arranging the measured values measured by each sensor in time series for each sensor. In the example shown in FIG. 7, at 3:00 AM (region indicated by reference numeral 601) on April 1, 2017, the measured value is “21.6” by the sensor identified by the sensor ID “01120-A1-4”. "Is measured, and it is shown that" 23.7 "is measured as a measured value by the sensor identified by the sensor ID" 01120-A1-5 ".

Further, the region indicated by reference numeral 602 in the time series data 600 represents the measured value measured by the sensor identified by the sensor ID “01120-A1-4”. Further, the region indicated by reference numeral 603 in the time series data 600 represents the measured value measured by the sensor identified by the sensor ID “01120-A1-5”.
The time-series data 600 is accumulated every time a measured value is acquired by the time-series data acquisition unit 824.

Returning to FIG. 2, the description of the diagnostic apparatus 82 will be continued.
The diagnostic time series data creation unit 826 is based on the reference data period and the diagnostic data period notified from the period determination unit 823, the inspection data table 200, and the time series data 600, and the diagnostic time series for each diagnostic group. Create the data. The diagnostic time series data is a measured value used in the abnormality diagnosis process. More specifically, the diagnostic time series data includes reference data and diagnostic data.

The reference data in the diagnostic time-series data may be a measured value for a reference data period that differs for each diagnostic unit. This is because the reference data period may differ for each diagnostic unit belonging to the diagnostic group.
The diagnostic data in the diagnostic time-series data is the measured value in the same reference data period in the diagnostic unit. This is because the reference data period is common to each diagnostic unit belonging to the diagnostic group.

Further, the diagnostic time-series data creation unit 826 performs correction processing on the reference data when creating the diagnostic time-series data. For example, the diagnostic time-series data creation unit 826 adds or subtracts the bias (pre-calibration bias or post-calibration bias) for each sensor registered in the inspection data table 200 to the reference data of the same sensor as a correction process. do. By the correction process, the diagnostic time series data creation unit 826 creates reference data having a bias close to zero. That is, the diagnostic time-series data creation unit 826 adds or subtracts a bias from the reference data so that the bias is close to zero. The diagnostic time-series data creation unit 826 stores the created diagnostic time-series data in the diagnostic time-series data storage unit 827.

The diagnostic time-series data storage unit 827 temporarily stores the diagnostic time-series data created by the diagnostic time-series data creation unit 826. FIG. 8 shows a specific example of the time series data for diagnosis. FIG. 8 is a diagram showing a specific example of diagnostic time series data. As shown in FIG. 8, the diagnostic time series data 700 includes a diagnostic group ID, a diagnostic unit ID, a sensor ID, a reference data, and a diagnostic data. The diagnostic time series data 700 is newly generated every time the abnormality diagnosis process is performed.

Returning to FIG. 2, the description of the diagnostic apparatus 82 will be continued.
The diagnostic unit 828 performs an abnormality diagnosis process based on the diagnostic time-series data 700 stored in the diagnostic time-series data storage unit 827. Existing technology is applied to the specific method of abnormality diagnosis processing. For example, as a specific method of abnormality diagnosis processing, the technique described in Reference 1 below may be used.
(Reference 1: Japanese Unexamined Patent Publication No. 2016-16658, bias estimation device and method, and failure diagnosis device and method)

The diagnosis result storage unit 829 is configured by using a storage device such as a magnetic hard disk device or a semiconductor storage device. The diagnosis result storage unit 829 stores the result of the abnormality diagnosis process performed by the diagnosis unit 828.
The reliability evaluation unit 830 evaluates the data reliability level as reference data based on the evaluation period notified from the period determination unit 823 and the number of days for determining the data reliability level.
The reliability storage unit 831 is configured by using a storage device such as a magnetic hard disk device or a semiconductor storage device. The reliability storage unit 831 stores the data reliability level evaluated by the reliability evaluation unit 830.

The display control unit 832 generates display screen data to be displayed on the display unit 833 based on the data reliability level stored in the reliability storage unit 831 and the diagnosis result stored in the diagnosis result storage unit 829. do. The display control unit 832 causes the display unit 833 to display the generated display screen data. The display screen data includes at least the diagnostic unit name, the data reliability level for each diagnostic unit, and the data reliability level for the diagnostic group.

The display unit 833 is an image display device such as a liquid crystal display, an organic EL (Electro Luminescence) display, and a CRT (Cathode Ray Tube) display. The display unit 833 displays the image data. The display unit 833 may be an interface for connecting the image display device to the diagnostic device 82. In this case, the display unit 833 generates a video signal for displaying the image data, and outputs the video signal to the image display device connected to the display unit 833.

FIG. 9 is a flowchart showing the flow of the abnormality diagnosis process performed by the diagnostic apparatus 82. The process of FIG. 9 is executed at the timing when the abnormality diagnosis process is performed.
The period determination unit 823 determines whether or not there is a diagnosis group that requires diagnosis (step S101). Specifically, first, the period determination unit 823 reads out the diagnostic table 500 stored in the related information storage unit 822. Next, the period determination unit 823 refers to the items of the diagnosis date and the diagnosis completion in the diagnosis table 500, and today is the diagnosis group targeted for the abnormality diagnosis processing, and the abnormality diagnosis processing is executed. Determine if there are any diagnostic groups. That is, the period determination unit 823 determines whether or not there is a diagnosis group whose diagnosis date is today and whose diagnosis has not been completed.

If there is a diagnosis group whose diagnosis date is today and there is a diagnosis group whose diagnosis is "not yet" completed, the period determination unit 823 determines that there is a diagnosis group that requires diagnosis.
On the other hand, if there is no diagnosis group whose diagnosis date is today, or if there is no diagnosis group whose diagnosis is "not yet", the period determination unit 823 determines that there is no diagnosis group that requires diagnosis.

If there is no diagnostic group that requires diagnosis (step S101-NO), the period determination unit 823 ends the abnormality diagnosis process.
On the other hand, when there is a diagnosis group requiring diagnosis (step S101-YES), the period determination unit 823 determines a diagnosis unit to be diagnosed from the diagnosis groups requiring diagnosis (step S102). For example, the period determination unit 823 determines the diagnostic unit that has not been subjected to the abnormality diagnosis process as the diagnostic unit to be diagnosed.

Next, the period determination unit 823 determines the reference data period in the determined diagnostic unit (step S103). Specifically, first, the period determination unit 823 reads out the inspection data table 200 and the diagnostic unit definition table 300 stored in the related information storage unit 822. Next, the period determination unit 823 refers to the read-out diagnostic unit definition table 300, and acquires the sensor ID belonging to the determined diagnostic unit. Next, the period determination unit 823 refers to the inspection data table 200 and selects the inspection data record corresponding to the acquired sensor ID.

Next, the period determination unit 823 refers to the inspection date item of the selected inspection data record, and determines the latest inspection date among the sensors belonging to the diagnostic unit as the first reference date. Then, the period determination unit 823 determines the period from a certain day after the determined first reference date to the day when the first number of days has elapsed as the reference data period. The period determination unit 823 notifies the diagnostic time series data creation unit 826 of the determined reference data period.

Next, the period determination unit 823 determines the diagnostic data period in the determined diagnostic unit (step S104). Specifically, first, the period determination unit 823 determines the diagnosis date as the second reference date. Next, the period determination unit 823 determines the period from the day before the second reference date to the day retroactive in the second number of days as the diagnostic data period. The period determination unit 823 notifies the diagnostic time series data creation unit 826 of the determined diagnostic data period.

The period determination unit 823 determines whether or not the period has been determined for all the diagnostic units (step S105). When the period has not been determined for all the diagnostic units (step S105-NO), the period determination unit 823 repeatedly executes the processes after step S102.
On the other hand, when the period is determined for all the diagnostic units (step S105-YES), the diagnostic time series data creation unit 826 has the reference data period and the diagnostic data period notified from the period determination unit 823, and the inspection data. Diagnostic time-series data is created based on the table 200 and the time-series data 600 (step S106). The diagnostic time-series data creation unit 826 temporarily stores the created diagnostic time-series data in the diagnostic time-series data storage unit 827.

The diagnostic unit 828 performs an abnormality diagnosis process based on the diagnostic time-series data 700 stored in the diagnostic time-series data storage unit 827 (step S107). After performing the abnormality diagnosis processing, the diagnosis unit 828 stores the result of the abnormality diagnosis processing in the diagnosis result storage unit 829. After that, the diagnosis unit 828 refers to the diagnosis table 500 and selects the abnormality diagnosis processing record 501 corresponding to the diagnosis group in which the abnormality diagnosis processing is performed. Then, the diagnosis unit 828 changes the diagnosis completion item of the selected abnormality diagnosis processing record 501 from “not yet” to “finished” (step S108). After that, the diagnostic apparatus 82 executes the processes after step S101.

FIG. 10 is a flowchart showing a flow of evaluation processing of the data reliability level performed by the diagnostic apparatus 82. The process of FIG. 10 may be executed after the process of FIG. 9 is performed, or may be executed during the process of FIG. 9 (for example, after step S107).
The period determination unit 823 determines a diagnostic unit for which the data reliability level is evaluated (step S201). For example, the period determination unit 823 determines the diagnostic unit that has not been selected since the process of FIG. 10 is executed as the diagnostic unit to be evaluated for the data reliability level. Next, the period determination unit 823 determines the evaluation period in the determined diagnostic unit (step S202).

Specifically, first, the period determination unit 823 reads out the inspection data table 200 and the diagnostic unit definition table 300 stored in the related information storage unit 822. Next, the period determination unit 823 refers to the read-out diagnostic unit definition table 300, and acquires the sensor ID belonging to the determined diagnostic unit. Next, the period determination unit 823 refers to the inspection data table 200 and selects the inspection data record corresponding to the acquired sensor ID. Next, the period determination unit 823 refers to the inspection date item of the selected inspection data record, and evaluates the number of days from the oldest date to the latest inspection date (inspection date) of the sensor belonging to the diagnostic unit. To be determined as. The period determination unit 823 notifies the reliability evaluation unit 830 of the information on the evaluation period of the determined diagnostic unit.

After that, the period determination unit 823 determines whether or not there is a diagnostic unit whose evaluation period has not been determined (step S203). For example, the period determination unit 823 refers to the diagnostic unit definition table 300, and determines the evaluation period depending on whether or not the evaluation period of all the diagnostic units registered in the diagnostic unit definition table 300 is determined. It may be determined whether or not there is a diagnostic unit. In this case, the period determination unit 823 determines that there is a diagnostic unit whose evaluation period has not been determined when the evaluation period of all the diagnostic units registered in the diagnostic unit definition table 300 has not been determined. Further, for example, the period determination unit 823 refers to the item of the diagnosis date in the diagnosis table 500, and determines whether or not the evaluation period of all the diagnosis units belonging to the diagnosis group to which the abnormality diagnosis process is performed today is determined. Accordingly, it may be determined whether or not there is a diagnostic unit whose evaluation period has not been determined. In this case, the period determination unit 823 determines that there is a diagnostic unit whose evaluation period has not been determined if the evaluation period of all the diagnostic units belonging to the diagnostic group subject to the abnormality diagnosis process has not been determined today. do.

If there is a diagnostic unit whose evaluation period has not been determined (step S203-YES), the period determination unit 823 repeatedly executes the processes after step S201.
On the other hand, when there is no diagnostic unit whose evaluation period has not been determined (step S203-NO), the reliability evaluation unit 830 performs data reliability based on the evaluation period of each diagnostic unit and the number of days for determining the data reliability level for each diagnostic group. The degree level is evaluated (step S204).

Specifically, the reliability evaluation unit 830 compares the evaluation period of the diagnostic unit with the number of days for determining the data reliability level, and when the evaluation period is equal to or less than the number of days for determining the data reliability level, the data reliability of the diagnostic unit Evaluate the degree level as "high". Further, the reliability evaluation unit 830 compares the evaluation period of the diagnostic unit with the number of days for determining the data reliability level, and when the evaluation period is longer than the number of days for determining the data reliability level, determines the data reliability level of the diagnostic unit. Evaluate as "low". The reliability evaluation unit 830 evaluates the data reliability level for each diagnostic unit. The reliability evaluation unit 830 may evaluate the data reliability level for each sensor belonging to the diagnostic unit.

The reliability evaluation unit 830 evaluates the data reliability level for all the diagnostic units, and then evaluates the data reliability level for the diagnostic group. Specifically, the reliability evaluation unit 830 evaluates the data reliability level of the diagnostic group to which the diagnostic unit having the data reliability level of "low" belongs as "low". That is, when the diagnostic group includes a diagnostic unit having a data reliability level of "low", the reliability evaluation unit 830 evaluates the data reliability level of the diagnostic group as "low".

On the other hand, the reliability evaluation unit 830 evaluates the data reliability level of the diagnostic group to which the diagnostic unit having the data reliability level of "low" does not belong as "high". That is, when the data reliability level of all the diagnostic units belonging to the diagnostic group is "high", the reliability evaluation unit 830 evaluates the data reliability level of the diagnostic group as "high". After that, the reliability evaluation unit 830 stores the data reliability level in the reliability storage unit 831 (step S205). For example, the reliability evaluation unit 830 associates the evaluation of the data reliability level of the diagnostic group with the evaluation of the data reliability level of all the diagnostic units belonging to the diagnostic group for each diagnostic group, and the reliability storage unit 831. To memorize.

Next, a method for determining each period and an evaluation process for the data reliability level will be described with specific examples. 11 and 12 are diagrams showing specific examples of the determination method for each period and the evaluation process of the data reliability level.
First, a specific example shown in FIG. 11 will be described. In the description of FIG. 11, it is assumed that the settings are as follows.
<Settings>
-Diagnosis date is August 1, 2017-Diagnosis unit A is composed of sensors 1, 2 and 3-The first number of days is 31 days, the second number of days is 7 days-The number of data reliability level judgment days is 31st ・ The inspection date of sensor 1 is February 3, 2017, the inspection date of sensor 2 is February 6, 2017, and the inspection date of sensor 3 is February 7, 2017.

In the case of the above settings, the latest inspection date among the sensors 1, 2 and 3 is February 7, 2017 of the sensor 3. Therefore, the period determination unit 823 determines February 7, 2017 as the first reference date. In this case, the period determination unit 823 determines the period from February 8, 2017 to March 11, 2017, which is the period from the day after the first reference date to the day when 31 days have passed, as the reference data period. In addition, the period determination unit 823 determines August 1, 2017, which is the diagnosis date, as the second reference date. In this case, the period determination unit 823 determines the period from July 25, 2017 to July 31, 2017, which is the period from the day before the second reference date to the day retroactive by 7 days, as the diagnostic data period.

Further, in the sensors 1, 2 and 3 belonging to the diagnostic unit A, the oldest inspection date is February 3, 2017 of the sensor 1, and the latest inspection date is February 7, 2017 of the sensor 3. It's a day. Therefore, the period determination unit 823 determines that the evaluation period of the diagnostic unit A is 4, which is the period from the oldest day to the latest day. The reliability evaluation unit 830 compares the evaluation period “4” determined by the period determination unit 823 with the data reliability level determination days “31”. Since the evaluation period “4” is equal to or less than the number of days for determining the data reliability level “31”, the reliability evaluation unit 830 evaluates the data reliability level of the diagnostic unit A as “high”.

Next, a specific example shown in FIG. 12 will be described. In the description of FIG. 12, it is assumed that the settings are as follows.
<Settings>
-Diagnosis date is August 1, 2017-Diagnosis unit B is composed of sensors 1, 2, 3, 4, 5, 6-The first number of days is 31 days, the second number of days is 7 days-Data The reliability level judgment days are 31 days. ・ The inspection date of the sensor 1 is February 7, 2017, the inspection date of the sensor 2 is February 8, 2017, the inspection date of the sensor 3 is February 9, 2017, and the sensor 4 The inspection date of the sensor 5 is March 5, 2017, the inspection date of the sensor 5 is March 6, 2017, and the inspection date of the sensor 6 is March 10, 2017.

In the case of the above settings, the latest inspection date among the sensors 1, 2, 3, 4, 5, and 6 is March 10, 2017 of the sensor 6. Therefore, the period determination unit 823 determines March 10, 2017 as the first reference date. In this case, the period determination unit 823 determines the period from March 11, 2017 to April 11, 2017, which is the period from the day after the first reference date to the day when 31 days have passed, as the reference data period. In addition, the period determination unit 823 determines August 1, 2017, which is the diagnosis date, as the second reference date. In this case, the period determination unit 823 determines the period from July 25, 2017 to July 31, 2017, which is the period from the day before the second reference date to the day retroactive by 7 days, as the diagnostic data period.

Further, in the sensors 1, 2, 3, 4, 5, and 6 belonging to the diagnostic unit B, the oldest inspection date is February 7, 2017 of the sensor 1, and the latest inspection date is the sensor 6. It is March 10, 2017. Therefore, the period determination unit 823 determines that the evaluation period of the diagnostic unit B is 32, which is the period from the oldest day to the latest day. The reliability evaluation unit 830 compares the evaluation period “32” determined by the period determination unit 823 with the data reliability level determination days “31”. Since the evaluation period “32” is equal to or longer than the data reliability level determination days “31”, the reliability evaluation unit 830 evaluates the data reliability level of the diagnostic unit B as “low”. Further, the reliability evaluation unit 830 also evaluates the data reliability level of the diagnostic group including the diagnostic unit B as “low”.

The diagnosis result for each diagnosis group obtained as described above and the data reliability level for each diagnosis group are displayed on the display unit 833 according to the user's request. FIG. 13 is a diagram showing an example of a display screen displayed on the display unit 833.
On the display screen shown in FIG. 13, the diagnosis result and the data reliability level of a certain diagnosis group xxxxx are displayed. For example, region 801 in FIG. 13 shows the data reliability level of the diagnostic group xxxxx. In this example, the data confidence level for diagnostic group xxxxx is shown to be "low". Area 802 in FIG. 13 shows the priority of inspection of each diagnostic unit. The inspection priority of each diagnostic unit is determined based on the diagnostic results. For example, the display control unit 832 displays the information of each diagnostic unit so that the inspection priority is higher in descending order of the abnormality score in the diagnosis result. In FIG. 13, it is shown that the diagnostic unit having the highest inspection order in the diagnostic group xxx, that is, the diagnostic unit in which the abnormality is likely to occur, is the system xxx.

The name of each diagnostic unit is shown in the area 803 in FIG. The area 804 in FIG. 13 shows the data reliability level of each diagnostic unit. As shown in FIG. 13, when a data shortage or an abnormality occurs in the calculation in the abnormality diagnosis process, the data reliability level of the corresponding diagnostic unit is not displayed. This is because if there is a shortage of data or an abnormality occurs in the calculation, the diagnosis has not been completed and it is not helpful to evaluate the reliability of the data.

According to the diagnostic device 82 configured as described above, based on the inspection date of the time-series data acquisition unit 824 that collects measured values from a plurality of sensors provided in the facility and the diagnostic unit composed of the plurality of sensors. It is provided with a period determination unit 823 that determines an extraction start position of reference data as a reference used for abnormality diagnosis processing of the diagnostic unit from the collected measured values. This makes it possible to determine the extraction start position of the reference data of the diagnostic unit that is the target unit of the abnormality diagnosis process. That is, the diagnostic device 82 can automatically determine the reference data. Therefore, it is possible to reduce the labor required by humans.

Further, the diagnostic device 82 determines the date close to the latest inspection date among the inspection dates of the plurality of sensors belonging to the diagnostic unit as the reference data extraction start position. Thereby, the measured value on the day close to the latest inspection date (for example, immediately after the inspection is performed) can be determined as the reference data. Therefore, the diagnostic device 82 can perform the abnormality diagnosis process by using the measured value with high accuracy as the reference data.

Further, the diagnostic device 82 determines the extraction start position of the diagnostic data to be the target of the abnormality diagnosis processing based on the execution date of the abnormality diagnosis processing. As a result, the extraction start position of the diagnostic data of the diagnostic unit, which is the target unit of the abnormality diagnosis processing, can be determined. That is, the diagnostic device 82 can automatically determine the diagnostic data. Therefore, it is possible to reduce the labor required by humans.

Further, the diagnostic device 82 determines a day close to the execution date of the abnormality diagnosis process as the extraction start position of the diagnostic data. Thereby, the measured value on the day close to the day when the abnormality diagnosis process is executed (for example, the day before the abnormality diagnosis process is executed) can be determined as the diagnostic data. Therefore, the diagnostic apparatus 82 can perform the abnormality diagnosis process with the latest measured value before executing the abnormality diagnosis process. As a result, abnormalities due to deterioration over time can be detected more accurately.

Further, the diagnostic device 82 corrects the reference data by calculating the bias obtained at the time of inspecting the sensor with respect to the reference data. Specifically, the diagnostic device 82 can create reference data having a bias close to zero. That is, the deviation of the reference data can be corrected. Therefore, the diagnostic device 82 can execute the abnormality diagnosis process with higher accuracy.

Conventionally, an inspection worker or a system administrator may not be able to easily grasp whether the measured value used for the abnormality diagnosis process is a highly reliable measurement value, that is, an accurate measurement value. there were. In addition, when inspecting whether the measured value is accurate or not, it may take time and effort because it requires cost and manpower.
On the other hand, the diagnostic device 82 includes a diagnostic unit 828 that performs abnormality diagnosis processing of the diagnostic unit, and a reliability evaluation unit 830 that evaluates the data reliability level of the measured value obtained from the diagnostic unit based on the diagnostic unit. A display unit 833 that displays the result of the abnormality diagnosis process in association with the data reliability level is provided. As a result, the operator performing the inspection and the system administrator can grasp at a glance whether or not the measured value obtained from the diagnostic unit on which the abnormality diagnosis process is performed is an accurate measured value. That is, if the operator performing the inspection or the system administrator evaluates that the measured value obtained from the diagnostic unit is inaccurate, the measured value used for the abnormality diagnosis process is inaccurate. It can be grasped at a glance as a value. Furthermore, since it is performed automatically, there is no need to manually perform the work. Therefore, it is possible to reduce the time and effort of the user.

(Second embodiment)
In the second embodiment, the data reliability level is evaluated by estimating the reliability model using a plurality of past inspection dates, the presence / absence of the configuration at the time of inspection, and the error measurement result for each sensor before and after calibration. do. Hereinafter, a specific description will be given.
FIG. 14 is a diagram for explaining the flow of processing in the second embodiment. In FIG. 14, it is assumed that the diagnostic unit C is composed of sensors 1, 2, 3, 4, 5, and 6. Further, in FIG. 14, the latest inspection date is August 1, 2017, the past inspection date (inspection 1) is February 1, 2017, and the past inspection date (inspection 2) is 20174. It is the 1st of the month, and the past inspection date (inspection 3) is June 1, 2017.

In the situation shown in FIG. 14, the sensors 4 to 6 have sensor error information Δx4, Δx5, Δx6 in the inspection on August 1, 2017, which is the latest on the start date of the reference data period (for example, August 2, 2017). Is obtained, these errors are corrected for the measured values in August 2017, the true data is estimated, and the data is used as the reference data. The reliability of these data is "high".
On the other hand, the sensors 1 to 3 have sensor error information Δx1 (i) measured in the past inspection 1 on February 1, 2017, inspection 2 on April 1, 2017, and inspection 3 on June 1, 2017. ), Δx2 (i), Δx3 (i), i = 1, 2, 3.

From the above example, the reference data for the sensors 1 to 3 is the reference data (2017 8) using the most recent Δx1 (3), Δx2 (3), and Δx3 (3) among the past inspection error data. Make corrections for the month). At this time, the reliability model for the values of each sensor 1 to 3 is calculated by the following.
Reliability of sensor 1 = σ1 ^* / σ1 ({Δx1 (1), Δx1 (2), Δx1 (3)})
Reliability of sensor 2 = σ2 ^* / σ2 ({Δx2 (1), Δx2 (2), Δx2 (3)})
Reliability of sensor 3 = σ3 ^* / σ3 ({Δx3 (1), Δx3 (2), Δx3 (3)})

Here, σi () represents the standard deviation with respect to the data in (), and σ (i) ^* is the normal error (resolution or measurement limit value) of each sensor. That is, in the present embodiment, the reliability of each sensor uses the reciprocal of the ratio of the past measured error standard deviation to the normal measurement limit error as the reliability index. The smaller the σi (), the higher the reliability. When the reliability is 1 or more, the reliability is set to "high", and when the reliability is less than 1, the reliability is set to "low", and multi-step reliability is defined as necessary. The method for deriving the reliability is not limited to the standard deviation, and may be derived by other statistical methods.

(Third embodiment)
In the third embodiment, as a method of correcting the reference data by the bias at the time of inspection, an error is obtained by using a plurality of past inspection dates, the presence / absence of the configuration at the time of inspection, and the error measurement result for each sensor before and after calibration. Estimate the growth model and correct the reference data with the bias estimated from the error growth model. Hereinafter, a specific description will be given.
FIG. 15 is a diagram for explaining the flow of processing in the third embodiment. In the example shown in FIG. 15, an error growth model for estimating the sensor error (as of August 1, 2017) with respect to the reference data is defined as follows.

-Inspections 1, 2, 3 and the reference data period start dates are t1, t2, t3, and tb, respectively.
The sensor errors measured in inspections 1, 2 and 3 are Δx (1), Δx (2) and Δx (3), and the estimated sensor error of the reference data period start date is Δx (b).
At this time, using the following error growth model, the estimated sensor error Δx (b) of the reference data period start date is obtained from the sensor error data (Δx (1), Δx (2), Δx (3)) at the time of past inspection. presume.

<Error growth model by least squares method>
a = {(t1-tm) (Δx (1) -Δxm) + (t2-tm) (Δx (2) -Δxm) + (t3-tm) (Δx (3) -Δxm)} / {(t1-) tm) ² + (t2-tm) ² + (t3-tm) ² }
b = Δxm-a · tm
However, Δxm is an average value of Δx (1), Δx (2), and Δx (3). That is, Δxm = {Δx (1) + Δx (2) + Δx (3)} / 3.
Further, tm is an average value of t1, t2, and t3. That is, tm = {t1 + t2 + t3} / 3.

By using the above error growth model, the estimated sensor error Δx (b) on the reference data period start date is estimated as Δx (b) = a · tb + b.

FIG. 16 is a diagram showing an example of sensor error estimation estimated using an error growth model. In FIG. 16, x indicates a measurement error on the inspection date. The square represents the estimated value of the sensor error on the reference date. The dotted line represents the estimation result by the error growth model.

The reliability model used in the second embodiment, the error growth model used in the third embodiment, the past inspection dates shown in the second embodiment and the third embodiment, and the configuration at the time of inspection. When the presence / absence of the above and the error measurement result for each sensor before and after calibration are displayed on the display unit 833, the result is as shown in FIG. Regarding the error growth model in FIG. 17, a diagram as shown in FIG. 16 for each sensor is displayed on the display unit 833 together with the diagram shown in FIG.

Hereinafter, modification common to each embodiment will be described.
The diagnostic system 100 has shown a configuration used as an abnormality diagnosis system for equipment installed in a building such as a building or a factory. The diagnostic system 100 is used as an abnormality diagnosis system for equipment installed in a treatment plant such as a water and sewage treatment plant, for example. It may be used.
In the present embodiment, the sensor is described as an example, but the equipment provided in the equipment does not have to be limited to the sensor. For example, the equipment provided in the equipment may be air conditioning equipment.

The diagnostic device 82 may be provided in a part or all of each system (for example, the first control system system 40, the first information system system 50, the second control system system 60, and the second information system system 70).

"A day after the first reference date" does not have to be limited to the day following the first reference date. For example, "a certain day after the first reference date" may not be a day that has passed too much from the first reference date, may be the day after the first reference date, or may be the day after the first reference date. It may be 3 days after.
"A day before the second base date" need not be limited to the day before the second base date. For example, "a day before the second reference date" may not be a day that goes back too far from the second reference date, may be the day before the second reference date, or may be the day before the second reference date. It may be 3 days before the day.

The diagnostic device 82 does not have to include the display unit 833. When configured in this way, the display control unit 832 causes the display unit 833 mounted on another device to display the generated display screen data via the network 90.
Further, the diagnostic device 82 does not have to include the reliability evaluation unit 830 and the reliability storage unit 831. When configured in this way, the period determination unit 823 does not have to determine the evaluation period. Further, the diagnostic device 82 does not have to execute the process shown in FIG.

The period determination unit 823, the diagnosis unit 828, the reliability evaluation unit 830, the reliability storage unit 831, the display control unit 832, and the display unit 833 may be configured as display devices.
The related information storage unit 822, the period determination unit 823, the time series data acquisition unit 824, and the time series data storage unit 825 may be configured as a determination device.
The related information storage unit 822, the period determination unit 823, the time series data acquisition unit 824, the time series data storage unit 825, and the diagnostic time series data creation unit 826 may be configured as a correction device.

The inspection result may be acquired by the inspection data acquisition unit 821 by reading the csv file prepared manually. In such a configuration, the diagnostic device 82 does not have to be communicably connected to the equipment maintenance device 52 and the equipment maintenance device 72.
The acquisition of the measured value (time-series data) by the time-series data acquisition unit 824 may be performed by reading the csv file prepared manually. In such a configuration, the diagnostic device 82 does not have to be communicably connected to the monitoring device 42 and the monitoring device 62.

According to at least one embodiment described above, a diagnosis is made from the collected data based on the inspection date of the collection unit that collects data from a plurality of devices provided in the facility and the inspection date of the unit composed of the plurality of devices. By having the period determination unit 823 for determining the extraction start position of the reference data as the reference used for the abnormality diagnosis processing of the unit, the time and effort of the user can be reduced.

Although some embodiments of the present invention have been described above, these embodiments are presented as examples and are not intended to limit the scope of the invention. These embodiments can be implemented in various other embodiments, and various omissions, replacements, and changes can be made without departing from the gist of the invention. These embodiments and variations thereof are included in the scope of the invention described in the claims and the equivalent scope thereof, as are included in the scope and gist of the invention.

100 ... Diagnostic system, 20 ... First sensor group, 30 ... Second sensor group, 40 ... First control system system, 41 ... Control device, 42 ... Monitoring device, 43 ... Router, 50 ... First information system system, 51 ... client device, 52 ... equipment maintenance device, 53 ... router, 60 ... second control system, 61 ... control device, 62 ... monitoring device, 63 ... router, 70 ... second control system, 71 ... client device, 72 ... Equipment maintenance device, 73 ... Router, 80 ... Data center, 81 ... Router, 82 ... Diagnostic device, 821 ... Inspection data acquisition unit, 822 ... Related information storage unit, 823 ... Period determination unit, 824 ... Time series data acquisition unit , 825 ... Time-series data storage unit, 826 ... Diagnostic time-series data creation unit, 827 ... Diagnostic time-series data storage unit, 828 ... Diagnosis unit, 829 ... Diagnosis result storage unit, 830 ... Reliability evaluation unit, 831 ... Reliability storage unit, 832 ... Display control unit, 833 ... Display unit

Claims (8)

A collection unit that collects data from multiple devices installed in the equipment,
Based on the inspection date of each device, a determination unit that determines the extraction start position of the reference data used as the reference for the abnormality diagnosis processing of the unit composed of the plurality of devices from the collected data.
Equipped with
The determination unit is a determination device that further determines a period for evaluating the reliability level of the reference data for each device. The determination device according to claim 1, wherein the determination unit determines the day after the inspection date of the plurality of devices belonging to the unit as the extraction start position. The determination device according to claim 1 or 2, wherein the determination unit determines an extraction start position of diagnostic data to be the target of the abnormality diagnosis process based on the execution date of the abnormality diagnosis process. The determination device according to claim 3, wherein the determination unit determines the day before the execution date of the abnormality diagnosis process as the extraction start position of the diagnostic data to be the target of the abnormality diagnosis process. The determination device according to any one of claims 1 to 4, and the determination device.
With a correction unit that corrects the reference data by calculating the bias obtained at the time of inspection of the equipment with respect to the reference data extracted from the data based on the extraction start position of the reference data determined by the determination device. ,
A correction device equipped with. With multiple devices installed in the equipment,
A collection unit that collects data from the plurality of devices,
Based on the inspection date of each device, a determination unit that determines the extraction start position of the reference data used as the reference for the abnormality diagnosis processing of the unit composed of the plurality of devices from the collected data.
Equipped with
The determination unit is a determination system that further determines a period for evaluating the reliability level of the reference data for each device . It is a determination method performed by a determination device that determines the extraction start position of reference data, which is a reference used for abnormality diagnosis processing of a unit composed of multiple devices.
A collection step that collects data from multiple devices in the equipment,
Based on the inspection date of each device, the determination step of determining the extraction start position of the reference data used as the reference for the abnormality diagnosis processing of the unit composed of the plurality of devices from the collected data, and the determination step.
Have,
A determination method for determining a period for evaluating the reliability level of the reference data for each device in the determination step. A collection step that collects data from multiple devices in the equipment,
Based on the inspection date of each device, the determination step of determining the extraction start position of the reference data used as the reference for the abnormality diagnosis processing of the unit composed of the plurality of devices from the collected data, and the determination step.
Let the computer run
In the determination step, a computer program for determining a period for evaluating the reliability level of the reference data for each device .

Images