JP6805554B2 - Monitoring device and monitoring method - Google Patents

Description

The present invention relates to a monitoring device and a monitoring method.

A method called Multivariate Statistical Process Control (MSPC) has been conventionally known as a method of analyzing operation data of equipment and plants in factories and the like to monitor and diagnose processes.

Further, in multivariate statistical process monitoring, a technique for giving a guideline for a plant manager or an operator to determine what kind of action should be taken in the event of an abnormality has been conventionally known (for example, patent documents). 1). In such a technique, for example, a model is created from operation data indicating a normal operation state of equipment, a plant, or the like, and then the occurrence of an abnormality is detected based on the model.

Japanese Unexamined Patent Publication No. 2014-96050

Here, for example, when the accuracy of abnormality detection deteriorates due to deterioration of equipment or plants, changes in environmental conditions (temperature, humidity, etc.), it is necessary to recreate the model. However, in the above-mentioned conventional technique, it is not possible to notify the operator or the like that the model needs to be recreated.

One embodiment of the present invention has been made in view of the above points, and an object of the present invention is to support the re-creation of a model.

In order to achieve the above object, one embodiment of the present invention is a monitoring device for detecting the occurrence of an abnormality in a device or equipment, and is a storage means for storing a model for detecting the occurrence of an abnormality in the device or equipment. And the acquisition means for acquiring the operation data indicating the operation state of the device or equipment, and when the operation data is acquired by the acquisition means, the operation data and the model stored in the storage means Based on the calculation means for calculating a predetermined index value, the determination means for determining whether or not the model needs to be recreated based on the index value calculated by the calculation means, and the determination means. When it is determined that the model needs to be recreated, the model has a notification means for notifying that the model needs to be recreated.

According to one embodiment of the present invention, it is possible to support the reconstruction of the model.

It is a system block diagram of an example of the monitoring system which concerns on 1st Embodiment. It is a hardware block diagram of an example of the monitoring apparatus which concerns on 1st Embodiment. It is a functional block diagram of an example of the monitoring device which concerns on 1st Embodiment. It is a flowchart which shows an example of the abnormality detection processing which concerns on 1st Embodiment. It is a figure which shows an example of the notification screen of remodeling. It is a functional block diagram of an example of the monitoring device which concerns on 2nd Embodiment. It is a flowchart which shows an example of the abnormality detection processing which concerns on the 2nd Embodiment. It is a flowchart which shows another example of the abnormality detection processing which concerns on 2nd Embodiment.

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

[First Embodiment]
<System configuration>
First, the system configuration of the monitoring system 1 according to the present embodiment will be described with reference to FIG. FIG. 1 is a system configuration diagram of an example of the monitoring system 1 according to the first embodiment.

As shown in FIG. 1, in the monitoring system 1 according to the present embodiment, the monitoring device 10 and one or more control devices 20 are connected to each other so as to be able to communicate with each other via a network N such as a LAN (Local Area Network). Has been done. Further, one or more devices 30 are connected to the control device 20. Here, the device 30 is various facilities, plants, etc. controlled by the control device 20, and performs, for example, a manufacturing process of various products such as food, semiconductors, and steel.

The monitoring device 10 is an information processing device that detects an abnormality generated in a manufacturing process of the device 30, and is, for example, a desktop PC, a notebook PC, or the like. The monitoring device 10 may be, for example, a smartphone, a tablet terminal, or the like.

A user of the monitoring device 10 (for example, an operator in a product manufacturing factory) can know that an abnormality has occurred in a manufacturing process or the like by using the monitoring device 10. The monitoring device 10 detects the occurrence of an abnormality, for example, based on a model created in advance from operation data indicating a normal operating state of the device 30.

Further, the monitoring device 10 determines whether or not it is necessary to recreate the model used for detecting the occurrence of an abnormality. Then, when the monitoring device 10 determines that the model needs to be recreated, the monitoring device 10 notifies the user to that effect. As a result, the user of the monitoring device 10 according to the present embodiment can know that it is necessary to recreate the model used for detecting the occurrence of an abnormality.

The control device 20 is a device that controls the device 30, and is, for example, a PLC (Programmable Logic Controller) or the like. The control device 20 acquires operation data indicating the operation state of the device 30 measured by, for example, various sensors, and transmits the operation data to the monitoring device 10.

The operating state of the device 30 is represented by, for example, measuring the temperature, pressure, flow rate, etc. (these are also referred to as "process variables") of predetermined parts of the device 30 with various sensors or the like. To. Therefore, the operation data is vector data represented by one or more process variables.

In addition, various sensors and the like measure the operating state of the device 30 at predetermined time intervals. A predetermined time interval in which various sensors or the like measure the operating state of the device 30 is also referred to as a sampling cycle. The sampling period is, for example, several tens of milliseconds to several tens of seconds.

As described above, the device 30 is various facilities, plants, or the like that are controlled by the control device 20 and perform process processing such as a manufacturing process.

The monitoring system 1 shown in FIG. 1 is an example and may have another configuration. For example, the monitoring system 1 according to the present embodiment may be configured to include a plurality of monitoring devices 10.

<Hardware configuration>
Next, the hardware configuration of the monitoring device 10 according to the present embodiment will be described with reference to FIG. FIG. 2 is a hardware configuration diagram of an example of the monitoring device 10 according to the first embodiment.

As shown in FIG. 2, the monitoring device 10 according to the present embodiment includes an input device 11, a display device 12, an external I / F 13, a RAM (Random Access Memory) 14, and a ROM (Read Only Memory) 15. Has. Further, the monitoring device 10 includes a CPU (Central Processing Unit) 16, a communication I / F 17, and a storage device 18. Each of these hardware is communicably connected by bus B.

The input device 11 is, for example, a keyboard, a mouse, a touch panel, or the like, and is used for the user to input each operation. The display device 12 is, for example, an LCD (Liquid Crystal Display) or the like, and displays the processing result of the monitoring device 10. At least one of the input device 11 and the display device 12 may be used by connecting to the bus B when necessary.

The external I / F 13 is an interface with an external device. The external device includes a recording medium 13a and the like. As a result, the monitoring device 10 can read and write the recording medium 13a via the external I / F 13.

The recording medium 13a includes, for example, a flexible disk, a CD, a DVD, an SD memory card, a USB memory, and the like. The recording medium 13a may store a program that realizes the present embodiment.

The RAM 14 is a non-volatile semiconductor memory capable of holding programs and data even when the power is turned off. The ROM 15 stores programs and data such as BIOS (Basic Input / Output System), OS (Operating System) settings, and network settings that are executed when the monitoring device 10 is started.

The CPU 16 is an arithmetic unit that realizes control and functions of the entire monitoring device 10 by reading programs and data from the ROM 15, the storage device 18, and the like onto the RAM 14 and executing processing.

The communication I / F 17 is an interface for connecting the monitoring device 10 to the network N. As a result, the monitoring device 10 can perform data communication with the control device 20 via the communication I / F17.

The storage device 18 is a non-volatile memory for storing programs and data, such as an HDD (Hard Disk Drive) and an SSD (solid state drive). The programs and data stored in the storage device 18 include a program that realizes the present embodiment, an OS that is basic software that controls the entire monitoring device 10, application software that provides various functions on the OS, and the like. The storage device 18 manages the stored programs and data by a predetermined file system, DB (database), or the like.

By having the above-mentioned hardware configuration, the monitoring device 10 according to the present embodiment can realize various processes as described later.

<Functional configuration>
Next, the functional configuration of the monitoring device 10 according to the present embodiment will be described with reference to FIG. FIG. 3 is a functional configuration diagram of an example of the monitoring device 10 according to the first embodiment.

As shown in FIG. 3, the monitoring device 10 according to the present embodiment includes a data acquisition unit 101, a data correction unit 102, an index value calculation unit 103, a model determination unit 104, a remodeling notification unit 105, and the like. It has an abnormality determination unit 106 and an abnormality notification unit 107. Each of these functional units is realized by a process of causing the CPU 16 to execute one or more programs installed in the monitoring device 10.

Further, the monitoring device 10 according to the present embodiment has a model storage unit 108. The storage unit can be realized by using, for example, a storage device 18. The storage unit may be realized by using a storage device or the like connected to the monitoring device 10 via the network N.

The data acquisition unit 101 acquires the operation data transmitted from the control device 20. The operation data is transmitted from the control device 20 for each sampling cycle, for example.

The data correction unit 102 corrects the operation data acquired by the data acquisition unit 101. Here, the correction of the operation data is, for example, a complementary process of missing a process variable included in the operation data, a time synchronization process of each process variable included in the operation data, and the like.

The index value calculation unit 103 is a predetermined index value for determining whether or not an abnormality has occurred based on the model stored in the model storage unit 108 and the operation data corrected by the data correction unit 102. Is calculated. Here, the predetermined index values are, for example, a Q statistic and a T ² statistic.

The model determination unit 104 determines whether or not it is necessary to recreate the model stored in the model storage unit 108 based on the index value calculated by the index value calculation unit 103. In other words, the model determination unit 104 determines the reliability of the model stored in the model storage unit 108 (that is, the accuracy of abnormality detection based on the model) based on the index value calculated by the index value calculation unit 103. Determine if it is sufficient.

When the model determination unit 104 determines that the model needs to be recreated, the remodeling notification unit 105 notifies the user to recreate the model. Here, the notification prompting the re-creation of the model is, for example, displaying a message indicating that the re-creation of the model is necessary on the display device 12, sending an e-mail containing the message, and the like.

When the model determination unit 104 determines that the model does not need to be recreated, the abnormality determination unit 106 determines whether or not an abnormality has occurred based on the index value calculated by the index value calculation unit 103. That is, the abnormality determination unit 106 detects an abnormality that has occurred in the process process based on the index value calculated by the index value calculation unit 103.

When the abnormality determination unit 106 determines that an abnormality has occurred, the abnormality notification unit 107 notifies that an abnormality has occurred. Here, the notification indicating that an abnormality has occurred is, for example, displaying a message indicating that an abnormality has occurred in the process process on the display device 12, sending an e-mail containing the message, or the like. Further, the abnormality notification unit 107 may notify the occurrence of an abnormality by, for example, a buzzer sound or the like, or may notify the occurrence of an abnormality by turning on a rotating light or the like.

The model storage unit 108 stores a model for detecting an abnormality generated in the process process. Here, the model is, for example, a normal model created from operation data indicating a normal operation state of the device 30 accumulated in advance.

<Details of processing>
Next, the details of the processing of the monitoring device 10 according to the present embodiment will be described. Hereinafter, in the monitoring device 10 according to the present embodiment, a process of detecting an abnormality generated in the process process will be described with reference to FIG. FIG. 4 is a flowchart showing an example of the abnormality detection process according to the first embodiment. The abnormality detection process shown in FIG. 4 is executed, for example, every sampling cycle.

First, the data acquisition unit 101 of the monitoring device 10 acquires the operation data transmitted from the control device 20 (step S401).

Next, the data correction unit 102 of the monitoring device 10 corrects the operation data acquired by the data acquisition unit 101 (step S402). That is, the data correction unit 102 performs, for example, complement processing for missing process variables included in the operation data, time synchronization processing for each process variable included in the operation data, and the like.

Next, the index value calculation unit 103 of the monitoring device 10 calculates a predetermined index value based on the model stored in the model storage unit 108 and the operation data corrected by the data correction unit 102 (step). S403). In the following, the operation data corrected by the data correction unit 102 is also simply referred to as “operation data”.

Here, it is assumed that the index value calculation unit 103 calculates the Q statistic and the T ² statistic as predetermined index values. However, the index value calculation unit 103 may calculate only one of the Q statistic and the T ² statistic. The Q statistic and the T ² statistic can be calculated from the operation data and the loading matrix obtained by performing principal component analysis (PCA) on the model.

Next, the model determination unit 104 of the monitoring device 10 needs to recreate the model stored in the model storage unit 108 based on the Q statistic and the T ² statistic calculated by the index value calculation unit 103. It is determined whether or not there is (step S404).

Here, the Q statistic is represented by the sum of the contributions of one or more Q statistics (this is referred to as "Q contribution"). Similarly, the T ² statistic is represented by the sum of the contributions of one or more T ² statistics (this is referred to as "T ² contribution").

Further, the accuracy of the Q statistic becomes higher as an index value when the variance calculated from the Q contribution of 1 or more (this is referred to as “Q variance”) is lower. Similarly, the accuracy of the T ² statistic becomes higher as an index value when the variance calculated from one or more T ² contributions (this is referred to as “T ² variance”) is lower.

Therefore, the model determination unit 104 determines that the model needs to be recreated, for example, when at least one of the Q variance and the T ² variance is larger than a predetermined threshold value set in advance.

The model determination unit 104 is not limited to this, and for example, when a predetermined number of Q contributions out of one or more Q contributions are larger (or smaller) than a predetermined threshold value, the model determination unit 104 recreates the model. May be determined to be necessary. Similarly, the model determination unit 104 needs to recreate the model when, for example, a predetermined number of T ² contributions among one or more T ² contributions are larger (or smaller) than a predetermined threshold value. It may be determined that.

Further, in the model determination unit 104, for example, among the Q contribution rates calculated by dividing the Q contribution degree by the Q statistic and multiplying by 100, a predetermined number of Q contribution rates are larger than a predetermined threshold value. If (or small), it may be determined that the model needs to be recreated. Similarly, in the model determination unit 104, for example, among the T ² contribution rates calculated by dividing the T ² contribution degree by the T ² statistic and multiplying by 100, a predetermined number of T ² contribution rates are determined. If it is greater than (or less than) a predetermined threshold, it may be determined that the model needs to be recreated.

As described above, in the monitoring system 1 according to the present embodiment, when the reliability of the index value calculated based on the model stored in the model storage unit 108 is low (that is, the accuracy is low), the model is concerned. Is determined to need to be recreated.

When the model determination unit 104 did not determine that the model needs to be recreated in step S404, the abnormality determination unit 106 of the monitoring device 10 generated an abnormality based on the Q statistic and the T ² statistic. Whether or not it is determined (step S405).

That is, the abnormality determination unit 106 determines whether or not at least one of the Q statistic and the T ² statistic is larger than a predetermined threshold value set in advance. Then, the abnormality determination unit 106 determines that an abnormality has occurred when at least one of the Q statistic and the T ² statistic is larger than a predetermined threshold value set in advance. As a result, the occurrence of an abnormality is detected.

When the abnormality determination unit 106 determines in step S405 that an abnormality has occurred, the abnormality notification unit 107 of the monitoring device 10 notifies that an abnormality has occurred in the process process (step S406).

The abnormality notification unit 107 may, for example, display a message indicating the occurrence of an abnormality on the display device 12, or send an e-mail containing the message to a predetermined terminal (for example, a smartphone such as an operator or a tablet terminal). You may send it. Further, the abnormality notification unit 107 may notify the occurrence of an abnormality by, for example, a buzzer sound or the like, or may notify the occurrence of an abnormality by turning on a rotating light or the like. As a result, a user such as an operator can know that an abnormality has occurred in the process process.

On the other hand, if the abnormality determination unit 106 does not determine that an abnormality has occurred in step S405, the monitoring device 10 ends the process.

When the model determination unit 104 determines in step S404 that the model needs to be recreated, the remodeling notification unit 105 of the monitoring device 10 gives a notification prompting the model to be recreated (step S407).

Here, the remodeling notification unit 105 notifies the display device 12 to recreate the model by displaying the remodeling notification screen G100 shown in FIG. 5, for example. The user can recreate the model by pressing the "recreate model" button on the remodeling notification screen G100 shown in FIG. 5, for example.

The remodeling notification unit 105 may, for example, send an e-mail containing a message prompting the re-creation of the model to a predetermined terminal (for example, a smartphone such as an operator, a tablet terminal, or the like).

As described above, in the monitoring system 1 according to the present embodiment, when the reliability of the model is low, the user can be notified that the model needs to be recreated. Therefore, a user such as an operator can perform more accurate abnormality detection by recreating the model stored in the model storage unit 108.

[Second Embodiment]
Next, the second embodiment will be described. In the second embodiment, when the number of times the model determination unit 104 determines that the model needs to be recreated is equal to or greater than a predetermined number of times, the model determination unit 104 notifies that the model needs to be recreated. ..

In the second embodiment, the differences from the first embodiment will be mainly described, and the portion where the same processing as that of the first embodiment and the portion having the same function will be described in the first embodiment. A reference numeral similar to that of the embodiment is assigned, and the description thereof will be omitted.

<Functional configuration>
First, the functional configuration of the monitoring device 10 according to the present embodiment will be described with reference to FIG. FIG. 6 is a functional configuration diagram of an example of the monitoring device 10 according to the second embodiment.

As shown in FIG. 6, the monitoring device 10 according to the present embodiment has a different function of the remodeling notification unit 105A from the first embodiment.

The remodeling notification unit 105A counts the number of times the model determination unit 104 determines that the model needs to be recreated. Then, when the count number exceeds a predetermined threshold value, the remodeling notification unit 105A gives a notification prompting the re-creation of the model.

<Details of processing>
Next, the details of the processing of the monitoring device 10 according to the present embodiment will be described. Hereinafter, in the monitoring device 10 according to the present embodiment, a process of detecting an abnormality generated in the process process will be described with reference to FIG. 7. FIG. 7 is a flowchart showing an example of the abnormality detection process according to the second embodiment.

When the model determination unit 104 determines in step S404 that the model needs to be recreated, the remodeling notification unit 105A of the monitoring device 10 counts the number of times it is determined that the model needs to be recreated. (Step S701). That is, the remodeling notification unit 105A adds "1" to N, for example, by using the variable N initialized to "0" when the monitoring device 10 is started.

Next, the remodeling notification unit 105A of the monitoring device 10 determines whether or not the count number exceeds a predetermined threshold value set in advance (step S702). That is, the remodeling notification unit 105A determines whether or not the value of the variable N is equal to or greater than a preset predetermined threshold value.

When it is determined in step S702 that the count number exceeds a predetermined threshold value, the remodeling notification unit 105A of the monitoring device 10 gives a notification prompting the model to be recreated and initializes the count number (that is, that is). , Initialize the variable N to "0") (step S703).

As described above, in the monitoring system 1 according to the present embodiment, it is necessary to count the number of times that the reliability of the model is determined to be low, and then recreate the model when the count number exceeds the threshold value. It is possible to notify the user that there is. As a result, the monitoring system 1 according to the present embodiment can determine whether or not it is necessary to recreate the model with higher accuracy.

Here, another example of the abnormality detection process according to the present embodiment will be described with reference to FIG. FIG. 8 is a flowchart showing another example of the abnormality detection process according to the second embodiment.

When the model determination unit 104 determines in step S404 that the model does not need to be recreated, the remodeling notification unit 105A of the monitoring device 10 initializes the count number (that is, initializes the variable N to "0". (Step S801).

As a result, in the monitoring system 1 according to the present embodiment, it is necessary to continuously count the number of times that the reliability of the model is determined to be low, and then recreate the model when the count number exceeds the threshold value. It is possible to notify the user that this is the case. As a result, the monitoring system 1 according to the present embodiment can determine whether or not it is necessary to recreate the model with higher accuracy.

The present invention is not limited to each of the above-described embodiments specifically disclosed, and various modifications and modifications can be made without departing from the scope of claims.

1 Monitoring system 10 Monitoring device 20 Control device 30 Equipment 101 Data acquisition unit 102 Data correction unit 103 Index value calculation unit 104 Model judgment unit 105 Remodeling notification unit 106 Abnormality judgment unit 107 Abnormality notification unit 108 Model storage unit

Claims (7)

A monitoring device that detects the occurrence of abnormalities in equipment or equipment.
A storage means for storing a model for detecting the occurrence of an abnormality in the device or equipment,
An acquisition means for acquiring operation data indicating the operation state of the device or equipment, and
When the operation data is acquired by the acquisition means, a calculation means for calculating the Q statistic and the T ² statistic based on the operation data and the model stored in the storage means, and
A determination means for determining whether or not the model needs to be recreated based on the Q statistic and the T ² statistic calculated by the calculation means.
When it is determined by the determination means that the model needs to be recreated, the notification means for notifying that the model needs to be recreated and the notification means.
Have a,
The determination means
When either the variance of the contribution of the Q statistic calculated by the calculation means or the variance of the contribution of the T ² statistic calculated by the calculation means is equal to or greater than a predetermined threshold value. A monitoring device that determines that the model needs to be recreated . A monitoring device that detects the occurrence of abnormalities in equipment or equipment.
A storage means for storing a model for detecting the occurrence of an abnormality in the device or equipment,
An acquisition means for acquiring operation data indicating the operation state of the device or equipment, and
When the operation data is acquired by the acquisition means, the Q statistic and T are based on the operation data and the model stored in the storage means. ² Calculation means for calculating statistics and
The Q statistic and T calculated by the calculation means. ² A determination means for determining whether or not the model needs to be recreated based on the statistics.
When it is determined by the determination means that the model needs to be recreated, the notification means for notifying that the model needs to be recreated and the notification means.
Have,
The determination means
Among the contributions of the Q statistic calculated by the calculation means, when a predetermined number of contributions are equal to or more than a predetermined threshold value, or the T calculated by the calculation means. ² A monitoring device that determines that the model needs to be recreated when a predetermined number of contributions of statistics are equal to or greater than a threshold value. The notification means is
The monitoring according to claim 1 or 2 , wherein when the number of times the determination means determines that the model needs to be recreated is equal to or greater than a predetermined threshold value, the determination means notifies that the model needs to be recreated. apparatus. The notification means is
The monitoring device according to claim 3 , wherein when the determination means does not determine that the model needs to be recreated, the number of times is initialized. The notification means is
The monitoring device according to any one of claims 1 to 4 , which indicates on a display that the model needs to be recreated, or notifies by sending an e-mail. In a monitoring method used for a monitoring device that detects the occurrence of an abnormality in a device or equipment and has a storage means for storing a model for detecting the occurrence of an abnormality in the device or equipment.
An acquisition procedure for acquiring operation data indicating the operation status of the equipment or equipment, and
When the operation data is acquired by the acquisition procedure, a calculation procedure for calculating the Q statistic and the T ² statistic based on the operation data and the model stored in the storage means, and
A determination procedure for determining whether or not the model needs to be recreated based on the Q statistic and the T ² statistic calculated by the calculation procedure.
When it is determined by the determination procedure that the model needs to be recreated, the notification procedure for notifying that the model needs to be recreated and the notification procedure.
Have a,
The determination procedure is
When either the variance of the contribution of the Q statistic calculated by the calculation procedure or the variance of the contribution of the T ² statistic calculated by the calculation procedure is equal to or greater than a predetermined threshold value. A monitoring method that determines that the model needs to be recreated . In a monitoring method used for a monitoring device that detects the occurrence of an abnormality in a device or equipment and has a storage means for storing a model for detecting the occurrence of an abnormality in the device or equipment.
An acquisition procedure for acquiring operation data indicating the operation status of the equipment or equipment, and
When the operation data is acquired by the acquisition procedure, the Q statistic and T are based on the operation data and the model stored in the storage means. ² Calculation procedure to calculate statistics and
The Q statistic and T calculated by the calculation procedure. ² A determination procedure for determining whether or not the model needs to be recreated based on the statistics, and
When it is determined by the determination procedure that the model needs to be recreated, the notification procedure for notifying that the model needs to be recreated and the notification procedure.
Have,
The determination procedure is
Of the contributions of the Q statistic calculated by the calculation procedure, when a predetermined number of contributions are equal to or more than a predetermined threshold value, or the T calculated by the calculation procedure. ² A monitoring method in which it is determined that the model needs to be recreated when a predetermined number of contributions of statistics are equal to or greater than a threshold value.

Images