Monitoring control device and monitoring control system

Abstract

Provided is a monitoring control device capable of improving the accuracy of maintenance of a monitored object, the monitoring control device being provided with: a correlation definition unit that defines a correlation between monitoring items among a plurality of monitoring items; a correlation model management DB that manages a plurality of correlation models using the plurality of monitoring items associated with each other as the correlation models; a monitoring data management DB that manages monitoring data for a plurality of monitoring items; a causal relationship calculation unit that, when an occurrence of an abnormality is detected in at least one of the plurality of monitoring items, calculates which monitoring item of the plurality of association models has a causal relationship with the occurrence of the abnormality on the basis of the plurality of association models managed by the association model management DB and the monitoring data managed by the monitoring data management database; and a cause specifying unit that specifies the cause of the occurrence of the abnormality on the basis of the monitoring items having the causal relationship.

Description

Technical Field

[0001] This disclosure relates to monitoring and control devices, and particularly to monitoring and control devices for monitoring factories such as water treatment plants. Background Technology

[0002] In recent years, monitoring systems for various factories, including water treatment plants, have been developed to infer the causal relationship of anomalies and provide information to factory operators in the event of anomalies.

[0003] For example, Patent Document 1 discloses a technique for defining causal relationship candidates by defining the correlation between tags of field machines and other field machines, such as the upstream and downstream positional relationship of piping configured in the factory, and constructing a causal relationship model using Bayesian networks based on process data.

[0004] The constructed causal relationship model is transformed into a form equivalent to a Quality Management Matrix (QMM) using the analysis department and then presented.

[0005] Existing technical documents

[0006] Patent documents

[0007] Patent Document 1: Japanese Patent Application Publication No. 2022-115643 Summary of the Invention

[0008] In Patent Document 1, the causal relationship is determined by defining the correlation between the tags of the field machine and other field machines, such as the upstream and downstream positional relationship of the piping. This has the problem that the causal relationship that can be determined is limited.

[0009] This disclosure was made to solve the problems described above, and its purpose is to provide a monitoring and control device that is not limited by the causal relationship between the monitored items and can improve the maintenance accuracy of the monitored objects.

[0010] This disclosure provides a monitoring and control device for monitoring multiple monitoring items. The monitoring and control device comprises: an association definition unit for defining associations between the multiple monitoring items; an association model management database for managing multiple association models, including the multiple monitoring items with associations defined by the association definition unit; a monitoring data management database for managing monitoring data of the multiple monitoring items; a causality calculation unit for calculating, when an anomaly is detected in at least one of the multiple monitoring items, which monitoring item in the multiple association models managed by the association model management database and the monitoring data managed by the monitoring data management database has a causal relationship with the anomaly; and a cause determination unit for determining the cause of the anomaly based on the monitoring item with the causal relationship calculated by the causality calculation unit, wherein the causality calculation unit calculates the monitoring item with the causal relationship to the anomaly based on the time-series changes in the monitoring data before and after the anomaly.

[0011] According to the monitoring and control device disclosed herein, based on multiple correlation models and monitoring data, it calculates which monitoring item in the multiple correlation models has a causal relationship with the occurrence of an anomaly, and determines the cause of the anomaly based on the monitoring item with the causal relationship. Therefore, the causal relationship between the identifiable monitoring items is not limited, which can improve the accuracy of the maintenance of the monitored objects. Attached Figure Description

[0012] Figure 1 This is a block diagram illustrating the structure of the monitoring and control device involved in this disclosure.

[0013] Figure 2 This is a diagram illustrating an example of a monitoring item information management DB of the monitoring and control device involved in this disclosure.

[0014] Figure 3 This is a diagram illustrating an example of a monitoring data management DB of the monitoring and control device involved in this disclosure.

[0015] Figure 4 This is a diagram illustrating an example of an association model between monitoring items defined in the association definition section of the monitoring and control device involved in this disclosure.

[0016] Figure 5 This is a schematic diagram illustrating a situation where an anomaly is detected in the monitoring and control device involved in this disclosure.

[0017] Figure 6 This is a graph showing the current and previous values ​​of the sensor output values ​​in the monitoring data management DB of the monitoring and control device involved in this disclosure.

[0018] Figure 7 This is a diagram illustrating an example of calculating the causal relationship between an anomaly and the occurrence of an anomaly in the monitoring and control device of Embodiment 1 of this disclosure.

[0019] Figure 8 This is a graph showing the time-series changes of monitoring data in the monitoring data management DB of the monitoring and control device according to Embodiment 1 of this disclosure before and after an anomaly occurs.

[0020] Figure 9 This diagram illustrates the determination of the cause of an anomaly in the cause determination unit of the monitoring and control device according to Embodiment 1 of this disclosure.

[0021] Figure 10 This is a graph showing the time-series changes of monitoring data in the monitoring data management DB of the monitoring and control device according to Embodiment 1 of this disclosure before and after an anomaly occurs.

[0022] Figure 11 This is a diagram showing an example of a display in the display section of the monitoring and control device according to Embodiment 1 of this disclosure.

[0023] Figure 12 This is a diagram illustrating an example of causal relationship calculation in the causal relationship calculation unit of the monitoring and control device according to Embodiment 2 of this disclosure.

[0024] Figure 13 This is a graph showing the time-series changes of monitoring data in the monitoring data management DB of the monitoring and control device according to Embodiment 2 of this disclosure before and after an anomaly occurs.

[0025] Figure 14 This is a diagram illustrating an example of causal relationship calculation in the causal relationship calculation unit of the monitoring and control device according to Embodiment 3 of this disclosure.

[0026] Figure 15 This is a graph showing the time-series changes of monitoring data in the monitoring data management DB of the monitoring and control device according to Embodiment 3 of this disclosure before and after an anomaly occurs.

[0027] Figure 16 This is a diagram illustrating an example of causal relationship calculation in the causal relationship calculation unit of the monitoring and control device according to Embodiment 4 of this disclosure.

[0028] Figure 17 This is a graph showing the time-series changes of monitoring data in the monitoring data management DB of the monitoring and control device according to Embodiment 4 of this disclosure before and after an anomaly occurs.

[0029] Figure 18This is a diagram illustrating an example of causal relationship calculation in the causal relationship calculation unit of the monitoring and control device according to Embodiment 5 of this disclosure.

[0030] Figure 19 This is a graph showing the time-series changes of monitoring data in the monitoring data management DB of the monitoring and control device according to Embodiment 5 of this disclosure before and after an anomaly occurs.

[0031] Figure 20 This is a diagram illustrating an example of causal relationship calculation in the causal relationship calculation unit of the monitoring and control device according to Embodiment 6 of this disclosure.

[0032] Figure 21 This is a graph showing the time-series changes of monitoring data in the monitoring data management DB of the monitoring and control device according to Embodiment 6 of this disclosure before and after an anomaly occurs.

[0033] Figure 22 This is a block diagram illustrating the structure of a modified example of the monitoring and control device involved in this disclosure.

[0034] Figure 23 This is a diagram showing the hardware structure of the monitoring and control device that implements embodiments 1 to 6 of this disclosure.

[0035] Figure 24 This is a diagram showing the hardware structure of the monitoring and control device that implements embodiments 1 to 6 of this disclosure.

[0036] Figure 25 This is a block diagram illustrating the structure of installing the monitoring and control device of embodiments 1 to 6 of this disclosure into a server computer constituting a cloud environment.

[0037] Figure 26 This is a block diagram illustrating the structure of the monitoring and control system according to Embodiment 7 of this disclosure. Detailed Implementation

[0038] <Implementation Method 1>

[0039] Figure 1 This is a block diagram illustrating the structure of the monitoring and control device 100 according to Embodiment 1 of this disclosure. Figure 1 As shown, the monitoring and control device 100 includes a monitoring item information management database (DB) 10, a monitoring item inter-relationship definition unit 20, an association model management database (DB) 30, a monitoring data management database (DB) 40, a causal relationship calculation unit 50, a cause determination unit 60, and a display unit 70.

[0040] DB10 manages monitoring project information, including equipment, machine, and sensor information, using monitoring data. Figure 2 This is a diagram showing an example of DB10's monitoring project information management in tabular form. Figure 2 In the monitoring section, the names of the plant, facility, equipment, machine, and sensor are shown. As a plant, the purification center is shown. As a facility, the water treatment facility is shown. As equipment, the primary sedimentation equipment and reaction tank equipment are shown. As machines, machines A, B, and C are shown. As sensor names, sensors A to E are shown.

[0041] The Inter-monitoring Item Correlation Definition Section 20 defines the correlations between monitoring items. These correlations can refer to physical correlations such as equipment structure correlations (e.g., equipment structures, machinery, and sensors), wiring correlations (e.g., power wiring in power systems), or process correlations (e.g., in water purification plants and wastewater treatment plants). They can also refer to correlations based on the experience values ​​of skilled monitoring personnel or correlations derived through statistical analysis. Furthermore, they include system correlations between systems such as monitoring control systems and equipment management systems. Moreover, they include correlations between information from different sensors and data sources for the same object, such as facilities, machinery, or roads, bridges, and tunnels, including inspection information, point group information, water level sensor information, vibration sensor information, camera image information, and weather information. This expands the scope of correlations between monitoring items. Furthermore, defining correlations between monitoring items means linking monitoring items together.

[0042] The relational model management DB30 manages the interrelated monitoring items defined by the monitoring item relationship definition section 20 as relational models. In this management, a graph database (DB) is used, and the relational model is managed as a tree structure. More specifically, a labeled attribute graph is used for management. Alternatively, an RDB (Relational Database) can also be used. Specific examples of the relational model management DB30 will be explained later.

[0043] The monitoring data management DB40 manages the monitoring data output by monitoring machines such as sensors according to time series, and inputs the monitoring data together with the correlation model managed by the correlation model management DB30 into the causal relationship calculation unit 50. Figure 3 This is a diagram showing an example of DB40 monitoring data management in tabular form. Figure 3 The table shows the current and previous values ​​of the output values ​​of sensors A through E, respectively.

[0044] When multiple sensors SC installed throughout the factory detect an anomaly, the causal relationship calculation unit 50 calculates which monitoring item in which correlation model is causally related to the anomaly based on the correlation model and the time series changes of monitoring data before and after the anomaly.

[0045] The cause determination unit 60 determines the monitoring items that are the cause of the abnormality based on the monitoring items that have a causal relationship calculated by the causal relationship calculation unit 50.

[0046] Display unit 70 is a display that shows the monitoring items that are causally related to the occurrence of the abnormality, as determined by cause determination unit 60, and the cause of the abnormality.

[0047] Next, the operation of each part of the monitoring and control device 100 will be explained using the association model management DB30 with labeled attribute graphs.

[0048] Figure 4 This illustrates an example of the association model between monitoring items defined in the inter-monitoring item association definition section 20, showing that according to... Figure 2 The example shown is a relational model defined in DB10 for monitoring project information management.

[0049] exist Figure 4 In the diagram, multiple monitoring items are arranged in a tree diagram. The water treatment facilities in the factory, which serves as the purification center, include primary sedimentation equipment and reaction tank equipment. The primary sedimentation equipment includes machine A and machine B, and the reaction tank equipment includes machine C. Machine A has sensor A, machine B has sensors B and C, and machine C has sensors D and E.

[0050] Each monitored item is enclosed in a box. The association model representing device relationships is defined by connecting monitored items with solid lines. The association model representing a water treatment process is defined by arrows with narrow dashed lines running from machine B towards machine C. The association model representing related relationships is defined by bidirectional arrows with a single-dotted dashed line between machine A and machine C. The association model representing electrical wiring within the same cabling is defined by arrows with wide dashed lines running from machine A towards machine B.

[0051] use Figure 5 This illustrates the relationship between the factory's sensors (SC) in the monitoring project association model defined in this way. Figure 1 This is an example of a situation where an anomaly has been detected. In Figure 5 The image illustrates, schematically, a situation where sensor D detects an anomaly. Figure 6 This displays the current and previous values ​​of the output value of sensor D in the monitoring data management DB40. For example... Figure 6 As shown, the current value of the output of sensor D is 20.0, and the previous value was 18.5.

[0052] An anomaly is detected when the current value of the sensor output differs significantly from the previous value, when the current value exceeds a predetermined threshold, or when the value is determined to be an outlier based on previous statistical results. Figure 6In the example, the difference between the current value and the previous value is 1.5, which is large, thus indicating that an exception has occurred.

[0053] Next, use Figure 7 This illustrates an example of causality calculation in the causality calculation unit 50. Figure 5 The diagram illustrates a situation where sensor D detects an anomaly, but... Figure 7 This schematically illustrates the possibility that an anomaly may also occur in sensor E.

[0054] Sensors associated with the sensor that detected the anomaly—that is, sensors causally related to the anomaly—are similarly likely to exhibit time-series changes before and after the anomaly occurred. Figure 8 In this process, the time-series changes in the monitoring data before and after the anomaly are calculated based on the current and previous values ​​of the output values ​​of sensors A~E in the monitoring data management DB40, and are represented as the change amount. Figure 8 In the data, the change in the output value of sensor D from its previous value is 1.5. The previous value of the output value of sensor E was 8.2, and the current value is 10.0, so the change from its previous value is 1.8, which is greater than the change of sensor D. Therefore, it is determined that sensor E and sensor D are also causally related to the occurrence of the anomaly. In addition, the changes in the previous values ​​of sensors A to C are all less than 0.5.

[0055] In this way, by calculating the change in each sensor's value from the previous value, sensors with time-series changes similar to those of the sensor that detected the anomaly can be identified as sensors causally related to the anomaly. Furthermore, causality can be calculated by using a statistical method employing Euclidean distance to calculate the time-series changes in monitoring data. When using Euclidean distance, the sensor that detected the anomaly and the sensor with the closest distance between the two points are identified as sensors causally related to the anomaly.

[0056] Next, use Figure 9 This explains the determination of the cause of the abnormality in section 60. Figure 9 This example illustrates how, in cases where there is also a possibility of an anomaly occurring in sensor E, sensor E is identified as a sensor causally related to the occurrence of the anomaly, just as sensor D, which detected the anomaly, is used as a sensor.

[0057] exist Figure 10 In this process, the time-series changes in the monitoring data before and after the anomaly are calculated based on the current and previous values ​​of the output values ​​of sensors A~E in the monitoring data management DB40, and are represented as the change amount. Figure 10In the above, the change in the output value of sensor D from the previous value is 1.5, and the change in the output value of sensor E from the previous value is 1.8, both of which are above 0.5.

[0058] When determining which sensors are causally related to the detected anomaly by the magnitude of their changes, this can be done by checking if a threshold is exceeded. For example, if the threshold for the change is set to 0.5, and the change in sensor E exceeds the threshold of 0.5, then, based on the association model representing device association, it is determined that sensors D and E are associated with machine C, and therefore, it is determined that sensors E and D are also causally related to the anomaly.

[0059] Furthermore, sensors D and E are connected to machine C via a common solid line. The anomaly detected based on sensor D is identified as originating in machine C. Here, the solid line connecting sensors D and E to machine C is referred to as an edge in the labeled attribute graph, and names can be defined for these edges. Figure 9 In the diagram, edge 1 is labeled as the name of the edge, and reason flag 1 is labeled as the property. Here, edge 1 is the name that defines "device association".

[0060] Thus, in the cause determination unit 60, by using sensor D to detect the occurrence of an anomaly, it is determined that sensor D and sensor E have a causal relationship with the occurrence of the anomaly, thereby determining that machine C is the cause of the anomaly. The result is displayed by the display unit 70 and provided to the user of the monitoring and control device 100.

[0061] Figure 11 An example of a display in display unit 70 is shown. Figure 11 In the diagram, sensor D emits light when it detects an anomaly. Sensors D and E are connected to machine C by a solid line, indicating a causal relationship between the anomaly and the anomaly. An arrow pointing from sensors D and E toward machine C indicates that machine C is the cause of the anomaly.

[0062] As explained above, in the monitoring and control device 100 of Embodiment 1, by defining the correlation between monitoring items such as machines and sensors that are associated with equipment, which are units finer than the components of the equipment, the causal relationship between the monitoring items and the occurrence of abnormalities can be determined, thereby identifying the machine that is the cause of the abnormality. Therefore, compared to the method of replacing all associated machines without determining the cause of the abnormality, the cost of maintenance in the factory can be reduced.

[0063] Furthermore, by defining system relationships between systems such as monitoring and control systems and equipment management systems, it is possible to perform monitoring and control between systems and to perform overall monitoring and control of the factory.

[0064] Furthermore, by calculating the change in each sensor from the previous value, sensors that exhibit time-series changes similar to those of the sensors that detected the anomaly are identified as having a causal relationship with the anomaly. This improves the accuracy of identifying the machine that caused the anomaly.

[0065] <Implementation Method 2>

[0066] Next, the monitoring and control device 100A according to Embodiment 2 of this disclosure will be described. Furthermore, a block diagram showing the structure of the monitoring and control device 100A will be provided. Figure 1 The monitoring and control device 100 of Embodiment 1 is the same. In the monitoring and control device 100A of Embodiment 2, the processing in the causal relationship calculation unit 50 and the cause determination unit 60 is different from that of the monitoring and control device 100.

[0067] use Figure 12 This illustrates an example of causality calculation in the causality calculation unit 50 of the monitoring and control device 100A. Figure 12 The image illustrates, schematically, a situation where sensor D detects an anomaly. Figure 13 In the process, the time series change of the monitoring data before and after the anomaly is calculated based on the current value and the previous value of the output values ​​of sensors A to E in the monitoring data management DB40 and is represented as the change amount, and the weights attached to sensors A to E are also represented.

[0068] That is, if a correlation model with a high causal relationship to the occurrence of an anomaly is known in advance in the correlation model defined in the correlation definition unit 20 between monitoring items, the causal relationship calculation unit 50 performs weighting on that correlation model. Here, the monitoring items of machine A, etc., are nodes in the labeled attribute graph, and the names of sensors with high causal relationships can be defined in the nodes. Figure 12 In this example, for machine A, a sensor with a causal relationship to machine A is defined as sensor A, and labeled as an attribute with a weight of 1.0. Similarly, for machines B and C, the names of sensors with high causal relationships are defined, and the weights of each sensor are labeled. Here, the weight is, for example, set as the value multiplied by the current value; in the case of a weight of 2.0, the current value becomes 2.0 times.

[0069] exist Figure 13 The diagram shows the weights assigned to sensors A through E, while the weights for sensors B and C are both set to 2.0. Figure 13 In the above, the change in the output value of sensor D from the previous value is 1.5, and the change in the output value of sensor E from the previous value is 1.8, both of which are above 0.5.

[0070] When the threshold for the amount of change is set to 0.5, the changes in sensors D and E exceed the threshold of 0.5. Here, based on the association model representing device association, it is determined that sensors D and E are sensors associated with machine C, therefore, it is determined that sensors D and E have a causal relationship with the occurrence of the anomaly.

[0071] In addition, since it is determined in advance that the causal relationship between machine B and sensors B and C is high, the weights of both are set to 2.0. Therefore, the current values ​​are 2.0 times each. The change in the output value from the previous value exceeds the threshold of 0.5, thus confirming that sensors B and C are causally related to the occurrence of the anomaly.

[0072] Furthermore, according to the correlation model representing the water treatment process, machine B and machine C are correlation models of the water treatment process. It is determined that the cause of the anomaly lies in machine B. Therefore, it can be determined that the cause of the anomaly is machine B and machine C.

[0073] As explained above, in the monitoring and control device 100A of Embodiment 2, not only is an association model defined by the physical connection between equipment and machines used to determine the machine that is the cause of the anomaly, but also an association model defined by the process of treatment, such as a water treatment process. Therefore, the cause of the anomaly can also be determined for machines that are indirectly related. Thus, the cause of the anomaly can be determined over a wider range, and the maintenance accuracy of the plant can be improved.

[0074] Furthermore, by weighting the association model with a high causal relationship to the occurrence of the anomaly in advance, it is possible to reliably determine the association model with a high causal relationship to the occurrence of the anomaly.

[0075] <Implementation Method 3>

[0076] Next, the monitoring and control device 100B according to Embodiment 3 of this disclosure will be described. Furthermore, a block diagram showing the structure of the monitoring and control device 100B will be provided. Figure 1 The monitoring and control device 100 of Embodiment 1 is the same. In the monitoring and control device 100B of Embodiment 3, the processing in the causal relationship calculation unit 50 and the cause determination unit 60 is different from that of the monitoring and control device 100.

[0077] use Figure 14 This illustrates an example of causality calculation in the causality calculation unit 50 of the monitoring and control device 100B. Figure 14 The image illustrates, schematically, a situation where sensor D detects an anomaly. Figure 15 In this process, the time-series changes in the monitoring data before and after the anomaly are calculated based on the current and previous values ​​of the output values ​​of sensors A~E in the monitoring data management DB40, and are represented as the change amount. Figure 15 In the diagram, the change in the output value of sensor D from its previous value is 1.5. The previous output value of sensor A was 11.0, and the current value is 11.5, so the change from the previous value is 0.5. Furthermore, the changes in the output values ​​of sensors B, C, and E from their previous values ​​are 0.

[0078] When the threshold for the amount of change is set to 0.5, the changes in both sensor A and sensor D exceed the threshold of 0.5. Here, based on the association model representing device association, it is determined that sensor A is associated with machine A, and sensor D is associated with machine C. Since the change in sensor A exceeds the threshold of 0.5, it is determined that sensor A has a causal relationship with the occurrence of the anomaly. Similarly, the change in sensor D exceeds the threshold of 0.5, therefore, it is determined that sensor D has a causal relationship with the occurrence of the anomaly.

[0079] Furthermore, based on the association model representing the relevant relationship, it is determined that machine C and machine A are related, so it can be determined that the cause of the anomaly detected by sensor D is machine A.

[0080] As explained above, in the monitoring and control device 100B of Embodiment 3, not only is a device association model defined by the physical connection between devices and machines used, but also an association model representing related associations is used to determine the machine that is the cause of the anomaly. Therefore, it is possible to determine the cause of the anomaly by considering the correlation based on the experience value of skilled monitoring personnel and the correlation derived through statistical analysis. Thus, it is possible to determine the cause of the anomaly that cannot be determined using other association models.

[0081] <Implementation Method 4>

[0082] Next, the monitoring and control device 100C according to Embodiment 4 of this disclosure will be described. Furthermore, a block diagram showing the structure of the monitoring and control device 100C will be provided. Figure 1 The monitoring and control device 100 shown in Embodiment 1 is the same. In the monitoring and control device 100C of Embodiment 4, the processing in the causal relationship calculation unit 50 and the cause determination unit 60 is different from that in the monitoring and control device 100.

[0083] use Figure 16 This illustrates an example of causality calculation in the causality calculation unit 50 of the monitoring and control device 100C. Figure 16 The image illustrates, schematically, a situation where sensor D detects an anomaly. Figure 17 In this process, the time-series changes in the monitoring data before and after the anomaly are calculated based on the current and previous values ​​of the output values ​​of sensors A~E in the monitoring data management DB40, and are represented as the change amount. Figure 17In the diagram, the change in the output value of sensor D from its previous value is 1.5. The previous output value of sensor A was 11.0, and the current value is 11.5, so the change is 0.5. The previous output value of sensor B was 3.0, and the current value is 3.6, so the change is 0.6. The previous output value of sensor C was 5.0, and the current value is 5.9, so the change is 0.9. Furthermore, the change in the output value of sensor E from its previous value is 0.

[0084] When the threshold for the amount of change is set to 0.5, the changes in sensors A, B, C, and D all exceed the threshold of 0.5, thus establishing a causal relationship between sensors A through D and the anomaly. Here, based on the association model representing device association, it is determined that sensor D is associated with machine C. Furthermore, based on the association model representing correlation, it is determined that machine C is correlated with machine A. Additionally, based on the association model representing power wiring, it is determined that machine A is associated with machine B. Therefore, it can be determined that the cause of the anomaly detected by sensor D is machine A.

[0085] As explained above, in the monitoring and control device 100C of Embodiment 4, it is possible to combine an association model defining device associations based on physical connections between devices and machines, an association model representing related relationships, and an association model representing power wiring to determine the machine that is the cause of the anomaly. Therefore, the accuracy of determining the cause of the anomaly can be improved, and the accuracy of factory maintenance can be improved.

[0086] <Implementation Method 5>

[0087] Next, the monitoring and control device 100D according to Embodiment 5 of this disclosure will be described. Furthermore, a block diagram showing the structure of the monitoring and control device 100D will be provided. Figure 1 The monitoring and control device 100 of Embodiment 1 is the same. In the monitoring and control device 100D of Embodiment 5, the processing in the causal relationship calculation unit 50 and the cause determination unit 60 is different from that of the monitoring and control device 100.

[0088] use Figure 18 This illustrates an example of causality calculation in the causality calculation unit 50 of the monitoring and control device 100D. Figure 18 The image illustrates, schematically, a situation where sensor D detects an anomaly. Figure 19 In this process, the time-series change of the monitoring data, calculated based on the current, previous, and previous-eighths output values ​​of sensors A through E in the monitoring data management DB40, is represented as the change quantity. Figure 19In this context, the change is a value calculated at specific times when a time-series change occurs. The previous value of sensor C's output was 5.9, the value two years prior was 5.0, and the current value is 5.9, indicating a time-series change between the previous and two years prior values. On the other hand, the previous value of sensor D's output was 18.5, the value two years prior was 18.5, and the current value is 20.0, indicating a time-series change between the current and previous values.

[0089] Based on the results, a causal relationship was established between sensors C and D and the occurrence of the anomaly. Furthermore, it was determined that a time-series change occurred in machine B prior to the time the anomaly occurred in machine C. Here, based on the association model representing device association, sensor D was identified as a sensor associated with machine C. Moreover, based on the association model representing the water treatment process, machines B and C form an association model for the water treatment process. Since the time-series change in the monitoring data of sensor C occurred before the time-series change in the monitoring data of sensor D, it can be determined that the cause of the anomaly detected by sensor D is machine B.

[0090] As explained above, in the monitoring and control device 100D of Embodiment 5, an association model of device association defined by the physical connection between equipment and machines, and an association model of the water treatment process, can be used. Furthermore, the timing of time-series changes in the generated monitoring data is considered to determine the machine that is causing the anomaly. Therefore, the accuracy of determining the cause of the anomaly can be improved, and the maintenance accuracy of the plant can be improved.

[0091] <Implementation Method 6>

[0092] Next, the monitoring and control device 100E according to Embodiment 6 of this disclosure will be described. Furthermore, a block diagram showing the structure of the monitoring and control device 100E will be provided. Figure 1 The monitoring and control device 100 of Embodiment 1 is the same. In the monitoring and control device 100E of Embodiment 6, the processing in the causal relationship calculation unit 50 and the cause determination unit 60 is different from that of the monitoring and control device 100.

[0093] use Figure 20 This illustrates an example of causality calculation in the causality calculation unit 50 of the monitoring and control device 100E. Figure 20 The image illustrates, schematically, a situation where sensor D detects an anomaly. Figure 21 In this context, the time-series change of the monitoring data, calculated based on the current, previous, previous-previous, and previous-previous values ​​of the output values ​​of sensors A through E in the monitoring data management DB40, is represented as the change quantity. Figure 21In the diagram, the change is a value calculated at specific times when a time-series change occurs. For sensor A, the current, previous, and previous-last values ​​are all 11.5, and the previous-last value is 11.0, indicating a time-series change between the two previous values. Similarly, for sensor C, the current and previous values ​​are 5.9, and the previous and previous-last values ​​are 5.0, indicating a time-series change between the two previous values. On the other hand, for sensor D, the previous, previous-last, and previous-last values ​​are 18.5, and the current value is 20.0, indicating a time-series change between the current and previous values.

[0094] Based on the results, it was determined that sensors C and D have a causal relationship with the occurrence of the anomaly. Furthermore, it was determined that there was a time-series change for machine B before the time the anomaly occurred in machine C, and a time-series change for machine A before that. Here, based on the association model representing device association, it was determined that sensor D is a sensor associated with machine C. Moreover, based on the association model representing the water treatment process, machines B and C are associated with the water treatment process; based on the association model representing electrical wiring, it was determined that machine A is a machine associated with machine B. Furthermore, the time-series change in the monitoring data of sensor C occurred before the time-series change in the monitoring data of sensor D, and the time-series change in the monitoring data of sensor A occurred before the time-series change in the monitoring data of sensor C. Therefore, it can be determined that the cause of the anomaly detected by sensor D is machine B and machine A.

[0095] As explained above, in the monitoring and control device 100E of Embodiment 6, it is possible to combine an association model of device association defined by the physical connection between devices and machines, an association model of water treatment process, and an association model representing power wiring, and also consider the timing of time series changes in the generated monitoring data to determine the machine that is the cause of the anomaly. Therefore, the accuracy of determining the cause of the anomaly can be improved, and the accuracy of factory maintenance can be improved.

[0096] <Variation Example>

[0097] In the monitoring and control devices 100-100E of embodiments 1 to 6 described above, the correlation between monitoring items defined by the monitoring item correlation definition unit 20 can include correlations based on the experience values ​​of skilled personnel monitoring and correlations derived through statistical analysis. However, when defining these, the experience values ​​and statistical values ​​of skilled personnel monitoring can be used as input information, and the correlation can be calculated using machine learning with artificial intelligence (AI), and the calculation results can be input into the monitoring item correlation definition unit 20.

[0098] exist Figure 22The image shows a monitoring control device 100~100E that inputs information about the correlation between monitored items calculated by AI200 into the monitoring item correlation definition unit 20.

[0099] By adopting such a structure, the correlation between surveillance items can be defined based on fuzzy information such as the experience value of skilled surveillance personnel, thus simplifying the processing in the correlation definition unit 20 between surveillance items.

[0100] <Hardware Structure>

[0101] Each component of the monitoring and control devices 100-100E described in embodiments 1-6 above can be configured using a computer and implemented by executing a program on the computer. That is, the monitoring and control devices 100-100E, for example, can be configured using a computer. Figure 23 The processing circuit 1000 shown is implemented in this way. In the processing circuit 1000, processors such as CPU (Central Processing Unit) and DSP (Digital Signal Processor) are used to execute programs stored in the storage device to realize the functions of each part.

[0102] Furthermore, dedicated hardware can also be applied to the processing circuit 1000. When the processing circuit 1000 is dedicated hardware, the processing circuit 1000 may correspond to, for example, a single circuit, a composite circuit, a programmable processor, a parallel programmable processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array), or a combination thereof.

[0103] The monitoring and control device 100~100E can implement the functions of each component with independent processing circuits, or it can integrate these functions and implement them with a single processing circuit.

[0104] In addition, Figure 24 The diagram illustrates the hardware structure of a processing circuit 1000 constructed using a processor. In this case, the functions of each part of the monitoring and control devices 100-100E are implemented through a combination with software (software, firmware, or both). The software is described as a program and stored in memory 1002. The processor 1001, which functions as the processing circuit 1000, implements the functions of each part by reading and executing the program stored in memory 1002 (storage device). In other words, it can be described as the process and method by which the program causes the computer to perform the actions of the constituent elements of the monitoring and control devices 100-100E.

[0105] Here, the memory 1002 may be, for example, non-volatile or volatile semiconductor memory such as RAM, ROM, flash memory, EPROM (Erasable Programmable Read Only Memory), EEPROM (Electrically Erasable Programmable Read Only Memory), HDD (Hard Disk Drive), magnetic disk, floppy disk, optical disk, compact disk, mini disk, DVD (Digital Versatile Disc) and its drive device, or any storage medium used in the future.

[0106] The above describes the structure for implementing the functions of each component of the monitoring and control device 100-100E using either hardware or software. However, it is not limited to this; it is also possible to implement some components of the monitoring and control device 100-100E using dedicated hardware and others using software. For example, some components can be implemented using a processing circuit 1000 as dedicated hardware, while others can be implemented by the processing circuit 1000, which acts as a processor 1001, reading and executing a program stored in the memory 1002.

[0107] As described above, the monitoring and control devices 100~100E can realize the above functions through hardware, software, or a combination thereof.

[0108] <Other Application Examples>

[0109] The monitoring and control devices 100-100E described in embodiments 1-6 above can be installed in a server computer constituting a cloud environment, and can input monitoring data to the server computer in the cloud via a communication network, use the server computer to determine the causal relationship, and display it on a display device connected to the communication network.

[0110] Figure 25 This is a block diagram illustrating the structure of a monitoring and control device 100-100E installed within a server computer 300 constituting a cloud environment.

[0111] like Figure 25 As shown, the server computer 300 includes a monitoring project information management DB10, a monitoring project inter-relationship definition unit 20, an association model management DB30, a monitoring data management DB40, a causal relationship calculation unit 50, and a cause determination unit 60. These functions are the same as those of the constituent elements of the monitoring and control devices 100-100E.

[0112] Monitoring data detected by multiple sensors SC located throughout the factory is input to a server computer 300 via a communication network. The server computer 300 determines the causal relationship and displays it on a display device DP connected to the communication network.

[0113] Factory managers use the causal relationships displayed on the DP display device to perform factory maintenance and management. By setting up the monitoring and control device 100~100E in the server computer 300 that constitutes the cloud environment, it can be accessed from anywhere, increasing convenience.

[0114] <Implementation Method 7>

[0115] Figure 26 This is a block diagram illustrating the structure of the monitoring and control system 400 according to Embodiment 7 of this disclosure. Figure 26 As shown, the monitoring and control system 400 includes a monitoring item information management DB10, a monitoring item correlation definition unit 20, a correlation model management DB30, a monitoring data management DB40, a causal relationship calculation unit 50, a cause determination unit 60, a display unit 70, and a sensor 80.

[0116] In addition to the components of the monitoring and control devices 100 to 100E, the monitoring and control system 400 also includes multiple sensors 80 installed at various locations in the factory.

[0117] By constructing such a monitoring and control system 400, the causal relationship of the anomaly can be determined, and the cause of the anomaly detected by the sensor 80 can be determined.

[0118] While this disclosure has been described in detail, the foregoing description is illustrative in all respects only, and this disclosure is not limited thereto. Numerous variations not illustrated can be conceived without departing from the scope of this disclosure.

[0119] Furthermore, this disclosure allows for free combination of various embodiments or appropriate modification or omission of various embodiments within its scope.

Claims

1. A monitoring and control device for monitoring multiple monitoring items, wherein, The monitoring and control device includes: The association definition section defines the associations between the monitoring items among the multiple monitoring items; The association model management database uses the multiple monitoring items that have associations defined by the association definition department as association models to manage multiple association models; The monitoring data management database manages the monitoring data of the multiple monitoring items. The causality calculation unit, when an anomaly is detected in at least one of the plurality of monitoring items, calculates which monitoring item in the plurality of association models has a causal relationship with the anomaly based on the plurality of association models managed by the association model management database and the monitoring data managed by the monitoring data management database. as well as The cause determination unit determines the cause of the anomaly based on the monitoring items for which a causal relationship exists, as calculated by the causality calculation unit. The causal relationship calculation unit calculates the monitoring items that have a causal relationship with the occurrence of the anomaly based on the time series changes of the monitoring data before and after the anomaly occurs.

2. The monitoring and control device according to claim 1, wherein, The related definition part defines This refers to equipment associations, which include at least facilities, equipment, machines, and sensors. At least including the wiring association of the power system's wiring, And at least the treatment process associations in water purification plants and wastewater treatment plants, As a correlation between the monitored items.

3. The monitoring and control device according to claim 2, wherein, The association definition section defines at least the correlations based on the experience values ​​of skilled personnel monitoring and the correlations derived through statistical analysis, as the associations between the monitored items.

4. The monitoring and control device according to claim 1, wherein, The association definition section defines the system association between the monitoring and control system and the equipment management system, which serves as the association between the monitored items.

5. The monitoring and control device according to claim 1, wherein, The association definition section defines the association of information from different sensors and data sources for the same object, as the association between the monitoring items.

6. The monitoring and control device according to claim 1, wherein, The causality calculation unit calculates the monitoring items that detected the anomaly and the monitoring items with similar time-series changes in the monitoring data before and after the anomaly as monitoring items that have a causal relationship with the anomaly.

7. The monitoring and control device according to claim 6, wherein, The causality calculation unit identifies, from the monitoring data of the plurality of monitoring items managed by the monitoring data management database, monitoring items whose change amount is similar to the change amount between the previous and current detection values ​​of the monitoring item that detected the anomaly as having a causal relationship with the occurrence of the anomaly.

8. The monitoring and control device according to claim 1, wherein, The causal relationship calculation unit: Among the multiple association models defined in the association definition section, if the association model with a high causal relationship to the occurrence of the anomaly is known in advance, the monitoring items for that association model are weighted. Weights are also considered to calculate the monitoring items that have a causal relationship with the occurrence of the anomaly.

9. The monitoring and control device according to claim 3, wherein, The cause determination unit combines the device association, the wiring association, the processing flow association, and the related association to determine the cause of the anomaly.

10. The monitoring and control device according to claim 3, wherein, The association definition unit takes the experience value and statistical value of the skilled personnel being monitored as input information, and the calculation result of the association between the monitored items calculated by machine learning based on artificial intelligence as input to define the relevant association.

11. The monitoring and control device according to claim 1, wherein, The device includes a display unit that shows the monitoring items that have the causal relationship with the occurrence of the anomaly and the cause of the anomaly.

12. A monitoring and control system for monitoring multiple monitoring items, wherein, The monitoring and control system includes: The association definition section defines the associations between the monitoring items among the multiple monitoring items; The association model management database uses the multiple monitoring items that have associations defined by the association definition department as association models to manage multiple association models; The monitoring data management database manages the monitoring data of the multiple monitoring items. The causality calculation unit, when an anomaly is detected in at least one of the plurality of monitoring items, calculates which monitoring item in the plurality of association models has a causal relationship with the anomaly based on the plurality of association models managed by the association model management database and the monitoring data managed by the monitoring data management database. as well as The cause determination unit determines the cause of the anomaly based on the monitoring items for which a causal relationship exists, as calculated by the causality calculation unit. The multiple monitoring items include multiple sensors. The causality calculation unit calculates the monitoring items that have a causal relationship with the occurrence of the anomaly based on the time series changes of the monitoring data before and after the anomaly occurred.

13. The monitoring and control system according to claim 12, wherein, have: A display section shows the monitoring items that have a causal relationship with the occurrence of the anomaly and the cause of the anomaly.

Images