Anomaly diagnosis device, anomaly diagnosis method, and program

Abstract

To enable an analyzer to specify a combination of sensors indicating a cause of anomaly in a monitor target with less labor and in a short time.SOLUTION: When an anomaly is detected, a combination of two sensors is selected from a plurality of sensors, the combination including: as one sensor, a sensor corresponding to an SN ratio up to a first predetermined order from a highest order when the SN ratios are arranged in descending order; and, as the other sensor, a sensor other than the one sensor and satisfying a predetermined selection condition. An index indicating a quantity of state of the one sensor included in each of the selected combinations of two sensors is set as a first index, and an index indicating a quantity of state of the other sensor is selected as a second index. For each combination of the selected first index and second index, a scale of the first index, a scale of the second index, and respective positions of the quantities of state specified by the scales of the first index and the second index are displayed on a display screen.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present disclosure relates to an abnormality diagnosis device, an abnormality diagnosis method, and a program. [Background technology]

[0002] For example, the Mahalanobis-Taguchi method (hereinafter referred to as the MT method) is used as a remote monitoring method for plants such as gas turbines. In the MT method, when the plant to be monitored is operating normally, multiple state quantities detected by various sensors installed in the monitored plant are acquired, and the acquired multiple state quantities are used as a reference. After that, the state quantities detected by the various sensors are acquired sequentially over time, and a signal space is determined using the acquired state quantities. Each time a signal space is obtained, the difference between the signal space and the unit space determined using the reference state quantities is calculated as the Mahalanobis distance. If the calculated Mahalanobis distance exceeds, for example, a predetermined threshold, it is determined that an abnormality has occurred in the monitored plant.

[0003] In the MT method, the signal-to-noise (SN) ratio is further calculated as an index showing the degree of influence on the Mahalanobis distance for each state quantity, i.e., for each sensor. By using this SN ratio, it is possible to identify which sensor detects an abnormality in the state quantity, in other words, which part of the monitored object that each sensor detects has an abnormality (for example, see Patent Documents 1 and 2). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2011-238148 A [Patent Document 2] JP 2021-140540 A Summary of the Invention [Problem to be solved by the invention]

[0005] The state quantity detected by the sensor with the highest S / N ratio is one of the factors that increases the Mahalanobis distance. However, abnormalities that occur in the monitored object are caused by various factors. For example, there are cases where the cause of the abnormality does not exist in the location detected by the sensor with the highest S / N ratio, and the change in the state of the location that causes the abnormality changes the state quantity of the location detected by the sensor with the highest S / N ratio.

[0006] In such cases, an analyst who analyzes an anomaly needs to perform an analysis task to identify the cause of the anomaly while referring to the relationship between the state quantities of multiple sensors. However, since a large number of sensors are installed on the monitored object, the number of sensor combinations becomes enormous. Therefore, there is a problem that, in the analysis task of identifying the cause of the anomaly, an analyst needs a lot of effort and time to identify which sensor combination to refer to.

[0007] The present invention has been made to solve the above-mentioned problems, and aims to provide an abnormality diagnosis device, an abnormality diagnosis method, and a program that enable an analyst to identify, with little effort and in a short time, a combination of sensors that indicates the cause of an abnormality occurring in a monitored object. [Means for solving the problem]

[0008] In order to solve the above problems, an abnormality diagnosis device according to the present disclosure includes an abnormality detection unit that performs abnormality detection by applying the Mahalanobis-Taguchi method to state quantities detected by each of a plurality of sensors provided to a monitoring target, and calculates an S / N ratio corresponding to each of the sensors; a sensor selection unit that, when the abnormality detection unit detects an abnormality, selects a combination of two sensors from the plurality of sensors, where one sensor includes a sensor corresponding to an S / N ratio up to a first predetermined rank counting from the top when the S / N ratios are arranged in descending order, and the other sensor includes a sensor other than the one sensor that satisfies a predetermined selection condition; and a display processing unit that selects an index representing the state quantity of one sensor included in each of the combinations of the two sensors selected by the sensor selection unit as a first index and an index representing the state quantity of the other sensor as a second index, and displays, for each combination of the first index and the second index selected, a scale of the first index, a scale of the second index, and each position of the state quantities identified by the scales of the first index and the second index on a display screen.

[0009] The abnormality diagnosis method according to the present disclosure includes an abnormality detection step of detecting an abnormality by applying the Mahalanobis-Taguchi method to state quantities detected by each of a plurality of sensors provided to a monitoring target, and calculating an S / N ratio corresponding to each of the sensors; a sensor selection step of selecting, when an abnormality is detected by the abnormality detection step, a combination of two sensors selected from the plurality of sensors, the combination including, as one sensor, a sensor corresponding to an S / N ratio in a first predetermined order counting from the top when the S / N ratios are arranged in descending order, and the other sensor including a sensor other than the one sensor that satisfies a predetermined selection condition; and a display processing step of selecting, as a first index, an index representing a state quantity of one sensor included in each of the combinations of two sensors selected by the sensor selection step, and selecting, as a second index, an index representing a state quantity of the other sensor, and displaying, for each combination of the first index and the second index selected, a scale of the first index, a scale of the second index, and each position of the state quantities identified by the scales of the first index and the second index on a display screen.

[0010] A program according to the present disclosure causes a computer to function as: anomaly detection means for detecting an anomaly by applying the Mahalanobis-Taguchi method to state quantities detected by each of a plurality of sensors provided on a monitored object, and calculating an S / N ratio corresponding to each of the sensors; a sensor selection means for selecting, when the anomaly detection means detects an anomaly, a combination of two sensors to be selected from the plurality of sensors, the combination including, as one sensor, a sensor corresponding to an S / N ratio in a first predetermined order counting from the top when the S / N ratios are arranged in descending order, and a sensor other than the one sensor that satisfies a predetermined selection condition, and a display processing means for selecting, as a first index, an index representing a state quantity of one sensor included in each of the combinations of two sensors selected by the sensor selection means, and selecting, as a second index, an index representing a state quantity of the other sensor, and displaying, for each combination of the first index and the second index selected, a scale of the first index, a scale of the second index, and each position of the state quantities specified by the scales of the first index and the second index on a display screen. Effect of the Invention

[0011] According to the abnormality diagnosis device, abnormality diagnosis method, and program disclosed herein, an analyst can identify, with little effort and in a short time, a sensor combination that indicates the cause of an abnormality occurring in a monitored object. [Brief description of the drawings]

[0012] [Figure 1] 1 is a block diagram showing a configuration of an abnormality diagnosis device according to an embodiment of the present disclosure, a plant that is a monitoring target of the abnormality diagnosis device, and a sensor and a display device that are connected to the abnormality diagnosis device. [Diagram 2] 4 is a diagram illustrating a data format of a table stored in a state quantity storage unit of the abnormality diagnosis device according to the embodiment of the present disclosure. FIG. [Diagram 3] 4 is a diagram showing a data format of a table stored in an S / N ratio storage unit of the abnormality diagnosis device according to the embodiment of the present disclosure. FIG. [Figure 4]4 is a flowchart showing a flow of processing by the abnormality diagnosis device according to the embodiment of the present disclosure. [Diagram 5] 13 is a diagram showing an example of a graph displayed on a display screen of a display device connected to the abnormality diagnosis device of the present disclosure. [Figure 6] FIG. 1 is a schematic block diagram illustrating a configuration of a computer according to at least one embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] Hereinafter, an abnormality diagnosis device, an abnormality diagnosis method, and a program according to an embodiment of the present disclosure will be described with reference to FIGS. 1 to 5. FIG. 1 is a block diagram showing the configuration of an abnormality diagnosis device 1 according to an embodiment of the present disclosure, a plant 30 that is a monitoring target of the abnormality diagnosis device 1, and sensors 20-1 to 20-N and a display device 2 connected to the abnormality diagnosis device 1. FIGS. 2 and 3 are diagrams showing data formats of tables stored in the state quantity storage unit 12 and the SN ratio storage unit 14 of the abnormality diagnosis device 1 according to an embodiment of the present disclosure, respectively. FIG. 4 is a flowchart showing a flow of processing by the abnormality diagnosis device 1 according to an embodiment of the present disclosure. FIG. 5 is a diagram showing an example of a graph displayed on a display screen of a display device 2 connected to the abnormality diagnosis device 1 of the present disclosure. Note that in each drawing, the same or corresponding configurations are designated by the same reference numerals and the description is omitted as appropriate.

[0014] (Example of configuration of an abnormality diagnosis device) 1, the abnormality diagnosis device 1 is connected to a display device 2 and sensors 20-1 to 20-N, and monitors the state of a plant 30, which is an object of monitoring by the abnormality diagnosis device 1, by using the sensors 20-1 to 20-N. The plant 30 is, for example, a gas turbine. The display device 2 is, for example, a liquid crystal display having a display screen.

[0015] The sensors 20-1 to 20-N are various sensors such as temperature sensors, pressure sensors, and voltage sensors, and are installed at locations in the plant 30 that require monitoring. Here, N is an integer equal to or greater than 2. Each of the sensors 20-1 to 20-N includes a clock or other internal timekeeping means, detects the state of the monitored location in the plant 30 in which it is installed at a constant time interval, and obtains the detection time (hereinafter, referred to as detection time) from the clocking means. Each of the sensors 20-1 to 20-N outputs state data including a state quantity indicating the amount of the detected state, information indicating the detection time of the state quantity, and sensor identification information of each of the sensors 20-1 to 20-N. The sensor identification information is information that is given in advance to each of the sensors 20-1 to 20-N so that each of the sensors 20-1 to 20-N can be identified. Here, the state quantity detected by the sensors 20-1 to 20-N is, for example, a temperature value in the case of a temperature sensor, a pressure value in the case of a pressure sensor, and a voltage value in the case of a voltage sensor. Information indicating the time obtained from a clock means provided inside each of the sensors 20-1 to 20-N is, for example, information expressed in the format of year, month, day, hour, minute, and second.

[0016] The abnormality diagnostic device 1 includes a state quantity collecting unit 11, a state quantity storing unit 12, an abnormality detecting unit 13, an S / N ratio storing unit 14, a sensor selecting unit 15, and a display processing unit 16.

[0017] The state quantity collecting unit 11 is connected to the sensors 20-1 to 20-N via, for example, an electric line or a communication line, and acquires state data output from each of the sensors 20-1 to 20-N. The state quantity collecting unit 11 corrects the difference in detection time of each of the acquired state data, and associates the detection time with the state data so that N pieces of state data exist at one detection time. The state quantity collecting unit 11 records, in the state quantity storage unit 12, the N state quantities included in the N pieces of state data, the sensor identification information corresponding to each of the N state quantities, and information indicating one detection time corresponding to the N state quantities.

[0018] As shown in FIG. 2, the state quantity storage unit 12 stores a table in a data format having items of "detection time", each of N pieces of "sensor identification information", and "type". In the "detection time" item, the state quantity collection unit 11 records information indicating the detection time. In each of the N pieces of "sensor identification information" items, the state quantity collection unit 11 records a state quantity corresponding to each of the sensor identification information items in the item. In the "type" item, one of "unit space", "signal space", or "abnormality" is recorded. When recording a record in the state quantity storage unit 12, the state quantity collection unit 11 records "signal space" in the "type" item.

[0019] A user of the abnormality diagnosis device 1, i.e., an analyst who analyzes an abnormality, operates an operation unit (not shown) provided in the abnormality diagnosis device 1 to select some records during the period when the plant 30 is operating normally from the records in the state quantity storage unit 12, and rewrites the "type" item of the selected records from "signal space" to "unit space." Note that rewriting from "signal space" to "unit space" is preferably performed, for example, by the analyst periodically, and consecutive records during the most recent period of about 2 to 3 months during which the plant 30 is operating normally are rewritten from "signal space" to "unit space" and recorded in the state quantity storage unit 12.

[0020] The anomaly detection unit 13 determines the unit space using the state quantities of records whose "type" field is "unit space" among the records recorded in the state quantity storage unit 12. The anomaly detection unit 13 continuously performs processing to monitor the processing of recording a new record in the state quantity storage unit 12 by the state quantity collection unit 11, and when a new record is recorded in the state quantity storage unit 12, determines a signal space using the state quantities of the newly recorded record.

[0021] The anomaly detection unit 13 applies the MT method to the state quantity in the signal space and the state quantity in the unit space to calculate the Mahalanobis distance in the signal space. When the calculated Mahalanobis distance exceeds a predetermined threshold, the anomaly detection unit 13 determines that an anomaly has occurred in the plant 30. When the anomaly detection unit 13 determines that an anomaly has occurred in the plant 30, it rewrites the "type" item of the record corresponding to the signal space at the time of the determination from "signal space" to "anomaly."

[0022] The abnormality detection unit 13 calculates an SN ratio, which is an index showing the degree of influence on the Mahalanobis distance for each of the sensors 20-1 to 20-N, based on each of the state quantities of the record in the state quantity storage unit 12 of the determination target when it is determined that an abnormality has occurred and the state quantity of the unit space. The abnormality detection unit 13 records information showing the detection time of the record in the state quantity storage unit 12 of the determination target when it is determined that an abnormality has occurred and the calculated SN ratio for each of the sensors 20-1 to 20-N in the SN ratio storage unit 14. When the recording in the SN ratio storage unit 14 is completed, the abnormality detection unit 13 outputs an abnormality occurrence notification signal including information showing the detection time recorded in the SN ratio storage unit 14.

[0023] 3, the SN ratio storage unit 14 stores a table in a data format having items of "detection time" and each of N pieces of "sensor identification information." In the "detection time" item, information indicating the detection time is recorded by the abnormality detection unit 13. In each of the N pieces of "sensor identification information" items, the abnormality detection unit 13 records an SN ratio corresponding to each of the sensor identification information in the item.

[0024] The sensor selection unit 15 selects a combination of two sensors from among the multiple sensors 20-1 to 20-N. The sensor selection unit 15 includes a sensor combination generation unit 15-1 and a two-dimensional MD (Mahalanobis-Distance) calculation unit 15-2. When the sensor combination generation unit 15-1 receives an abnormality occurrence notification signal output by the abnormality detection unit 13, the sensor combination generation unit 15-1 detects a record corresponding to information indicating a detection time included in the received abnormality occurrence notification signal from the S / N ratio storage unit 14. The sensor combination generation unit 15-1 arranges the N pieces of sensor identification information included in the detected record in descending order of S / N ratio. The sensor combination generation unit 15-1 selects sensor identification information between the top and a predetermined rank (hereinafter referred to as a first predetermined rank).

[0025] The sensor combination generation unit 15-1 generates combination data in which the selected sensor identification information is included as one sensor identification information, and sensor identification information other than the one sensor identification information and corresponding to sensors 20-1 to 20-N that satisfy a predetermined selection condition is included as the other sensor identification information.

[0026] The two-dimensional MD calculation unit 15-2 reads out state quantities corresponding to two pieces of sensor identification information included in the combination data generated by the sensor combination generation unit 15-1 from the state quantity storage unit 12, and determines a unit space and a signal space using the read out state quantities. The two-dimensional MD calculation unit 15-2 calculates the two-dimensional Mahalanobis distance of the signal space based on the determined unit space and signal space.

[0027] The two-dimensional MD calculation unit 15-2 calculates the two-dimensional Mahalanobis distance corresponding to each of the combination data generated by the sensor combination generation unit 15-1, and arranges the combination data in descending order of the two-dimensional Mahalanobis distance. The two-dimensional MD calculation unit 15-2 selects combination data between the top and a predetermined rank (hereinafter referred to as the second predetermined rank).

[0028] The display processing unit 16 is connected to the display device 2 via, for example, an electric line. The display processing unit 16 reads out state quantities corresponding to each of the two sensor identification information included in the combination data selected by the two-dimensional MD calculation unit 15-2 from the state quantity storage unit 12. The display processing unit 16 selects an index representing the state quantity of one sensor of the two sensor identification information included in the combination data selected by the two-dimensional MD calculation unit 15-2 as a first index, and an index representing the state quantity of the other sensor as a second index. For each combination of the selected first index and second index, the display processing unit 16 generates image data of a graph represented by the scale of the first index and the scale of the second index, in which the positions of each of the read state quantities are indicated by the scales of the first index and the second index. When generating the image data of the graph, the display processing unit 16 generates the image data of the graph so that the positions of the state quantities are indicated in different forms for each type of the read state quantity, i.e., for each type of the read state quantity, i.e., "unit space", "signal space", and "abnormality". The display processing unit 16 outputs the image data of the generated graph to the display device 2. The display device 2 uses the image data of the graph output by the display processing unit 16 to display an image of the graph on the display screen.

[0029] (Processing by abnormality diagnosis device) 4 is a flowchart showing the flow of processing performed by sensor selection unit 15 and display processing unit 16 after abnormality detection unit 13 outputs an abnormality occurrence notification signal in abnormality diagnosis device 1. In the following, the above-mentioned predetermined selection condition is a condition that all sensors 20-1 to 20-N are selected except for sensors 20-1 to 20-N corresponding to one of the sensor identification information, and the above-mentioned first predetermined ranking is the highest ranking and the above-mentioned second predetermined ranking is the third ranking, will be described.

[0030] The sensor combination generation unit 15-1 takes in and acquires an abnormality occurrence notification signal output by the abnormality detection unit 13 (S1). The sensor combination generation unit 15-1 reads out a record corresponding to information indicating a detection time included in the acquired abnormality occurrence notification signal from the S / N ratio storage unit 14. The sensor combination generation unit 15-1 arranges the N pieces of sensor identification information included in the read out record in descending order of S / N ratio. The sensor combination generation unit 15-1 selects the sensor identification information that is highest in the order, i.e., the sensor identification information with the highest S / N ratio (S2).

[0031] The sensor combination generation unit 15-1 selects the sensor identification information with the largest SN ratio as the main sensor identification information (hereinafter referred to as main sensor identification information) when generating the combination. The sensor combination generation unit 15-1 generates N-1 combination data in which the main sensor identification information is included as one of the sensor identification information and each of the sensor identification information other than the main sensor identification information is included as the other sensor identification information according to a predetermined selection condition. When generating the combination data, the sensor combination generation unit 15-1 assigns an identification label indicating that the main sensor identification information is the main sensor identification information to generate the combination data.

[0032] The sensor combination generation unit 15-1 associates, with each of the generated N-1 pieces of combination data, information indicating the detection time in the "detection time" field of the record read from the S / N ratio storage unit 14 when generating the N-1 pieces of combination data. The sensor combination generation unit 15-1 outputs the N-1 pieces of combination data, each associated with information indicating the detection time, to the two-dimensional MD calculation unit 15-2 (S3).

[0033] The two-dimensional MD calculation unit 15-2 receives N-1 pieces of combination data, each of which is associated with information indicating a detection time, output by the sensor combination generation unit 15-1. The two-dimensional MD calculation unit 15-2 selects one of the received N-1 pieces of combination data.

[0034] The two-dimensional MD calculation unit 15-2 reads out all state quantities corresponding to the two sensor identification information pieces included in the selected combination data, which state quantities have the "type" item set to "unit space", from the state quantity storage unit 12. The two-dimensional MD calculation unit 15-2 determines the unit space corresponding to the selected combination data, using the read out state quantities whose "type" item set corresponds to "unit space".

[0035] The two-dimensional MD calculation unit 15-2 detects one record corresponding to information indicating a detection time associated with the selected combination data from the state quantity storage unit 12. The two-dimensional MD calculation unit 15-2 reads out two state quantities corresponding to each of the two sensor identification information included in the selected combination data from the detected one record. The two-dimensional MD calculation unit 15-2 determines a signal space corresponding to the selected combination data using the two read out state quantities. The two-dimensional MD calculation unit 15-2 calculates a two-dimensional Mahalanobis distance of the signal space using the signal space and a unit space determined corresponding to the selected combination data. The two-dimensional MD calculation unit 15-2 further associates the calculated two-dimensional Mahalanobis distance with the selected combination data. The two-dimensional MD calculation unit 15-2 records the combination data in which information indicating a detection time and the two-dimensional Mahalanobis distance are associated with each other in an internal storage area (S4).

[0036] The two-dimensional MD calculation unit 15-2 selects any one of the combination data that is not the target of the S4 process, and repeats the process of S4 for the selected combination data until there is no more combination data to be processed in S4 (loops L1s to L1e). The process of the flowchart in FIG. 4 is performed every time the abnormality detection unit 13 outputs an abnormality occurrence notification signal. Therefore, in the first process of S4 in the process of loops L1s to L1e, the two-dimensional MD calculation unit 15-2 initializes the internal storage area before recording the combination data in which the information indicating the detection time and the two-dimensional Mahalanobis distance are associated with each other in the internal storage area. This erases the combination data from the previous abnormality occurrence notification signal.

[0037] The two-dimensional MD calculation unit 15-2 reads out from an internal storage area N-1 pieces of combination data in which the two-dimensional Mahalanobis distance is associated with information indicating the detection time. The two-dimensional MD calculation unit 15-2 arranges the read N-1 pieces of combination data in which the two-dimensional Mahalanobis distance is associated with information indicating the detection time in descending order of the two-dimensional Mahalanobis distance. The two-dimensional MD calculation unit 15-2 selects the three combination data from the top to the third place. The two-dimensional MD calculation unit 15-2 further associates information indicating the order with the three selected combination data in which information indicating the detection time is associated, and outputs the combined data to the display processing unit 16 (S5).

[0038] The display processing unit 16 takes in three combination data output by the two-dimensional MD calculation unit 15-2, which are combination data associated with information indicating the order and information indicating the detection time. The display processing unit 16 performs a process of generating image data of a graph from each of the combination data according to the order, for example, in the procedure shown below. That is, the display processing unit 16 first selects the combination data ranked first, and reads out the sensor identification information to which an identification label indicating that it is the main sensor identification information and the sensor identification information to which no identification label is assigned, which are included in the selected combination data. Here, it is assumed that the sensor corresponding to the sensor identification information to which an identification label is assigned is the sensor 20-A (where A is an integer between 1 and N), and the sensor corresponding to the sensor identification information to which no identification label is assigned is the sensor 20-P (where P is an integer between 1 and N, which is different from A).

[0039] The display processing unit 16 sets an index representing the state quantity of the sensor 20-A as a first index, and sets an index representing the state quantity of the sensor 20-P as a second index. The display processing unit 16 reads out from the state quantity storage unit 12 the state quantities corresponding to the sensor identification information of the sensors 20-A and 20-P, whose "type" item is "unit space", and assigns a label indicating "unit space" to the read out state quantities.

[0040] The display processing unit 16 reads out from the state quantity storage unit 12 the state quantities corresponding to each of the sensor identification information of the sensors 20-A and 20-P, the state quantities in which the date of the detection time in the "detection time" field matches the date indicated by the information indicating the detection time associated with the first-place combination data, and the state quantities in which the "type" field is "signal space" or "abnormality." In other words, the display processing unit 16 reads out from the state quantity storage unit 12 the state quantities corresponding to each of the sensor identification information of the sensors 20-A and 20-P, the state quantities of "signal space" or "abnormality" detected on the same day as the day the abnormality occurred.

[0041] When the display processing unit 16 reads out from the state quantity storage unit 12 the state quantity of "signal space" or "abnormality" detected on the day when an abnormality occurred corresponding to each of the sensor identification information of the sensors 20-A and 20-P, the display processing unit 16 also reads out information in the "type" field corresponding to the read out state quantity, i.e., information on "signal space" or "abnormality." The display processing unit 16 assigns to the read out state quantity the information in the "type" field read out corresponding to the state quantity.

[0042] As a result, the display processing unit 16 obtains state quantities corresponding to each of the sensor identification information of the sensors 20-A, 20-P that correspond to the first-place combination data, with each state quantity being assigned one of the information types "unit space," "signal space," or "abnormal."

[0043] The display processing unit 16 defines an index representing the state quantity of the sensor 20-A as a first index, and defines an index representing the state quantity of the sensor 20-P as a second index. The display processing unit 16 generates image data of a scatter plot graph, in which the scale of the first index is the vertical axis, the scale of the second index is the horizontal axis, and the positions of the state quantities specified by the scales of the first index and the second index, and the types of the state quantities are indicated. In order to display the positions and types of the state quantities, the display processing unit 16 generates image data of the graph so that, for example, a circular mark is displayed at the position of the state quantity to which "unit space" is assigned, a diamond mark is displayed at the position of the state quantity to which "signal space" is assigned, and a star mark is displayed at the position of the state quantity to which "abnormality" is assigned. The display processing unit 16 outputs image data of the generated graph to the display device 2. The display device 2 imports the image data of the graph output by the display processing unit 16, and displays an image of the graph on the display screen using the image data of the imported graph (S6). As a result, for example, the graph shown in FIG. 5(a) is displayed on the display screen of the display device 2.

[0044] The display processing unit 16 then selects the combination data ranked second and performs the process of S6. One of the sensors corresponding to the main sensor identification information included in the combination data ranked second is the sensor 20-A. Here, the other sensor included in the combination data ranked second is the sensor 20-Q (where Q is an integer between 1 and N, and is an integer different from A and P). In this case, the display processing unit 16 performs the process of S6 on the combination data ranked second to generate image data of a scatter plot graph with the first index corresponding to the sensor 20-A as the vertical axis and the second index corresponding to the sensor 20-Q as the horizontal axis. As a result, for example, a graph shown in FIG. 5(b) is displayed on the display screen of the display device 2.

[0045] Finally, the display processing unit 16 selects the third-ranked combination data and performs the process of S6. One sensor corresponding to the main sensor identification information included in the third-ranked combination data is the sensor 20-A. Here, the other sensor included in the third-ranked combination data is the sensor 20-R (where R is an integer between 1 and N, and is an integer different from A, P, and Q). In this case, the display processing unit 16 generates image data of a scatter plot graph with the first index corresponding to the sensor 20-A as the vertical axis and the second index corresponding to the sensor 20-R as the horizontal axis by the process of S6 for the third-ranked combination data. As a result, for example, a graph shown in FIG. 5(c) is displayed on the display screen of the display device 2 (loop L2s to L2e). When the process of loops L2s to L2e ends, the process of the flowchart shown in FIG. 4 ends.

[0046] (Actions and Effects of the Abnormality Diagnosis Device) When the process of FIG. 4 is completed, three graphs, which are shown in FIG. 5(a), (b), and (c), are displayed on the display screen of the display device 2. The graph shown in FIG. 5(a) is a graph showing the relationship between the state quantity of the sensor 20-P, which has the largest two-dimensional Mahalanobis distance, and the state quantity of the sensor 20-A, in relation to the state quantity of the sensor 20-A. The graph shown in FIG. 5(b) is a graph showing the relationship between the state quantity of the sensor 20-Q, which has the second largest two-dimensional Mahalanobis distance, and the state quantity of the sensor 20-A, in relation to the state quantity of the sensor 20-A. The graph shown in FIG. 5(c) is a graph showing the relationship between the state quantity of the sensor 20-R, which has the third largest two-dimensional Mahalanobis distance, and the state quantity of the sensor 20-A, in relation to the state quantity of the sensor 20-A. In addition, since five star-shaped marks indicating the type of "abnormality" are displayed in the graphs of FIG. 5(a), (b), and (c), these are graphs displayed on the display screen of the display device 2 when the abnormality detection unit 13 outputs the abnormality occurrence notification signal for the fifth time. The analyst refers to these three graphs and analyzes how far the state quantity indicating an abnormality is from the unit space, and thereby can identify the location where the abnormality is occurring in the plant 30.

[0047] The number of combinations of selecting two sensors 20-1 and 20-N from the N sensors 20-1 to 20-N is given as follows: N C2, that is, N×(N-1) / 2. When the number of N is large, the number of combinations becomes enormous. In response to this, the sensor selection unit 15 of the abnormality diagnosis device 1 narrows down the combinations to be displayed from among the N×(N-1) / 2 combinations based on two indices, the magnitude of the SN ratio and the magnitude of the two-dimensional Mahalanobis distance. The two indices, the magnitude of the SN ratio and the magnitude of the two-dimensional Mahalanobis distance, can be said to be indices indicating the degree of relevance to an abnormality. Therefore, the combinations narrowed down by the sensor selection unit 15 are combinations with a high relevance to an abnormality, and the graph displayed on the display screen of the display device 2 is a graph showing the relationship between the combinations with a high relevance to an abnormality. In other words, by narrowing down the combinations based on the two indices, the magnitude of the SN ratio and the magnitude of the two-dimensional Mahalanobis distance, it is possible to increase the possibility that the graph displayed on the display screen of the display device 2 includes a graph showing the characteristics of the state quantities of the sensors 20-1 to 20-N that indicate the cause of the abnormality that has occurred. Therefore, by using the abnormality diagnostic device 1, an analyst can specify the combination of sensors that indicates the cause of an abnormality occurring in the plant 30 that is the monitoring target with little effort and in a short time.

[0048] <Other configuration examples of the embodiment> (Another example of a configuration with a specific order) The first and second predetermined ranks, which are the premise of the process in Fig. 4, are merely examples, and the first and second predetermined ranks may be arbitrary ranks. For example, if the second predetermined rank is set to the top, the two-dimensional MD calculation unit 15-2 outputs only the combination data with the largest two-dimensional Mahalanobis distance to the display processing unit 16. In the example of Fig. 4, the display processing unit 16 displays only the graph in Fig. 5(a) on the display screen of the display device 2.

[0049] By lowering the rank indicated by the first predetermined rank, the number of combination data generated by the sensor combination generation unit 15-1 increases, and the processing load on the two-dimensional MD calculation unit 15-2 increases, but on the other hand, it becomes possible to extract combinations that are highly related to the occurring abnormality over a wider range. Therefore, the first predetermined rank is determined in consideration of the analysis range and the degree of the processing load on the two-dimensional MD calculation unit 15-2. Note that, if the first predetermined rank is set to second rank or lower, the sensor combination generation unit 15-1 may select sensor identification information of any one or more ranks between the top rank and the first predetermined rank.

[0050] Since the second predetermined rank is related to the number of graphs that the display processing unit 16 displays on the display screen of the display device 2, by lowering the order indicated by the second predetermined rank, the display processing unit 16 displays more graphs on the display screen of the display device 2. Therefore, the analyst will refer to more graphs for analysis, and the labor and time required for the analyst's analysis will increase, but on the other hand, since the analyst's analysis targets increase, the possibility of identifying the cause of the abnormality occurring more accurately can be increased. Therefore, the second predetermined rank is determined in consideration of the labor and time required for the analyst's analysis and the required accuracy. Note that, when the second predetermined rank is set to the second rank or lower, the two-dimensional MD calculation unit 15-2 may select combination data of any one or more ranks between the top rank and the second predetermined rank.

[0051] (Another example of a specific selection condition (part 1)) The predetermined selection condition assumed in the process of Fig. 4 is one example, and the following predetermined selection condition may be applied. For example, the predetermined selection condition may be a condition that the number of sensors is two or more than the predetermined number of arbitrary sensors 20-1 to 20-N other than the sensor 20-A corresponding to the main sensor identification information. By using this condition, the amount of combination data generated by the sensor combination generation unit 15-1 can be reduced, and therefore the processing load on the two-dimensional MD calculation unit 15-2 can be reduced.

[0052] (Another example of a specific selection condition (part 2)) As a predetermined selection condition, a condition may be applied that the sensors 20-1 to 20-N are other than the sensor 20-A corresponding to the main sensor identification information, and that the degree of correlation between the state quantity of the sensor 20-A and the state quantity of the sensors 20-1 to 20-N other than the sensor 20-A is high in relation to a predetermined threshold. A specific process when this condition is applied will be described. For example, in the process of S3 in FIG. 4, the sensor combination generation unit 15-1 determines the main sensor identification information, and then reads out from the state quantity storage unit 12 the state quantity of the main sensor identification information, which has the item "type" as "unit space". The sensor combination generation unit 15-1 selects one of the sensor identification information other than the main sensor identification information, and reads out from the state quantity storage unit 12 the state quantity of the selected sensor identification information, which has the item "type" as "unit space". The sensor combination generating unit 15-1 calculates a correlation coefficient based on the state quantity of the unit space of the main sensor identification information and the state quantity of the unit space of the selected sensor identification information.

[0053] The sensor combination generation unit 15-1 calculates the correlation coefficient corresponding to each of the sensor identification information other than the main sensor identification information by the above procedure. The sensor combination generation unit 15-1 selects the sensor identification information in which the absolute value of the correlation coefficient exceeds a predetermined threshold. The sensor combination generation unit 15-1 generates combination data including the main sensor identification information on one side and each of the sensor identification information in which the absolute value of the correlation coefficient exceeds a predetermined threshold on the other side. The sensor combination generation unit 15-1 associates, with each of the generated combination data, information indicating the detection time in the "detection time" field of the record of the SN ratio storage unit 14 referred to when the combination data was generated. The sensor combination generation unit 15-1 outputs the combination data associated with the information indicating the detection time to the two-dimensional MD calculation unit 15-2. After that, the processing of loops L1s to L1e is performed.

[0054] This enables the sensor combination generation unit 15-1 to generate combination data in which one sensor identification information is main sensor identification information that is highly relevant to the occurring abnormality and that is selected based on the magnitude of the S / N ratio, and the other sensor identification information is sensor identification information of sensors 20-1 to 20-N that detects a state quantity that shows a high degree of correlation with the state quantity corresponding to the main sensor identification information.

[0055] In this case, the sensor combination generation unit 15-1 narrows down the number of sensor identification information selected as the other sensor, compared to when the other sensor identification information is all sensor identification information other than the main sensor identification information, and therefore the number of combination data to be generated can be reduced. Therefore, it is possible to reduce the processing load on the two-dimensional MD calculation unit 15-2. Furthermore, the sensor combination generation unit 15-1 does not randomly narrow down the number of sensor identification information selected as the other sensor, but narrows down the sensor identification information of the sensors 20-1 to 20-N that detect state quantities having a high degree of correlation with the state quantity of the sensor 20-1. Therefore, in the process of S6, the display processing unit 16 can further increase the possibility that a graph showing the characteristics of the state quantities of the sensors 20-1 to 20-N that indicate the cause of the occurring abnormality is included in the graph displayed on the display screen of the display device 2.

[0056] The sensor combination generation unit 15-1 may calculate a coefficient of determination instead of calculating the correlation coefficient. Since the coefficient of determination is expressed as a numerical value between 0 and 1, it is not necessary to compare the absolute value with a threshold value as in the case of the correlation coefficient, which is expressed as a numerical value between -1 and 1. Instead, it is sufficient to compare the coefficient of determination with a threshold value.

[0057] (Another example of the configuration of the specified selection condition (part 3)) As the predetermined selection condition, a condition that the sensors 20-1 to 20-N are other than the sensor 20-A corresponding to the main sensor identification information and are predetermined sensors 20-1 to 20-N for the sensor 20-A corresponding to the main sensor identification information may be applied. For example, for each sensor 20-1 to 20-N, other sensors 20-1 to 20-N that are estimated to have a high relevance, such as other sensors 20-1 to 20-N with a large two-dimensional Mahalanobis distance or other sensors 20-1 to 20-N that were useful in identifying the cause of the abnormality, can be specified in advance. List data indicating the correspondence between each of the sensors 20-1 to 20-N specified in this way and the other sensors 20-1 to 20-N corresponding to each of them by the sensor identification information is generated in advance. The generated list data is stored in advance in a storage area inside the sensor combination generation unit 15-1.

[0058] After determining the main sensor identification information in the process of S3 in FIG. 4, the sensor combination generation unit 15-1 refers to the list data stored in the internal storage area and reads out the sensor identification information associated with the main sensor identification information in the list data. The sensor combination generation unit 15-1 generates combination data including the main sensor identification information on one side and each of the sensor identification information read out from the list data on the other side. The sensor combination generation unit 15-1 associates, with each of the generated combination data, information indicating the detection time in the "detection time" field of the record of the SN ratio storage unit 14 referred to when the combination data was generated. The sensor combination generation unit 15-1 outputs the combination data associated with the information indicating the detection time to the two-dimensional MD calculation unit 15-2. After that, the processes of loops L1s to L1e are performed.

[0059] This enables the sensor combination generation unit 15-1 to generate combination data in which one sensor identification information is main sensor identification information that is highly relevant to the occurring abnormality and selected based on the magnitude of the S / N ratio, and the other sensor identification information is sensor identification information of sensors 20-1 to 20-N that is estimated to be highly relevant to the sensor 20-A corresponding to the main sensor identification information.

[0060] In this case, the sensor combination generation unit 15-1 narrows down the number of sensor identification information selected as the other sensor, compared to when the other sensor identification information is all sensor identification information other than the main sensor identification information, and therefore the number of combination data to be generated can be reduced. Therefore, it is possible to reduce the processing load on the two-dimensional MD calculation unit 15-2. Furthermore, the sensor combination generation unit 15-1 does not randomly narrow down the number of sensor identification information selected as the other sensor, but narrows down the sensor identification information of 20-1 to 20-N that is estimated to be highly related to the sensor 20-A of the main sensor identification information. Therefore, in the process of S6, the display processing unit 16 can further increase the possibility that a graph showing the characteristics of the state quantities of the sensors 20-1 to 20-N that indicate the cause of the occurring abnormality is included in the graph displayed on the display screen of the display device 2.

[0061] In addition, when the sensor combination generation unit 15-1 applies either the predetermined selection condition of another configuration example (part 2) of the predetermined selection condition or the predetermined selection condition of another configuration example (part 3) of the predetermined selection condition as the predetermined selection condition, the sensor combination generation unit 15-1 may generate combination data using sensor identification information corresponding to sensors 20-1 to 20-N arbitrarily selected from the sensors 20-1 to 20-N that satisfy the predetermined selection condition, rather than using sensor identification information corresponding to all of the sensors 20-1 to 20-N that satisfy the predetermined selection condition as described above.

[0062] In addition, the sensor combination generation unit 15-1 may apply both the predetermined selection conditions of another configuration example (part 2) of the predetermined selection conditions and the predetermined selection conditions of another configuration example (part 3) of the predetermined selection conditions as the predetermined selection conditions, and generate combination data using sensor identification information corresponding to all of the sensors 20-1 to 20-N that meet the predetermined selection conditions, or may generate combination data using sensor identification information corresponding to sensors 20-1 to 20-N arbitrarily selected from the sensors 20-1 to 20-N that meet the predetermined selection conditions.

[0063] (About graph types) In the above embodiment, the display processing unit 16 indicates the types of state quantities, "unit space", "signal space", and "abnormality", by marks of different shapes. In contrast, the display processing unit 16 may display the graph in any form, such as by indicating the types of state quantities, "unit space", "signal space", and "abnormality", by marks of different colors, as long as the difference between the types of state quantities, "unit space", "signal space", and "abnormality", can be understood. In this case, the display processing unit 16 does not necessarily need to display the graph in a form that makes the difference between the three types, "unit space", "signal space", and "abnormality", clear. For example, the display processing unit 16 may regard the state quantity of "abnormality" as a part of the state quantity of "signal space" and display the graph in a form that makes the difference between the state quantity of "unit space" and the state quantity of "signal space" clear, or may display only the position of the state quantity without considering the types, "unit space", "signal space", and "abnormality".

[0064] In the above embodiment, as shown in Fig. 5, the display processing unit 16 displays a scatter diagram graph in which an index representing the state quantity of the sensor 20-A corresponding to the main sensor identification information to which an identification label is attached is set as a first index, the first index is represented on the vertical axis, and the second index is represented on the horizontal axis. Alternatively, a scatter diagram graph in which the first index is represented on the horizontal axis and the second index is represented on the vertical axis may be used. Alternatively, the sensor combination generating unit 15-1 may generate combination data without assigning an identification label, and the display processing unit 16 may set as a first index an index representing the state quantity of the sensors 20-1 to 20-N corresponding to one sensor identification information randomly selected from two sensor identification information included in the combination data, and set as a second index an index representing the state quantity of the sensors 20-1 to 20-N corresponding to the remaining other sensor identification information. In addition, the display processing unit 16 may generate image data of a graph other than a scatter plot, so long as the graph shows the scales of the first index and the second index, the positions of each of the state quantities specified by the scales of the first index and the second index, and the types of each of the state quantities.

[0065] (Graph display) In the above embodiment, the process of Fig. 4 is performed each time the abnormality detection unit 13 outputs an abnormality occurrence notification signal. In this case, if the abnormality detection unit 13 outputs multiple abnormality occurrence notification signals within a short period of time, three graphs will be additionally displayed on the display screen of the display device 2 for each abnormality occurrence notification signal. In this way, in order to prevent many graphs from being displayed on the display screen of the display device 2, for example, the display processing unit 16 may perform a process of selecting and displaying graphs on the display screen of the display device 2 as follows.

[0066] When the display processing unit 16 receives the combination data output by the two-dimensional MD calculation unit 15-2, in which information indicating the order is associated with information indicating the detection time, the display processing unit 16 records the received combination data in an internal storage area. The analyst operates the display processing unit 16 to specify the detection time via an operation unit provided in the abnormality diagnosis device 1 (not shown). The display processing unit 16 reads out the three combination data corresponding to the detection time received from the operation unit from the internal storage area, and performs the processing of loops L2s to L2e. This makes it possible to display only the images of the three graphs corresponding to the detection time specified by the analyst on the display screen of the display device 2.

[0067] The analyst may specify the detection time and the ranking to the display processing unit 16 via an operation unit (not shown). In this case, the display processing unit 16 reads one combination data corresponding to the detection time and ranking received from the operation unit from an internal storage area, and performs the process of S6. This makes it possible to display only one graph image corresponding to the detection time and ranking specified by the analyst on the display screen of the display device 2.

[0068] (About the 2D MD calculation part) In the abnormality diagnosis device 1 of the above embodiment, the sensor selection unit 15 may be a sensor selection unit 15 that includes only the sensor combination generation unit 15-1 and does not include the two-dimensional MD calculation unit 15-2. In this case, the display processing unit 16 displays a graph on the display screen of the display device 2 based on the combination data generated by the sensor combination generation unit 15-1. Therefore, although the abnormality diagnosis device 1 cannot obtain the effect of narrowing down using the two-dimensional Mahalanobis distance, it is possible to eliminate the processing load of the two-dimensional MD calculation unit 15-2. In this case, the predetermined selection condition applied in the sensor combination generation unit 15-1 may be any of the above-mentioned predetermined selection conditions.

[0069] <Other embodiments> Although the embodiments of the present disclosure have been described in detail above with reference to the drawings, the specific configuration is not limited to this embodiment, and design changes and the like that do not depart from the gist of the present disclosure are also included.

[0070] <Computer configuration> FIG. 6 is a schematic block diagram showing a configuration of a computer according to at least one embodiment. The computer 90 includes a processor 91, a main memory 92, a storage 93, and an interface 94. The sensors 20-1 to 20-N and the display device 2 are connected to the interface 94. The above-mentioned abnormality diagnosis device 1 is implemented in the computer 90. The operations of the state quantity collection unit 11, the abnormality detection unit 13, the sensor selection unit 15, and the display processing unit 16 of the above-mentioned abnormality diagnosis device 1 are stored in the storage 93 in the form of a program. The processor 91 reads the program from the storage 93 and loads it in the main memory 92, and executes the above-mentioned processing according to the program. The processor 91 also secures storage areas in the main memory 92 corresponding to the internal storage areas of the sensor combination generation unit 15-1 and the two-dimensional MD calculation unit 15-2, the state quantity storage unit 12, and the SN ratio storage unit 14 according to the program. The storage areas of the state quantity storage unit 12 and the SN ratio storage unit 14 may be secured in the storage 93.

[0071] The program may be for realizing a part of the functions to be performed by the computer 90. For example, the program may be for realizing the functions by combining with other programs already stored in the storage 93 or by combining with other programs implemented in other devices. In other embodiments, the computer may include a custom LSI (Large Scale Integrated Circuit) such as a PLD (Programmable Logic Device) in addition to or instead of the above configuration. Examples of PLDs include PAL (Programmable Array Logic), GAL (Generic Array Logic), CPLD (Complex Programmable Logic Device), and FPGA (Field Programmable Gate Array). In this case, some or all of the functions to be realized by the processor may be realized by the integrated circuit.

[0072] Examples of the storage 93 include a hard disk drive (HDD), a solid state drive (SSD), a magnetic disk, a magneto-optical disk, a compact disc read only memory (CD-ROM), a digital versatile disc read only memory (DVD-ROM), and a semiconductor memory. The storage 93 may be an internal medium directly connected to the bus of the computer 90, or an external medium connected to the computer 90 via an interface 94 or a communication line. In addition, when the program is distributed to the computer 90 via a communication line, the computer 90 that receives the program may load the program into the main memory 92 and execute the above-mentioned process. In at least one embodiment, the storage 93 is a non-transitory tangible storage medium.

[0073] <Additional Notes> The abnormality diagnostic device 1 described in the above embodiment can be understood, for example, as follows.

[0074] (1) The abnormality diagnosis device 1 according to the first aspect includes an abnormality detection unit 13 that detects an abnormality by applying the Mahalanobis-Taguchi method to state quantities detected by each of a plurality of sensors 20-1 to 20-N provided to a monitoring target, and calculates an S / N ratio corresponding to each of the sensors 20-1 to 20-N. When the abnormality detection unit 13 detects an abnormality, the abnormality detection unit 13 selects two sensors from the plurality of sensors 20-1 to 20-N, and when the S / N ratios are arranged in descending order, the sensors 20-1 to 20-N corresponding to the S / N ratios counting from the top and up to a first predetermined rank that is determined in advance are included as one of the sensors, and a sensor other than the one sensor is The system includes a sensor selection unit 15 that selects a combination in which a sensor 20-1 to 20-N that satisfies a predetermined selection condition is included as the other sensor, and a display processing unit 16 that selects an index representing a state quantity of one sensor 20-1 to 20-N included in each of the combinations of two sensors selected by the sensor selection unit 15 as a first index and an index representing a state quantity of the other sensor 20-1 to 20-N as a second index, and displays on a display screen a scale of the first index, a scale of the second index, and each position of the state quantity specified by the scale of the first index and the second index for each combination of the selected first index and second index. According to this aspect and each of the following aspects, an analyst can specify a sensor combination that indicates the cause of an abnormality occurring in a monitoring target with little effort and in a short time.

[0075] (2) The abnormality diagnosis device 1 according to a second aspect is the abnormality diagnosis device 1 according to (1), in which the sensor selection unit 15 calculates a two-dimensional Mahalanobis distance for each combination of the two sensors selected from the state quantities of the two sensors included in the combination, and when the two-dimensional Mahalanobis distances are sorted in descending order, selects a combination of the two sensors from among the combinations corresponding to the two-dimensional Mahalanobis distances counting from the top to a second predetermined rank that is determined in advance. According to this aspect, it is possible to narrow down the combinations of the two sensors based on the magnitude of the two-dimensional Mahalanobis distance.

[0076] (3) The abnormality diagnosis device 1 according to a third aspect is the abnormality diagnosis device 1 according to (1) or (2), in which the predetermined selection condition is a condition that all sensors 20-1 to 20N other than the one sensor are selected. According to this aspect, it is possible to generate combinations of two sensors by selecting all sensors that can be combined with the one sensor selected based on the S / N ratio.

[0077] (4) The abnormality diagnosis device 1 according to a fourth aspect is the abnormality diagnosis device 1 according to (1) or (2), in which the predetermined selection condition is a condition that the sensors are any predetermined number of two or more sensors 20-1 to 20-N other than the one sensor. According to this aspect, the other sensor of the combination of two sensors can be narrowed down to a predetermined number of sensors 20-1 to 20-N.

[0078] (5) The abnormality diagnosis device 1 according to a fifth aspect is the abnormality diagnosis device 1 according to (1) or (2), and the predetermined selection condition is that the sensors 20-1 to 20-N are sensors other than the one sensor, and the degree of correlation between the state quantity of the one sensor and the state quantity of the sensor other than the one sensor shows a high degree of correlation in relation to a predetermined threshold value. According to this aspect, it is possible to narrow down combinations of two sensors based on the degree of correlation.

[0079] (6) The abnormality diagnosis device 1 according to a sixth aspect is the abnormality diagnosis device 1 according to any one of (1), (2), and (5), in which the predetermined selection condition is a condition that the sensor is a sensor other than the one sensor and is a sensor 20-1 to 20-N that is predetermined for the one sensor. According to this aspect, the other sensor of the combination of two sensors can be narrowed down to the predetermined sensors 20-1 to 20-N.

[0080] (7) The abnormality diagnosis device 1 according to a seventh aspect is the abnormality diagnosis device 1 according to any one of (1) to (6), wherein the display processing unit 16 displays the positions of the state quantities specified by the scales of the first index and the second index in a different form for each type of the state quantities. According to this aspect, since the positions of the state quantities are displayed in a different form for each type of the state quantities on the display screen, it is possible to distinguish, for example, whether each of the state quantities is a unit space, a signal space, or a state quantity when an abnormality occurs on the display screen.

[0081] (8) The abnormality diagnosis device 1 according to an eighth aspect is the abnormality diagnosis device 1 according to any one of (1) to (7), wherein the first predetermined rank is the highest. According to this aspect, the one sensor selected based on the S / N ratio can be the sensor with the largest S / N ratio.

[0082] (9) The abnormality diagnosis device 1 according to a ninth aspect is the abnormality diagnosis device 1 according to any one of (1) to (8), wherein the second predetermined rank is the highest rank. According to this aspect, the combination of two sensors to be displayed on the display screen can be the combination with the maximum two-dimensional Mahalanobis distance. [Explanation of symbols]

[0083] 1. Abnormality diagnosis device 2 Display device 11 State quantity collection unit 12 State quantity memory unit 13 Anomaly detection section 14 SN ratio storage section 15 Sensor selection section 15-1 Sensor combination generation unit 15-2 Two-dimensional MD calculation section 16 Display processing section 20-1~20-N Sensor 30 Plant

Claims

1. an anomaly detection unit that detects an anomaly by applying the Mahalanobis-Taguchi method to state quantities detected by each of a plurality of sensors provided on the monitoring target, and calculates an S / N ratio corresponding to each of the sensors; a sensor selection unit that, when the anomaly detection unit detects an anomaly, selects a combination of two sensors from the plurality of sensors, the combination including, as one sensor, a sensor corresponding to an S / N ratio in a first predetermined order counting from the highest when the S / N ratios are arranged in descending order, and including, as the other sensor, a sensor other than the one sensor that satisfies a predetermined selection condition; a display processing unit that selects an index representing a state quantity of one of the sensors included in each of the combinations of the two sensors selected by the sensor selection unit as a first index and an index representing a state quantity of the other sensor included in each of the combinations of the two sensors selected by the sensor selection unit as a second index, and displays on a display screen a scale of the first index, a scale of the second index, and each position of the state quantities specified by the scales of the first index and the second index for each combination of the selected first index and the selected second index; An abnormality diagnosis device comprising:

2. The sensor selection unit a two-dimensional Mahalanobis distance is calculated from state quantities of the two sensors included in each of the selected combinations of the two sensors, and when the two-dimensional Mahalanobis distances are sorted in descending order, a combination of the two sensors is again selected from among the combinations corresponding to the two-dimensional Mahalanobis distances up to a second predetermined rank counting from the top when the two-dimensional Mahalanobis distances are sorted in descending order; The abnormality diagnosis device according to claim 1 .

3. The predetermined selection condition is: All sensors other than the one sensor are included. The abnormality diagnosis device according to claim 1 or 2.

4. The predetermined selection condition is: The condition is that the number of sensors is two or more, and the number of sensors is a predetermined number other than the one sensor. The abnormality diagnosis device according to claim 1 or 2.

5. The predetermined selection condition is: a condition that the sensor is a sensor other than the one sensor, and a degree of correlation between a state quantity of the one sensor and a state quantity of the sensor other than the one sensor shows a high degree of correlation in relation to a predetermined threshold value; The abnormality diagnosis device according to claim 1 or 2.

6. The predetermined selection condition is: a condition that the sensor is a sensor other than the one sensor and is a sensor that is predetermined with respect to the one sensor; The abnormality diagnosis device according to claim 1 or 2.

7. The display processing unit is When displaying the positions of the state quantities specified by the scales of the first index and the second index, the positions of the state quantities are displayed in different forms for each type of the state quantities. The abnormality diagnosis device according to claim 1 or 2.

8. The first predetermined rank is the highest rank. The abnormality diagnosis device according to claim 1 or 2.

9. The second predetermined rank is the highest rank. The abnormality diagnosis device according to claim 2 .

10. an anomaly detection step of detecting an anomaly by applying the Mahalanobis-Taguchi method to state quantities detected by each of a plurality of sensors provided on the monitoring target, and calculating an S / N ratio corresponding to each of the sensors; a sensor selection step of selecting, when an abnormality is detected by the abnormality detection step, a combination of two sensors selected from the plurality of sensors, the combination including, as one sensor, a sensor corresponding to an S / N ratio in a first predetermined order counting from the top when the S / N ratios are arranged in descending order, and including, as the other sensor, a sensor other than the one sensor that satisfies a predetermined selection condition; a display processing step of selecting an index representing a state quantity of one sensor included in each of the combinations of two sensors selected by the sensor selection step as a first index and an index representing a state quantity of the other sensor included in each of the combinations of two sensors selected by the sensor selection step as a second index, and displaying on a display screen a scale of the first index, a scale of the second index, and each position of the state quantities specified by the scales of the first index and the second index for each combination of the selected first index and the selected second index; An abnormality diagnosis method comprising:

11. Computer, an anomaly detection means for detecting an anomaly by applying the Mahalanobis-Taguchi method to state quantities detected by each of a plurality of sensors provided on the monitored object, and calculating an S / N ratio corresponding to each of the sensors; a sensor selection means for selecting a combination of two sensors from among the plurality of sensors when the abnormality detection means detects an abnormality, the combination including, as one sensor, a sensor corresponding to an S / N ratio in a first predetermined order counting from the highest when the S / N ratios are arranged in descending order, and including, as the other sensor, a sensor other than the one sensor that satisfies a predetermined selection condition; a display processing means for selecting an index representing a state quantity of one of the sensors included in each of the combinations of the two sensors selected by the sensor selection means as a first index and an index representing a state quantity of the other sensor as a second index, and for each combination of the first index and the second index selected, displaying on a display screen a scale of the first index, a scale of the second index, and each position of the state quantities specified by the scales of the first index and the second index; A program to function as a

Images