Failure sign monitoring system and failure sign monitoring method

Abstract

To provide a failure sign monitoring system and failure sign monitoring method making it possible to prevent detection oversight or erroneous detection at a time when a monitoring model created with an automatic update feature is used to monitor a sign of a failure of a plant.SOLUTION: A plant monitoring system is adapted to a plant including sensors A1 to Am, etc., and sensors B1 to Bn which measure the conditions of respective apparatuses, uses a monitoring model to monitor the plant, and regularly updates the monitoring model. The plant monitoring system includes a management server 43 that, before updating the monitoring model by using sensor values obtained during a present period, checks a variant of each of the sensor values, which are obtained during the present period, by using each of sensor values, which are obtained during a period preceding the present period, as a reference, and that uses the sensor values obtained by removing a sensor value whose variant is larger than a predesignated range, or the sensor values, which are obtained during another period, to update the monitoring model.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a failure sign monitoring system and a failure sign monitoring method for monitoring signs before a failure occurs, and more particularly to a failure sign monitoring system and a failure sign monitoring method for monitoring various plants. [Background technology]

[0002] Power plants in power plants use various types of equipment such as steam turbines and boilers to generate electricity. In such power plants, signs of failure are monitored for the installed equipment (see, for example, Patent Document 1). According to the monitoring system described in this document, System Invariant Analysis Technology (SIAT) is used to monitor for signs of failure. The invariant analysis technology utilizes a network fault response engine. The network fault response engine monitors the entire system by detecting disruptions (breaks) in the relationships between observation points.

[0003] This invariant analysis technology is applied to plant monitoring. Specifically, there is a strong correlation between the measurements of each sensor (hereinafter referred to as "sensor values") in a plant. Conventional monitoring systems automatically analyze sensor values ​​collected from observation points, comprehensively find and model correlations (invariants), which are invariant relationships between sensor values ​​between two points, and create a correlation model (hereinafter referred to as "monitoring model") for monitoring equipment, etc. After this, the monitoring system uses the monitoring model to examine the correlation between, for example, a sensor installed on a pipe in a power plant and a sensor installed on a steam turbine for any "unusual" behavior, that is, any disruption of the relationship. The monitoring system can then detect abnormalities in the power plant from the "unusual" behavior, enabling real-time monitoring of the entire plant. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2017-021702 Summary of the Invention [Problem to be solved by the invention]

[0005] The conventional monitoring system mentioned above has the following problem. Specifically, this monitoring system has an automatic update function that automatically updates the monitoring model periodically. However, when using the automatic update function, there is a concern of "missed detection" and "false positive detection." For example, there is a missed detection (Example 1) as shown in Figure 6. If the monitoring system evaluates the period "third week of month xx, year 20xx" using a monitoring model based on sensor values ​​from "second week of month xx, year 20xx," it is possible to detect "Trouble I." Note that in Figure 6, the sensor values ​​are represented by two curves, upper and lower. The lower curve g1 shows the sensor value when there is no trouble and the upper curve g2 shows the sensor value when a trouble, i.e., a failure, occurs.

[0006] However, if the monitoring system evaluates the period "4th week of xx month of 20xx year" using a monitoring model based on sensor values ​​from "3rd week of xx month of 20xx year," it will learn the sensor value fluctuations for "Trouble I" in the third week as normal, and may not be able to detect "Trouble II" which involves similar sensor value fluctuations.

[0007] There is also the possibility of false positives (Example 2) as shown in Figure 7. For example, even if the sensor value from a given sensor is constant due to maintenance in "week 3 of month xx, year 20xx," when the monitoring system creates a model and evaluates "week 4 of month xx, year 20xx," it will learn that the constant sensor value from the sensor is normal behavior. For this reason, the monitoring system may detect a status change in "week 4 of month xx, year 20xx" even when nothing is happening.

[0008] An object of the present invention is to solve the above-mentioned problems and to provide a failure sign monitoring system and a failure sign monitoring method that make it possible to prevent missed detections and false detections when monitoring for signs of plant failures using a monitoring model created with an automatic update function. [Means for solving the problem]

[0009] In order to solve the above problem, the invention of claim 1 is a plant monitoring system that is used for a plant that is equipped with various devices and various sensors that measure the states of these devices, monitors the plant using a monitoring model created based on correlations between the sensors, and periodically updates this monitoring model, and before updating the monitoring model using each sensor value for a current period, a variance ratio, which is a ratio between the variance calculated from each sensor value in the current period and the variance calculated from each sensor value in a period prior to the current period, is compared with a preset reference value, and if the variance ratio is greater than the reference value, sensor values ​​whose variance ratio is greater than the reference value are eliminated; The failure sign monitoring system is characterized by comprising a processing means for updating the monitoring model.

[0013] The invention of claim 2 is a failure sign monitoring method used for a plant equipped with various devices and various sensors for measuring the states of these devices, which monitors the plant using a monitoring model created based on correlations between the sensors and periodically updates the monitoring model, and before updating the monitoring model using each sensor value of a current period, a variance ratio, which is the ratio of variance calculated from each sensor value of the current period to variance calculated from each sensor value of a period prior to the current period, is compared with a preset reference value, and if the variance ratio is greater than the reference value, sensor values ​​whose variance ratio is greater than the reference value are excluded, and the monitoring model is updated periodically. Update The failure sign monitoring method is characterized by the following: [Effects of the Invention]

[0014] Claim 1 and 2According to the invention, when periodically updating a monitoring model, if the sensor values ​​for the current period have large variations, the monitoring model is created from the sensor values ​​obtained by removing the sensor values ​​with large variations or from the sensor values ​​obtained from another period. As a result, when monitoring for signs of failure in equipment installed in a plant in the next period using the updated monitoring model, the monitoring model created using sensor values ​​with large variations is not used, thereby preventing missed detections and false detections of signs of failure and preventing a decrease in monitoring accuracy.

[0015] Claim 1 and 2 According to the invention, by calculating the variance of the sensor values ​​in the current period, it becomes possible to check whether the variations in the sensor values ​​in the current period are large.

[0016] Claim 1 and 2 According to the invention, by calculating the variance ratio between the variance of the sensor values ​​in the previous period and the variance of the sensor values ​​in the current period, it is possible to check using a simple calculation method whether the variance of the sensor values ​​in the current period is large. [Brief explanation of the drawings]

[0018] [Figure 1] 1 is a configuration diagram showing a plant monitoring system according to a first embodiment of the present invention. [Figure 2] FIG. 10 is a diagram illustrating an example of sensor data. [Figure 3] FIG. 10 is a diagram illustrating an example of time-series sensor values. [Figure 4] 10 is a flowchart illustrating an example of a determination process. [Figure 5] FIG. 10 is a diagram illustrating a part of the calculation for calculating variance from sensor values. [Figure 6] FIG. 10 is an explanatory diagram illustrating an example of a detection failure. [Figure 7] FIG. 10 is an explanatory diagram illustrating an example of erroneous detection. DETAILED DESCRIPTION OF THE INVENTION

[0019] Next, each embodiment of the present invention will be described in detail with reference to the drawings.

[0020] (Embodiment 1) A plant monitoring system according to this embodiment is shown in Fig. 1. The plant monitoring system of Fig. 1 is used in a power plant where a power plant 10 is installed, and mainly comprises a sensor monitoring device 20 and a model monitoring device 40. The sensor monitoring device 20 is connected to the model monitoring device 40 via an in-house communication network 30 such as an intranet so that data communication with the model monitoring device 40 is possible.

[0021] The power plant 10 generates electricity using nuclear or thermal power, and this embodiment illustrates an example in which nuclear power is used. The power plant 10 uses a large number of devices, such as a nuclear reactor, a turbine, a generator, a pump, and piping, although these devices are not shown. The power plant 10 generates electricity using these devices. The power plant 10 is equipped with various sensors, such as A1 sensor ~ A m Sensor ... B1 sensor ~ B n Sensor Each sensor is installed. a1~a m ... b1~b n to the sensor monitoring device 20.

[0022] The sensor monitoring device 20 receives sensor values ​​a1 to a2 from the sensors installed in the power plant 10. m , ···, b1~b n The sensor monitoring device 20 receives the sensor values ​​a1 to a m , ···, b1~b nWhen receiving the sensor data, the sensor data shown in Fig. 2 is created. The sensor data indicates the sensor and records the installation point where the sensor is installed in correspondence with the sensor identification information for identifying the sensor. The sensor data also records the sensor value of the sensor, the measurement date and time, etc. in correspondence with the sensor identification information. In Fig. 2, the sensor values ​​a1 to a m A1 sensor to A2 are installed in the condenser of the power plant 10. m The sensor values ​​b1 to b n The sensors B1 to B are installed in the feedwater pump system of the power plant 10. n The following describes the configuration based on a sensor example.

[0023] The sensor monitoring device 20 transmits the created sensor data to the model monitoring device 40 via the in-house communication network 30 every time the transmission time elapses. The transmission time is set in the sensor monitoring device 20 as needed, such as in units of seconds, minutes, hours, or days.

[0024] The model monitoring device 40 receives sensor data from the sensor monitoring device 20 via the in-house communication network 30. The model monitoring device 40 uses the received sensor data to monitor the power plant 10. To this end, the model monitoring device 40 comprises a communication control unit 41, a data server 42, a management server 43, and clients 441 to 44. k The model monitoring device 40 includes a communication control unit 41, a data server 42, a management server 43, and clients 441 to 44. k are connected via a LAN (Local Area Network) or the like so that data can be sent and received.

[0025] The communication control unit 41 is connected to a data server 42, a management server 43, and clients 441 to 444. k to the in-house communication network 30. For example, when the communication control unit 41 receives sensor data from the sensor monitoring device 20 via the in-house communication network 30, it sends the sensor data to the data server .

[0026] The data server 42 is a storage device that stores data related to the power plant 10. For example, when the data server 42 receives sensor data from the sensor monitoring device 20 via the in-house communication network 30 and the communication control unit 41, the data server 42 stores the sensor data. Furthermore, when the data server 42 receives a data transmission request from the management server 43, the data server 42 extracts the corresponding sensor data and sends it to the management server 43.

[0027] Clients 441-44 k are computers operated by the person in charge of operating the power plant 10, and are used to operate the power plant 10. k In the data entry area 441, various instructions and the like required for the operation of the power plant 10 are input by a person in charge. For example, in order to create each monitored object model of the power plant 10 as needed, the clients 441 to 444 are k Model creation instructions are input to clients 441 to 444. k When a model creation instruction is input, the model creation instruction is sent to the management server 43.

[0028] Also, clients 441-44 k When receiving various data, such as alarm data described later, from the management server 43, the clients 441 to 444 output an alarm. k When the sensor list data is received from the management server 43, the display of this data is performed.

[0029] The management server 43 is a computer that performs various processes for monitoring the power plant 10 and each piece of equipment. First, the management server 43 creates a monitoring model of the power plant 10. Next, the management server 43 monitors the power plant 10 and the equipment based on the created monitoring model. Below, we will explain in order the creation of the monitoring model of the power plant 10 and the monitoring of the power plant 10 etc. using the monitoring model.

[0030] The management server 43 creates a monitoring model during normal operation in order to monitor the power plant 10. The monitoring model is obtained by deriving correlations between sensors using invariant analysis technology based on time-series data from each sensor attached to the power plant 10. The monitoring model during normal operation is created by the management server 43 based on sensor values ​​when the power plant 10 was operating normally. In other words, since the state of the power plant 10 is stable during constant-thermal operation, the distribution of sensor values ​​is also stable. The management server 43 creates the monitoring model based on this state.

[0031] Furthermore, in this embodiment, the management server 43 automatically updates the monitoring model at predetermined intervals, for example, once a week. The predetermined interval for automatically updating the monitoring model can be any interval, such as several days or several weeks. The management server 43 sends a data transmission request to the data server 42 at a preset timing. The management server 43 also sends a data transmission request to the clients 441 to 444. k When a model creation instruction is received from the data server 42 through manual input by a person in charge, the data server 42 also sends a data transmission request.

[0032] Thereafter, the management server 43 receives each sensor data from the data server 42. Each sensor data received by the management server 43 is data during normal operation of the power plant 10, and is, for example, as shown in FIG. 2. Then, the management server 43 creates a time-series sensor value list using the sensor values ​​obtained from each sensor data. An example of this time-series sensor value list is shown in FIG. 3. FIG. 3 shows a list of sensor values ​​divided into one-week intervals as a predetermined period, and for the sensor value a1 of the A1 sensor, the time-series sensor value a1(p1), , a1(p t ) a1(q1), , a1(q t ) a1(r1),..., a1(r t ) a1(s1), ···, a1(s t ) In other words, the time series sensor value for the first week of "20xx year xx month" is a1(p1), , a1(p t ) The time series sensor value for the second week is a1(q1), , a1(q t ) In addition, the time series sensor value for the third week is a1(r1),..., a1(r t ) The time series sensor value for the fourth week is a1(s1), ···, a1(s t ) Then, the time series sensor value a1(s t ) was the final value for this week.

[0033] When the management server 43 finishes creating such a list of time-series sensor values, it would have conventionally set the monitoring model to be used in the next week based on the sensor values ​​a1(s1), . . . , a1(s t ) is used to create the monitoring model. In other words, the current period, week 4, is used as the model learning period.

[0034] In contrast to this, this embodiment is configured as follows. That is, when the management server 43 finishes creating the list of time-series sensor values, it performs a judgment process. An example of this judgment process is shown in FIG. 4. The management server 43 performs a judgment process based on the last sensor value a1(s t ), the management server 43 starts the determination process before updating the surveillance model. When the determination process starts, the management server 43 receives the sensor values ​​a1(s1), . . . , a1(s t ) variance σ i That is, in step S1, the management server 43 calculates the sensor values ​​a1(s1), . . . , a1(s t) is stable. For this purpose, the management server 43 checks whether the distribution of the sensor values ​​a1(s1), . . . , a1(s t ) average value a 1av is calculated using the following formula:

[0035]

number

[0036] Next, the management server 43 calculates the sensor values ​​a1(s1), . . . , a1(s t ) for each of the average values ​​a 1av After that, the management server 43 calculates the difference between the sensor values ​​a1(s1), . . . , a1(s t ) and the average value a 1av Next, the management server 43 calculates the square of the difference between the average value a 1av Furthermore, the management server 43 divides the calculated sum by the value t to obtain the variance σ of this week (fourth week), which is the model learning period. i Calculate.

[0037]

number

[0038] The value t here is the sensor value a1(s1),...,a1(s t ) is the number of data points t.

[0039] When step S1 is completed, the management server 43 similarly calculates the variance σ i-1 , a1(r1), , a1(r t ) is used to calculate (step S2).

[0040] When step S2 is completed, the management server 43 performs an F-test on the distribution of the sensor values ​​(step S3). That is, the management server 43 calculates the variance σ i and the variance of the previous period (week 3) σi-1 The variance ratio F i Calculate the variance ratio F i is a value used to determine whether the model learning period (week 4) and the previous period (week 3) have equal variances. For example,

[0041]

number

[0042] Then, the management server 43

[0043]

number

[0044] The variance ratio F i Conversely,

[0045]

number

[0046] Then, the management server 43

[0047]

number

[0048] The variance ratio F i Calculate.

[0049] In order to perform the F test, the management server 43 uses a predetermined reference value F s The judgment criteria value F s is a value that serves as a criterion for the F test, and is calculated based on two periods when the power plant is normal, for example, the variance of the first week and the variance of the second week. Also, the variance ratio F is a value input to the management server 43 by the person in charge operating the client 441, for example, that is, a value set according to the characteristics of the sensors of the power plant. i This is the standard value for judgment.

[0050] The management server 43 calculates the variance ratio F i Regarding

[0051]

number

[0052] If the relationship holds, the variance σ for this week (week 4), which is the model learning period, i and the variance of the previous week (week 3) i-1 In other words, the management server 43 determines in step S3 that the sensor values ​​during the model learning period are the same as the a1(s1), ···, a1(s t ) Variation in the sensor value of the previous period a1(r1),..., a1(r t ) It is determined that the variation in the

[0053] After the determination in step S3, the management server 43 calculates the sensor values ​​a1(s1), . . . , a1(s t ) to create an optimal monitoring model (step S4).

[0054] In step S4, the management server 43 calculates the sensor values ​​a1(s1), . . . , a1(s t ) to create a monitoring model. That is, the management server 43 checks the strength of the correlation between sensors and creates a monitoring model using correlated sensors. At this time, the management server 43 creates in advance a monitoring model that focuses on the strength of the correlation between sensors according to the plant state (under inspection, starting up, operating at rated thermal output, undergoing regular testing, or shut down).

[0055] On the other hand, the management server 43 uses the variance ratio F calculated in step S3 i Regarding

[0056]

number

[0057] If the relationship does not hold, for example, clients 441 to 444 k In step S5, the management server 43 generates alarm data informing the clients 441 to 44 that the distribution of the sensor values ​​is unstable as an abnormality, and transmits the alarm data to the clients 441 to 444. k Send to.

[0058] After step S5, the management server 43 changes the model learning period or the variance ratio F i The management server 43 then removes sensor values ​​that are greater than the set reference value from the model learning period, corrects or changes the model learning period, and creates a monitoring model based on the corrected or changed model learning period (step S6). For example, as shown in FIG. 6, if the sensor values ​​fluctuate during the model learning period due to a problem, or as shown in FIG. 7, if the sensor values ​​become constant during the model learning period due to maintenance, the management server 43 uses the variance ratio F i This allows the management server 43 to correct and change the model learning period, so that in step S6, when updating the monitoring model, the management server 43 can create an optimal monitoring model equivalent to that when the distribution of sensor values ​​is stable.

[0059] When step S4 or step S6 is completed, the management server 43 completes the determination process. Then, the management server 43 performs real-time monitoring or monitoring by the core invariant method using the optimal monitoring model for the next week.

[0060] The above is the configuration of the failure sign monitoring system according to this embodiment. Next, a failure sign monitoring method using this failure sign monitoring system will be described.

[0061] Typically, the model monitoring device 40 monitors the state of the power plant 10 using a monitoring model. The management server 43 of the model monitoring device 40 automatically updates the monitoring model every model learning period, which is a predetermined period, for example, every week. At this time, the management server 43 determines whether there is a variation in the sensor value for this week, i.e., the fourth week, compared with the sensor value for the third week, using the determination process shown in FIG. 4.

[0062] If there is no variation between the sensor values ​​of the fourth week and the sensor values ​​of the third week, the management server 43 creates a monitoring model using the sensor values ​​of the fourth week and updates the monitoring model.

[0063] If there is a deviation in the sensor value of the fourth week compared to the sensor value of the third week, the management server 43 corrects or changes the model learning period, creates a monitoring model based on the sensor value of the corrected or changed model learning period, and updates the monitoring model. At the same time, the management server 43 creates alarm data and sends it to the clients 441 to 444. k Send to clients 441-44 k When alarm data is received, the system notifies the person in charge that the distribution of sensor values ​​during the model learning period is not stable.

[0064] Then, in the next week, the management server 43 of the model monitoring device 40 performs real-time monitoring or monitoring of the power plant 10 by the core invariant method using the updated monitoring model.

[0065] Thus, according to this embodiment, a function for performing a test (F test) based on the variance ratio is added to the sensor value distribution. That is, a filter for the test based on the variance ratio (F test) is applied to the sensor value distribution. As a result, if the distribution of the sensor values ​​is not stable, it is possible to change the model learning period or the variance ratio F i This eliminates sensor values ​​that are greater than the set reference value from the model learning period, making it possible to prevent "missed detections" and "false positive detections" when the monitoring model is automatically updated. This prevents a decrease in monitoring accuracy.

[0066] (Embodiment 2) In this embodiment, the variation in each sensor value for the current period is determined as follows: In this embodiment, components that are the same as or considered to be the same as those in the first embodiment described above are given the same reference numerals, and their description will be omitted.

[0067] In this embodiment, the determination process (FIG. 4) is simplified. That is, the management server 43 determines the variance σ of this week (fourth week), which is the model learning period, in the same manner as in step S1 of the determination process (FIG. 4). i Let a1(s1),...,a1(s t ) is calculated.

[0068] The management server 43 stores in advance a variance reference value that is calculated from each variance in past automatic updates and represents the normal state of the sensor value. The management server 43 then compares this variance reference value with the variance σ for this week (week 4). i and compare the sensor values ​​a1(s1), , a1(s t ) is stable.

[0069] Thereafter, the management server 43 performs step S4 or steps S5 and S6 similar to the determination process (FIG. 4).

[0070] Thus, according to this embodiment, by adding processing for simplified judgment based on variance to the sensor value distribution, it is possible to prevent "missed detections" and "false detections" when automatically updating the monitoring model, thereby preventing a decrease in monitoring accuracy. [Explanation of symbols]

[0071] 10 Power Plant 20 Sensor monitoring device 40 Model Surveillance Device 41 Communication control unit 42 Data Server 43 Management Server (Processing Means) 441~44k client

Claims

1. A plant monitoring system that is used for a plant that is equipped with various devices and various sensors that measure the states of the devices, monitors the plant using a monitoring model created based on correlations between the sensors, and periodically updates the monitoring model, a processing means for comparing a variance ratio, which is a ratio between a variance calculated from each sensor value in the current period and a variance calculated from each sensor value in a period prior to the current period, with a preset reference value before updating the surveillance model using each sensor value in the current period, and, if the variance ratio is greater than the reference value, excluding sensor values ​​whose variance ratio is greater than the reference value and updating the surveillance model; A failure sign monitoring system comprising:

2. A failure sign monitoring method is used for a plant that is equipped with various devices and various sensors that measure the states of the devices, the method monitors the plant using a monitoring model created based on correlations between the sensors, and periodically updates the monitoring model, before updating the surveillance model using each sensor value for the current period, a variance ratio, which is a ratio between a variance calculated from each sensor value for the current period and a variance calculated from each sensor value for a period prior to the current period, is compared with a preset reference value, and if the variance ratio is greater than the reference value, sensor values ​​whose variance ratio is greater than the reference value are excluded and the surveillance model is updated; A failure sign monitoring method characterized by:

Images