JP7453049B2 - Abnormal sign monitoring system, abnormal sign monitoring method, and program - Google Patents

Description

The present invention relates to an abnormality sign monitoring system and the like.

As a technique for estimating the cause of a sign of abnormality in plant equipment, for example, the technique described in Patent Document 1 is known. That is, Patent Document 1 discloses an abnormality parameter estimating unit that acquires measured values of a plant, compares them with normal measured values stored in a normal measured value database, and estimates abnormal parameters, and a physical model. A device status monitoring system is described that includes a faulty device estimator that estimates a faulty device from abnormal parameters.

Japanese Patent Application Publication No. 2020-9080

For example, in a power generation plant, multiple state quantities such as pressure, temperature, and flow rate of equipment are detected, but these state quantities are mainly used for the operation of the power plant and are used to monitor equipment for signs of abnormality. is often not sufficient. To give a specific example, in a pump installed in a power generation plant, the pressure, temperature, and flow rate of fluid in the pump affect the operation of the power generation plant, so detected values thereof are often acquired.

On the other hand, since pump vibration does not particularly affect the operation of a power generation plant, a vibration sensor is often not installed on the pump. If such an undetected state quantity (for example, displacement of vibration of a pump) exists, it may not be possible to sufficiently narrow down the cause of a sign of abnormality in a device such as a pump. Patent Document 1 does not describe a technique for presenting a result of estimating the cause of a device abnormality sign using an undetected state quantity in order to narrow down the cause of the device abnormality sign.

Therefore, it is an object of the present invention to provide an abnormality sign monitoring system and the like that uses undetected state quantities to present the results of estimating the causes of abnormality signs in equipment.

In order to solve the above-mentioned problems, the present invention includes a detected value acquisition unit that acquires detected values of a plurality of state quantities related to plant equipment from a plurality of sensors, and a detection value acquisition unit that acquires detected values of a plurality of state quantities regarding plant equipment from a plurality of sensors, If the detection values of the plurality of sensors acquired by the detection value acquisition unit include an abnormal detection value based on predetermined normal data including the detection values of the plurality of sensors, the abnormal detection value is a first abnormality cause estimation unit that estimates the cause of the abnormality sign of the device based on a predetermined correspondence between the state quantity detected as a detection value and the cause of the abnormality sign of the device; In addition to the detected value, the cause of the abnormality sign of the device when it is assumed that a predetermined undetected state quantity that is not detected in the device is abnormal, and/or the undetected state quantity is normal. a second abnormality cause estimating unit that estimates the cause of the abnormality sign of the device when it is assumed that an additional effect calculation unit that calculates an additional effect when assuming that The present invention is characterized by comprising a display control section that causes a display means to display, as the additional effect, the difference between the cause of the abnormality sign of the device and the undetected state amount.

According to the present invention, it is possible to provide an abnormality sign monitoring system and the like that presents a cause estimation result of a device abnormality sign using undetected state quantities.

FIG. 1 is a functional block diagram of the abnormality sign monitoring system according to the first embodiment. FIG. 2 is an explanatory diagram showing an example of time-series data regarding a pump in the abnormality sign monitoring system according to the first embodiment. FIG. 2 is an explanatory diagram showing an example of the degree of abnormality of the detected value of each sensor in the abnormality sign monitoring system according to the first embodiment. FIG. 2 is an explanatory diagram showing an example of an abnormality detection result by a parameter abnormality detection unit of the abnormality sign monitoring system according to the first embodiment. FIG. 2 is an explanatory diagram showing an example of a physical model database of the abnormality sign monitoring system according to the first embodiment. FIG. 6 is an explanatory diagram when the detected value of vibration of the pump is simulated as being abnormal by the parameter state simulating unit of the abnormality sign monitoring system according to the first embodiment. FIG. 7 is an explanatory diagram of data in which a result of simulation by a parameter state simulator is incorporated into a result of detection by a parameter abnormality detector in the abnormality sign monitoring system according to the first embodiment. It is an example of a screen display of the diagnostic result output device of the abnormality sign monitoring system according to the first embodiment. 3 is a flowchart of processing executed by the abnormality sign monitoring system according to the first embodiment. FIG. 7 is an explanatory diagram when the detected value of the vibration of the pump is simulated as being normal by the parameter state simulating section of the abnormality sign monitoring system according to the second embodiment. In the abnormality sign monitoring system according to the second embodiment, it is an explanatory diagram of data in which the results of simulation by the parameter state simulation unit are incorporated into the detection results by the parameter abnormality detection unit. It is an example of a screen display of the diagnostic result output device of the abnormality sign monitoring system according to the second embodiment. FIG. 7 is an explanatory diagram showing an example of a physical model database of an abnormality sign monitoring system according to a fourth embodiment. It is an example of a screen display of the diagnostic result output device of the abnormality sign monitoring system according to the fourth embodiment.

≪First embodiment≫
FIG. 1 is a functional block diagram of an abnormality sign monitoring system 100 according to the first embodiment.
The abnormality sign monitoring system 100 is a system that monitors the presence or absence of abnormality signs in equipment (for example, the pump 21) of the plant 200. Here, the "abnormality sign" of the device is a harbinger of an abnormality occurring in the device. In addition, the types of plants 200 include, but are not limited to, power plants (such as nuclear power plants), chemical plants, manufacturing plants, and water treatment plants.

Note that the abnormality sign monitoring system 100 may be configured with one device, or may be configured with a plurality of devices (servers, etc.) connected in a predetermined manner via a signal line or a network. In the following, prior to explaining the abnormality sign monitoring system 100, the pump 21, which is one of the objects to be monitored for abnormality signs, will be explained, and each sensor that detects the state quantity related to the pump 21 will be explained.

The pump 21 shown in FIG. 1 is a device that pumps a predetermined fluid, and is provided in the plant 200. Although not shown, the plant 200 (for example, a power generation plant) is provided with various devices such as a turbine, a generator, and a condenser.

In the example of FIG. 1, a pressure sensor 31, a temperature sensor 32, and a flow rate sensor 33 are provided as sensors for detecting a plurality of state quantities related to the pump 21. The pressure sensor 31 is a sensor that detects the pressure of fluid discharged from the pump 21, and is provided in the piping k1 on the discharge side of the pump 21. The temperature sensor 32 is a sensor that detects the temperature of the fluid discharged from the pump 21, and is provided in the pipe k1. The flow rate sensor 33 is a sensor that detects the flow rate of fluid discharged from the pump 21, and is provided in the pipe k1.

The detected value of the pressure sensor 31 is outputted to the detected value acquisition device 1 (detected value acquisition unit), which will be described next, via the wiring w1. Similarly, the detected values of the temperature sensor 32 and the flow rate sensor 33 are also output to the detected value acquisition device 1 via the wiring w2 and w3. Note that instead of (or together with) the discharge side of the pump 21, a predetermined sensor (not shown) may be provided in the piping k2 on the suction side of the pump 21.

<Configuration of abnormality sign monitoring system>
As shown in FIG. 1, the abnormality sign monitoring system 100 includes a detected value acquisition device 1, a time series data storage section 2, a normal time database 3, a parameter abnormality detection section 4, a physical model database 5, and a first An abnormality cause estimation unit 6 is provided. In addition to the above-described configuration, the abnormality sign monitoring system 100 also includes a parameter state simulating section 7, a second abnormality cause estimation section 8, an additional effect calculation section 9, a display control section 10, and a diagnostic result output device 11. (display means).

The detected value acquisition device 1 is a device that acquires detected values of a plurality of state quantities (pressure, temperature, flow rate) regarding the pump 21 (equipment of the plant 200) from a pressure sensor 31, a temperature sensor 32, and a flow rate sensor 33. Then, the detected value acquisition device 1 associates the momentary detected values of the pressure sensor 31, temperature sensor 32, and flow rate sensor 33 with the date and time when the detected values were acquired, and the identification information of each sensor. The predetermined time-series data is stored in the time-series data storage section 2. Note that, as the detected value acquisition device 1, for example, a predetermined process computer is used.
The time series data storage unit 2 stores predetermined time series data acquired by the detected value acquisition device 1. As such a time series data storage section 2, a magnetic disk device, an optical disk device, a semiconductor storage device, etc. are used.

FIG. 2 is an explanatory diagram showing an example of time-series data regarding the pump (see also FIG. 1 as appropriate).
“Date and time” shown in FIG. 2 is the date and time when the detected value acquisition device 1 acquired the detected values of the pressure sensor 31, temperature sensor 32, and flow rate sensor 33. In FIG. 2, only the date is shown in the "Date and Time" column for simplification, but in reality, the time series data includes the date and time when each detected value was acquired.

For example, on August 1, 2000, the pressure on the discharge side of the pump 21 was 7.21 [MPa], the temperature was 270.1 [°C], and the flow rate was 500 [kg/s]. is one of the detected values acquired at a predetermined time on August 1, 2000. Actually, on August 1, 2000, each detection value is acquired at predetermined time intervals (every 0.1 seconds, every few seconds, every few minutes, etc.) and stored in the time-series data storage section 2. The same applies to each detection value after August 2, 2000.

Further, each detected value of "pressure" shown in FIG. 2 is associated with identification information of the pressure sensor 31. Similarly, each detected value of "temperature" is associated with identification information of the temperature sensor 32, and each detected value of "flow rate" is associated with identification information of the flow rate sensor 33.

The normal database 3 shown in FIG. 1 includes momentary detection values of the pressure sensor 31, temperature sensor 32, and flow rate sensor 33, the date and time when the detection values were acquired, and identification information of each sensor. The associated time series data is stored as predetermined normal data. This normal data includes detection values of each sensor acquired during a period in which it is known that the pump 21 is normal.

The parameter abnormality detection unit 4 detects the detected values of the pressure sensor 31, temperature sensor 32, and flow rate sensor 33 (stored in the time-series data storage unit 2) acquired by the detected value acquisition device 1, based on the normal data described above. It is determined whether or not the time-series data (currently available time-series data) contains an abnormal detected value. Note that the word "parameter" is synonymous with "state quantity" such as pressure, temperature, and flow rate. It is also assumed that predetermined normal data is stored in the normal state database 3 in advance of the start of processing in the parameter abnormality detection section 4.

The parameter abnormality detection unit 4 reads predetermined time series data (see FIG. 2) from the time series data storage unit 2 and also reads normal data from the normal time database 3. Then, the degree of abnormality of the time-series data in the time-series data storage section 2 is calculated based on the distribution of normal data. This "degree of abnormality" is a numerical value indicating the degree of abnormality of each detected value of the pressure sensor 31, temperature sensor 32, and flow rate sensor 33.

The parameter abnormality detection unit 4 calculates, for example, a Mahalanobis distance based on normal data for each detected value (see FIG. 2), and uses this Mahalanobis distance as the degree of abnormality. To give a specific example, in the time series data of FIG. 2, the detected value of the pressure sensor 31 on August 1, 2000 is 7.21 [MPa]. The parameter abnormality detection unit 4 calculates the Mahalanobis distance for this 7.21 [MPa] based on the normal data of the detected value of the pressure sensor 31, and uses this Mahalanobis distance as the degree of pressure abnormality.

Note that the same applies to each detection value of the pressure sensor 31 at other dates and times, and also to each detection value of the temperature sensor 32 and the flow rate sensor 33. That is, in the example of this embodiment, the parameter abnormality detection unit 4 calculates the degree of abnormality individually for each of the three state quantities (pressure, temperature, and flow rate).

If the degree of abnormality of the detected value of a predetermined sensor (for example, the pressure sensor 31) is less than a predetermined threshold, the parameter abnormality detection unit 4 determines that the detected value of that sensor is normal. On the other hand, when the degree of abnormality of the detected value of a predetermined sensor (for example, pressure sensor 31) is equal to or greater than the predetermined threshold, the parameter abnormality detection unit 4 determines that the detected value of that sensor is abnormal. The above-mentioned predetermined threshold value is a threshold value that serves as a criterion for determining whether or not the detection value of each sensor is abnormal, and is set in advance. In this embodiment, as an example, a case will be described in which the magnitude of the predetermined threshold value regarding the degree of abnormality is 1.0.

FIG. 3 is an explanatory diagram showing an example of the degree of abnormality of the detected value of each sensor.
Note that each cell in FIG. 3 corresponds to each cell in FIG. 2 regarding time series data. Furthermore, in the example of FIG. 3, as indicated by frame line G1, the degree of temperature abnormality on August 4, August 5, and August 6, 2000 is 1.0 (predetermined threshold) or higher. . This is because the temperature detection values on these dates were significantly higher than the normal distribution (see Figure 2). Although it is simplified in FIG. 3, for example, after the temperature abnormality exceeds 1.0 from a predetermined time on August 4, 2000, for a predetermined period of time (in the example of FIG. 3, on August 5, 2000) , August 6th, 2017).

On the other hand, in the example of FIG. 3, regarding the pressure and flow rate on the discharge side of the pump 21, the degree of abnormality at each date and time is all less than 1.0. Furthermore, during the period from August 1 to August 3, 2000, the abnormality degree of the temperature on the discharge side of the pump 21 was also less than 1.0.

FIG. 4 is an explanatory diagram showing an example of an abnormality detection result by the parameter abnormality detection section.
Note that each cell in FIG. 4 corresponds to each cell in FIG. 2 regarding time series data and each cell in FIG. 3 regarding abnormality degree. Further, in the example of FIG. 4, as indicated by a frame line G2, an abnormal temperature on the discharge side of the pump 21 was detected on August 4, 2000, August 5, and August 6, 2000.

In this way, when the detection value of the pump 21 includes an abnormal detection value (see frame line G2 in FIG. 4), the parameter abnormality detection unit 4 detects the first abnormality cause estimation unit 6 and the second abnormality cause. Predetermined data is output to the estimator 8. The predetermined data described above includes identification information (corresponding to the state quantity) of the sensor from which the abnormal detection value was obtained, and the date and time when the abnormal detection value was obtained.

The physical model database 5 shown in FIG. 1 is used to estimate the cause of a sign of abnormality in the pump 21 based on the state amount when there is an abnormality in a predetermined state amount (for example, "temperature" shown in FIG. 4). used for.

FIG. 5 is an explanatory diagram showing an example of the physical model database 5 (see also FIG. 1 as appropriate).
In the example of FIG. 5, cases where the detected value of a predetermined sensor is abnormal are indicated by a circle, while cases where the detected value of the sensor is normal are left blank. The physical model database 5 is a database that shows the correspondence between the state quantity (marked with a circle) in which an abnormality has been detected and the cause of the abnormality sign of the pump 21. It is set. Further, a predetermined physical model database 5 is set in advance for each device included in the plant 200.

For example, in the plant 200 shown in FIG. 1, while pressure, temperature, and flow rate on the discharge side of the pump 21 are detected, vibrations of the pump 21 (see also FIG. 5) are not particularly detected. This is because even if vibrations of the pump 21 are not detected, there is no particular problem in the operation of the plant 200. Therefore, the pump 21 is often not equipped with a vibration sensor (not shown). However, if some abnormality sign occurs in the pump 21, the cause of the abnormality sign may be narrowed down by detecting the vibration of the pump 21.

Therefore, in the first embodiment, regarding the pump 21, the cause estimation result of the abnormality sign based on the presence or absence of abnormality in pressure, temperature, and flow rate (see FIG. 4), and the result of estimating the cause of the abnormality sign based on the presence or absence of abnormality in the pressure, temperature, and flow rate, and furthermore, the The differences between the result of estimating the cause of the abnormality sign and the result are presented to the user. This is one of the main features of the first embodiment.

The first abnormality cause estimating unit 6 shown in FIG. If so, the cause of the abnormality sign of the pump 21 is determined based on a predetermined correspondence relationship (see FIG. 5) between the state quantity detected as an abnormal detection value and the cause of the abnormality sign of the pump 21 (equipment). Estimate.

In the example of FIG. 4, the "temperature" on the discharge side of the pump 21 has become abnormal after August 4, 2000. As described above, an abnormality in "temperature" is detected by the parameter abnormality detection section 4. When an abnormality in a predetermined state quantity is detected in this way, the first abnormality cause estimation unit 6 refers to the physical model database 5 (see FIG. 5) and estimates the cause of the abnormality sign of the pump 21. Note that the "vibration" of the pump 21 (see FIG. 5), which is an undetected parameter (undetected state quantity), is not particularly used in the processing of the first abnormality cause estimation unit 6.

For example, if the "temperature" on the discharge side of the pump 21 is abnormal, but the "pressure" and "flow rate" are normal (after August 4, 2000 in FIG. 4), the first abnormality cause estimation unit 6 , perform the following processing. That is, the first abnormality cause estimation unit 6 estimates that the cause of the abnormality sign of the pump 21 is "pump bearing wear" or "pump seal wear" based on the physical model database 5 (see FIG. 5). The first abnormality cause estimation unit 6 outputs the estimation result of the cause of the abnormality sign of the pump 21 to the additional effect calculation unit 9. Note that the data output from the first abnormality cause estimation unit 6 to the additional effect calculation unit 9 includes, in addition to the cause of the abnormality sign of the pump 21, the state quantity (sensor identification) that is the basis for estimating the cause of the abnormality sign. information) is also included.

The parameter state simulation unit 7 shown in FIG. 1 has a function of assuming that a predetermined undetected state parameter (for example, "vibration" of the pump 21: see FIG. 5) that is not detected in the pump 21 is abnormal. There is. Note that while the normal state database 3 (see FIG. 5) includes the "vibration" of the pump 21, in reality, the detected value of the "vibration" of the pump 21 has not been obtained. "Vibration" is identified as an undetected parameter.

FIG. 6A is an explanatory diagram when the parameter state simulator simulates that the detected value of vibration of the pump is abnormal (see also FIG. 1 as appropriate).
In the example of FIG. 6A, during the period from August 1 to August 3, 2000, when the pressure, temperature, and flow rate on the discharge side of the pump 21 were all normal (see FIG. 4), the parameter state simulation unit 7 , it is simulated that the vibration of the pump 21, which is an undetected parameter, is normal. In this way, when the actually detected values of pressure, temperature, and flow rate are all normal, it is simulated that the vibration of the pump 21 is also normal.

On the other hand, after August 4, 2020, when an abnormality in the temperature of the pump 21 was detected (see FIG. 4), the parameter state simulation unit 7 simulates an abnormality in the vibration of the pump 21, which is an undetected parameter. In this way, when the parameter abnormality detection section 4 detects an abnormality in a predetermined state quantity (for example, the temperature of the pump 21), the parameter state simulating section 7 simulates an abnormality in the vibration of the pump 21. Good too. Note that the format of the data generated by the parameter state simulator 7 is similar to the format of the data generated by the parameter abnormality detector 4 (see FIG. 4).

Incidentally, when the temperature etc. of the pump 21 is abnormal, the parameter state simulation unit 7 either simulates the undetected parameter as "abnormal" (first embodiment) or simulates the undetected parameter as "normal". Whether to do so (second embodiment) is set in advance.

In addition to the abnormal detected value (in the example of FIG. 4, the temperature on the discharge side of the pump 21), the second abnormality cause estimation unit 8 shown in FIG. The cause of the abnormality sign of the pump 21 is estimated when a parameter (undetected state quantity: in the example of FIG. 6A, "vibration" of the pump 21) is assumed to be abnormal.

FIG. 6B is an explanatory diagram of data in which the results of simulation by the parameter state simulator are incorporated into the results of detection by the parameter abnormality detector.
The normality/abnormality of pressure, temperature, and flow rate shown in FIG. 6B is a detection result by the parameter abnormality detection unit 4, and is the same as in FIG. 4. On the other hand, the normality/abnormality of vibration shown in FIG. 6B is the result of simulation by the parameter state simulator 7, and is the same as that in FIG. 6A. In this way, the second abnormality cause estimating unit 8 (see FIG. 1) determines whether the pump Estimating the causes of 21 abnormal signs.

In the example of FIG. 6B, after August 4, 2020, as shown by frame lines G3 and G4, there is an abnormality in the temperature on the discharge side of the pump 21 and the vibration of the pump 21 (the vibration is an assumption) . The second abnormality cause estimation unit 8 refers to the physical model database 5 (see FIG. 5) and estimates that the cause of the abnormality sign of the pump 21 is "pump bearing wear." Then, the second abnormality cause estimation unit 8 outputs the estimation result of the cause of the abnormality sign of the pump 21 to the additional effect calculation unit 9. Note that the data output from the second abnormality cause estimation unit 8 to the additional effect calculation unit 9 includes the cause of the abnormality sign of the pump 21 as well as the type of undetected parameter that is the basis for estimating the cause of the abnormality sign. Contains information indicating.

The additional effect calculation unit 9 calculates the cause of the abnormality sign of the pump 21 (equipment) estimated by the first abnormality cause estimation unit 6, the cause of the abnormality sign of the pump 21 estimated by the second abnormality cause estimation unit 8, Compare. Then, the additional effect calculation unit 9 calculates the difference between the estimation result of the first abnormality cause estimation unit 6 and the estimation result of the second abnormality cause estimation unit 8 by using an undetected parameter (for example, “vibration” of the pump 21). ) is identified as an additional effect if detected. Here, the term "additional effect" means that the cause of the sign of abnormality in the pump 21 can be further narrowed down when it is assumed that an undetected parameter is detected.

In the example described above, the first abnormality cause estimation unit 6 estimates that the cause of the abnormality sign of the pump 21 is "pump bearing wear" or "pump seal wear" (see FIGS. 4 and 5). On the other hand, based on the assumption that the vibration (undetected parameter) of the pump 21 is abnormal, the second abnormality cause estimation unit 8 estimates that the cause of the abnormality sign of the pump 21 is "pump bearing wear" ( (See Figures 5 and 6B). The additional effect calculation unit 9 calculates the additional effect when the undetected parameter is added, for example, by subtracting the estimation result of the second abnormality cause estimation unit 8 from the estimation result of the first abnormality cause estimation unit 6. Then, the additional effect calculation section 9 outputs the calculation result of the additional effect to the display control section 10.

Note that the calculation result of the additional effect includes, in addition to the additional effect related to narrowing down the cause of the abnormal sign, a state quantity (for example, vibration of the pump 21) corresponding to this additional effect. Furthermore, in addition to the additional effects described above, the estimation results of the first abnormality cause estimation section 6 and the second abnormality cause estimation section 8 may also be output to the display control section 10.

The display control unit 10 outputs the calculation results of the additional effect calculation unit 9 to the diagnosis result output device 11, and controls the display of the diagnosis result output device 11 in a predetermined manner. That is, the display control unit 10 displays the cause of the abnormality sign of the pump 21 (equipment) estimated by the first abnormality cause estimation unit 6 and the cause of the abnormality sign of the pump 21 estimated by the second abnormality cause estimation unit 8. , are displayed on the diagnostic result output device 11 (display means) together with undetected parameters (undetected state quantities: for example, vibration of the pump 21).

The diagnostic result output device 11 is, for example, a display, and displays the calculation results of the additional effect calculation section 9 in a predetermined manner under the control of the display control section 10. In addition to displaying the calculation result of the additional effect calculation section 9, the diagnosis result output device 11 may also output a predetermined sound.

FIG. 7 is an example of a screen display of the diagnosis result output device.
In the example of FIG. 7, the diagnosis result output device 11 displays that an abnormality was detected in the temperature on the discharge side of the pump 21 on August 4, 2000. Based on the detection values of the sensors actually installed, "pump bearing wear" and "pump seal wear" are listed as the estimated causes of the signs of abnormality in the pump 21.

Further, in the example of FIG. 7, if it is confirmed that the "vibration" of the pump 21 is abnormal, it is displayed that the cause of the sign of abnormality of the pump 21 can be narrowed down to "pump bearing wear." This part is an additional effect obtained by detecting the "vibration" (undetected state quantity) of the pump 21. As a result, the user can understand that if a vibration sensor (not shown) is newly installed in the pump 21, there is a high possibility that the cause of the sign of abnormality in the pump 21 can be narrowed down in the future.

<Processing in the abnormality sign monitoring system>
FIG. 8 is a flowchart of processing executed by the abnormality sign monitoring system (see also FIG. 1 as appropriate).
In step S101, normal data is acquired and stored. That is, the detected value of each sensor during a predetermined period during which it is known that each device including the pump 21 is operating normally is acquired by the detected value acquisition device 1, and is further stored in the time series data storage unit 2. be remembered.

In step S102, the detection values of each sensor are acquired and stored (detection value acquisition step). That is, the detected values of each sensor including the pressure sensor 31, temperature sensor 32, and flow rate sensor 33 are acquired by the detected value acquisition device 1, and further stored in the time series data storage section 2.

In step S103, it is determined whether an abnormality in a parameter (state quantity) is detected. That is, the parameter abnormality detection unit 4 determines whether or not the detected values of the pressure sensor 31, temperature sensor 32, and flow rate sensor 33 include abnormal detected values. If no parameter abnormality is detected in step S103 (S103: No), the process in the abnormality sign monitoring system 100 returns to step S102. On the other hand, if an abnormality in the parameters is detected in step S103 (S103: Yes), the process in the abnormality sign monitoring system 100 proceeds to step S104.

In step S104, it is determined that there is a sign of abnormality in the device. For example, the parameter abnormality detection unit 4 determines that there is a sign of abnormality in the pump 21, which is one of the devices.
Next, in step S105, the cause of the abnormality sign is estimated (first abnormality cause estimation step). That is, the first abnormality cause estimation unit 6 refers to the physical model database 5 and estimates the cause of the abnormality sign of the pump 21 based on the detection result of the parameter abnormality detection unit 4.

In step S106, an abnormality in an undetected parameter (undetected state quantity) is simulated. For example, if the vibration of the pump 21 is an undetected parameter, the parameter state simulator 7 simulates (assumes) that the vibration of the pump 21 is abnormal.
Next, in step S107, after simulating the abnormality of the undetected parameter, the cause of the abnormality sign is estimated (second abnormality cause estimation step). That is, the second abnormality cause estimation unit 8 refers to the physical model database 5 and determines the cause of the abnormality sign of the pump 21 based on the detection result of the parameter abnormality detection unit 4 and the simulation result of the abnormality of the undetected parameter. presume.

In step S108, the additional effect of the undetected parameter is calculated. For example, when the vibration of the pump 21 is added as an undetected parameter, the additional effect calculation unit 9 calculates how the cause of the abnormality sign of the pump 21 can be specifically narrowed down (that is, the additional effect).
In step S109, the diagnosis result is output (display control step). That is, the additional effect calculated in step S108 is output to the diagnosis result output device 11 by the display control unit 10.
After performing the process in step S109, the series of processes in the abnormality sign monitoring system 100 ends (END).

Note that when a vibration sensor (not shown) is additionally installed in the pump 21, when it is confirmed that the plant 200 is operating normally in a subsequent predetermined period, the detection value of the vibration sensor for that predetermined period is Predetermined normal data including the normal data is stored in the normal database 3.

<Effect>
According to the first embodiment, based on the assumption that a predetermined undetected parameter (vibration) among a plurality of parameters (pressure, temperature, flow rate, vibration) used for monitoring signs of abnormality in the pump 21 is abnormal. , the cause of the sign of abnormality in the pump 21 is estimated. Then, the cause estimation results of the anomaly sign are compared between the case where the undetected parameter is not included and the case where the undetected parameter is included, and the comparison results are presented.

This allows the user to understand how the cause of the abnormality sign can be narrowed down if a predetermined sensor is newly installed to detect undetected parameters. Therefore, when rationalizing the maintenance work of the plant 200, the user can easily consider what kind of sensor should be newly installed in each device. In addition, for example, by newly installing a vibration sensor (not shown) that detects vibrations of the pump 21, if a similar abnormality sign occurs subsequently, the cause of the abnormality sign can be narrowed down, allowing maintenance and inspection work to be carried out. becomes easier to do.

Furthermore, even if a new vibration sensor is not installed in the pump 21, predetermined data presented to the diagnostic result output device 11 can be used for maintenance of the plant 200, such as when a worker periodically inspects the vibration of the pump 21. It can be used as a basis for judgment when reviewing inspection work.

Note that it is possible for the administrator to grasp the relationship between the predetermined undetected parameter and the cause of the abnormality sign using only the information in the physical model database 5 (see FIG. 3). However, there may be multiple undetected parameters for one piece of equipment, and there are many pieces of equipment in the plant 200, so it takes a lot of effort for the administrator to consider all the combinations. It takes. Further, for example, in a power generation plant, signs of equipment abnormality rarely occur, so it is not necessarily desirable to provide all of the large number of sensors corresponding to undetected parameters.

In contrast, according to the first embodiment, when an abnormality sign occurs in a device such as the pump 21, the cause estimation result of the abnormality sign is presented when the undetected parameter is assumed to be abnormal. This allows the user (administrator) to easily understand what kind of sensor would be effective to add in case a similar abnormality sign occurs in the future.

≪Second embodiment≫
The second embodiment differs from the first embodiment in that when an abnormal detected value exists for a device such as the pump 21, the parameter state simulator 7 simulates (assumes) that the undetected parameter is normal. are different. Note that other aspects (such as the configuration of the abnormality sign monitoring system 100: see FIG. 1) are the same as those in the first embodiment. Therefore, the parts that are different from the first embodiment will be explained, and the explanation of the overlapping parts will be omitted.

FIG. 9A is an explanatory diagram when the detected value of vibration of the pump is simulated to be normal by the parameter state simulation unit of the abnormality sign monitoring system according to the second embodiment (see FIG. 1 as appropriate).
Note that, similarly to FIG. 4 of the first embodiment, it is assumed that an abnormality in the temperature on the discharge side of the pump 21 is detected after August 4, 2020. Even in such a case, as shown in FIG. 9A, the parameter state simulation unit 7 simulates (assumes) that the vibration of the pump 21, which is an undetected parameter, is normal.

Then, in addition to the abnormal detected value (for example, the temperature on the discharge side of the pump 21), the second abnormality cause estimating unit 8 calculates a predetermined undetected parameter (undetected state quantity) that is undetected in the pump 21 (equipment). : In the example of FIG. 9A, the cause of the sign of abnormality in the pump 21 is estimated when it is assumed that the "vibration" of the pump 21 is normal.

FIG. 9B is an explanatory diagram of data in which the results of simulation by the parameter state simulator are incorporated into the detection results by the parameter abnormality detector.
In the example of FIG. 9B, after August 4, 2020, the temperature on the discharge side of the pump 21 is abnormal as shown by the frame G5, but on the other hand, the pressure, flow rate, and vibration on the discharge side of the pump 21 are abnormal. is normal (vibration is an assumption). Then, the second abnormality cause estimating unit 8 refers to the physical model database 5 (see FIG. 5) and estimates that the cause of the abnormality sign of the pump 21 is "pump seal wear."

FIG. 10 is an example of a screen display of the diagnosis result output device.
In the example of FIG. 10, if it is confirmed that the "vibration" of the pump 21 is normal, it is displayed that the cause of the sign of abnormality in the pump 21 can be narrowed down to "pump seal wear." This part is an additional effect obtained by detecting the "vibration" (undetected state quantity) of the pump 21.

<Effect>
According to the second embodiment, if it is assumed that the undetected parameter is "normal", the user can understand how the cause of the abnormality sign is narrowed down. Therefore, when rationalizing the maintenance work of the plant 200, etc., it becomes easier for the user to consider what kind of sensor should be newly installed in each device.

≪Third embodiment≫
In the third embodiment, the parameter state simulating unit 7 determines whether to simulate that the undetected parameter is abnormal or to simulate that the undetected parameter is normal, based on the detected value of each sensor. This embodiment differs from the first embodiment in this point. Note that other aspects (such as the configuration of the abnormality sign monitoring system 100: see FIG. 1) are the same as those in the first embodiment. Therefore, the parts that are different from the first embodiment will be explained, and the explanation of the overlapping parts will be omitted.

The parameter state simulating unit 7 shown in FIG. Determine as follows. For example, if the rotational speed command value of the motor (not shown) that is the drive source of the pump 21 is very small (that is, below a predetermined value), or if the output power to this motor is very small, Abnormalities rarely occur in the vibration of the pump 21. In such a case, the parameter state simulator 7 assumes that the vibration of the pump 21, which is an undetected parameter, is normal.

Further, when the rotational speed command value of a motor (not shown) that is a drive source of the pump 21 is very small, the flow rate of the pump 21 is also usually very small. Therefore, when the detected value of the flow rate sensor 33 is less than or equal to a predetermined value, the parameter state simulator 7 may assume that the vibration of the pump 21, which is an undetected parameter, is normal.
In addition, the parameter state simulation unit 7 determines whether the vibration of the pump 21 is normal or abnormal based on the rotational speed command value of the motor that is the drive source of the pump 21 and the magnitude of the detection value of each sensor. Either one of them may be assumed.

That is, when the second abnormality cause estimation unit 8 estimates the cause of the abnormality sign of the pump 21 (equipment), the magnitude of the detected value of at least one of the pressure sensor 31, the temperature sensor 32, and the flow rate sensor 33; Alternatively, one of abnormality and normality of the undetected parameter (undetected state quantity) may be assumed so as to match the magnitude of a predetermined command value used to drive the pump 21.

<Effect>
According to the third embodiment, abnormal vibration of the pump 21, which is an undetected parameter, is determined so as to match the rotational speed command value of the motor (not shown) of the pump 21 and the detected value of the flow sensor 33. Either normal is assumed. Therefore, when a vibration sensor (not shown) is newly installed in the pump 21, it is possible to appropriately show the user how to specifically narrow down the cause of the sign of abnormality in the pump 21.

<Fourth embodiment>
The fourth embodiment is characterized in that, in the physical model database 5A (see FIG. 11), the sensitivity of undetected parameters (for example, vibrations and sounds of the pump 21) to signs of abnormality in the pump 21 is associated with each other. This is different from the first embodiment. Note that other aspects (such as the configuration of the abnormality sign monitoring system 100: see FIG. 1) are the same as those in the first embodiment. Therefore, the parts that are different from the first embodiment will be explained, and the explanation of the overlapping parts will be omitted.

FIG. 11 is an explanatory diagram showing an example of the physical model database 5A of the abnormality sign monitoring system according to the fourth embodiment (see FIG. 1 as appropriate).
The example shown in FIG. 11 differs from the first embodiment (see FIG. 5) in that information regarding the "sound" of the pump 21 is included in the physical model database 5A. In addition, the triangular mark written in the "Sound" column of the pump 21 indicates that "sound" has lower sensitivity than "vibration" of the pump 21 in detecting signs of abnormality in the pump 21 (probability of detecting signs of abnormality). (low). In other words, information that "vibration" of the pump 21 is more sensitive than "sound" (higher probability of detecting an abnormality sign) is included in advance in the physical model database 5A.

In this way, multiple types of undetected parameters (undetected state quantities) are associated with the causes of abnormality signs of the pump 21 (equipment), and are also associated with the high sensitivity of the pump 21 to abnormality signs. There is. Note that the information indicating the sensitivity of the undetected parameter to a sign of abnormality of the pump 21 may be a relationship between the relative sensitivity of "vibration" and "sound" of the pump 21, Alternatively, the sensitivity to signs of abnormality of the pump 21 may be quantified in a predetermined manner.

FIG. 12 is an example of a screen display of the diagnosis result output device.
In the example of FIG. 12, in order to further narrow down the cause of the sign of abnormality in the pump 21, detection of the vibration of the pump 21 is "very effective," while detection of the sound of the pump 21 is "slightly effective." is displayed. In this way, displays such as "very effective" and "slightly effective" are also included in the display of the high sensitivity of the pump 21 to signs of abnormality.

In this way, when the display control unit 10 displays the difference between the estimation result of the first abnormality cause estimation unit 6 and the estimation result of the second abnormality cause estimation unit 8 on the diagnostic result output device 11 (display means), Information indicating the sensitivity of each of a plurality of types of undetected parameters (undetected state quantities) is also displayed on the diagnostic result output device 11 (display means).

<Effect>
According to the fourth embodiment, for example, a plurality of types of undetected parameters (vibration, sound) for narrowing down the cause of a sign of abnormality in the pump 21 can be presented to the user. Furthermore, by showing the sensitivity to signs of abnormality of the pump 21 for each unmeasured parameter, it is possible to present the user with information for making a decision when installing a new predetermined sensor.

≪Modification example≫
Although the abnormality sign monitoring system 100 according to the present invention has been described above based on each embodiment, the present invention is not limited to these descriptions, and various changes can be made.
For example, in the first embodiment, a case has been described in which the process (S107) of the second abnormality cause estimation unit 8 is performed after the process of the first abnormality cause estimation unit 6 (S105 in FIG. 8) is performed. The order of these processes may be reversed. Furthermore, the processing by the first abnormality cause estimation unit 6 (S105) and the processing by the second abnormality cause estimation unit 8 (S107) may be performed in parallel.

In addition, when the display control unit 10 displays the difference between the estimation result of the first abnormality cause estimation unit 6 and the estimation result of the second abnormality cause estimation unit 8 on the diagnosis result output device 11 (display means), A suggestion to add a predetermined sensor corresponding to the parameter (undetected state quantity) may also be displayed on the diagnostic result output device 11 (display means). This allows the user to understand specifically what kind of sensor should be added. Note that the display control unit 10 may also cause the diagnostic result output device 11 to display information indicating the installation location in the case of adding a predetermined sensor.

Moreover, each embodiment can be combined as appropriate. For example, the first embodiment in which it is assumed that the undetected parameter (undetected state quantity) is abnormal, and the second embodiment in which it is assumed that the undetected parameter is normal may be combined as appropriate. Then, in addition to the abnormal detected value (for example, the temperature on the discharge side of the pump 21), it is assumed that a predetermined undetected parameter that is not detected in the pump 21 (equipment) is abnormal, and a sign of an abnormality in the pump 21 is detected. The second abnormality cause estimation unit 8 may estimate the cause of the abnormality sign of the pump 21 when the undetected parameter is assumed to be normal. This allows the user to understand how the cause of the abnormality sign is narrowed down when undetected parameters are detected.

Furthermore, the first embodiment and the second embodiment may be combined to perform the following processing in the abnormality sign monitoring system 100. That is, the second abnormality cause estimating unit 8 determines the cause of the abnormality sign of the pump 21 (equipment) when it is assumed that the undetected parameter (undetected state quantity) is abnormal, and the cause of the abnormality sign that the undetected parameter is normal. Both of the causes of the abnormal sign of the pump 21 in the hypothetical case are estimated. Then, the display control unit 10 determines the cause of the abnormality sign of the pump 21 when it is assumed that the undetected parameter is abnormal, and the cause of the abnormality sign of the pump 21 when it is assumed that the undetected parameter is normal. The differences between the two are also displayed on the diagnostic result output device 11 (display means). This allows the user to understand from multiple angles how the cause of the abnormality sign is narrowed down when undetected parameters are detected.
In addition, for example, the second embodiment and the fourth embodiment may be combined, or the third embodiment and the fourth embodiment may be combined.

Further, in each embodiment, a case has been described in which the parameter abnormality detection unit 4 detects an abnormality in the state quantities (pressure, temperature, flow rate) related to the pump 21 based on the Mahalanobis distance, but the present invention is not limited to this. For example, when the pressure on the discharge side of the pump 21 deviates from a predetermined range set in advance, the parameter abnormality detection section 4 may detect a pressure abnormality. The same applies to the temperature and flow rate on the discharge side of the pump 21. In addition to the Mahalanobis distance, the parameter abnormality detection unit 4 may detect an abnormality in the state quantity related to the pump 21 based on a statistical method such as well-known clustering.

Furthermore, in each embodiment, in addition to detecting the pressure, temperature, and flow rate of the pump 21, a case has been described in which the vibration of the pump 21 is used as an undetected parameter. flow rate and vibration) can be changed as appropriate. Further, as the undetected parameter, a parameter whose detection accuracy is not so high or a parameter whose data tends to be insufficient may be used. Each embodiment can also be applied to estimating the cause of abnormal signs of other equipment included in the plant 200. In this case, it is assumed that a predetermined physical model database 5 (see FIG. 5) is set in advance for each device.

Further, in the first embodiment, in the physical model database 5 (see FIG. 5), the cause of the abnormality sign of the pump 21 is associated in advance with the combination of the presence or absence of an abnormality in the pressure, temperature, flow rate, and vibration of the pump 21. Although we have explained the case where there is For example, if the pressure of the fluid discharged from the pump 21 is abnormally low and the flow rate of the pump 21 is also abnormally low, a pipe leak (see Fig. 5) may occur, based on normal data. The physical model database 5 may include information indicating the magnitude relationship of the detected values of the state quantities.

Further, in addition to the parameter abnormality detection section 4 and the first abnormality cause estimation section 6, the processing of the parameter state simulation section 7, the second abnormality cause estimation section 8, the additional effect calculation section 9, the display control section 10, etc. It may be executed as a program. The above program can be provided via a communication line, or can be written on a recording medium such as a CD-ROM and distributed. Further, the normal state database 3 and the physical model database 5 may be stored in a storage means (not shown) of the computer.

Further, each embodiment is described in detail to explain the present invention in an easy-to-understand manner, and the present invention is not necessarily limited to having all the configurations described. Furthermore, it is possible to add, delete, or replace some of the configurations of the embodiments with other configurations. Further, each of the above-described configurations, functions, etc. may be realized in part or in whole by, for example, designing an integrated circuit. Further, each of the configurations, functions, etc. described above may be realized by software by a processor interpreting and executing a program for realizing each function. Further, the mechanisms and configurations described above are those considered necessary for explanation, and not all mechanisms and configurations are necessarily shown in the product.

1 Detected value acquisition device (detected value acquisition unit)
2 Time series data storage unit 3 Normal time database 4 Parameter abnormality detection unit 5,5A Physical model database (predetermined correspondence relationship)
6 First abnormality cause estimation section 7 Parameter state simulation section 8 Second abnormality cause estimation section 9 Additional effect calculation section 10 Display control section 11 Diagnosis result output device (display means)
21 Pump (equipment)
31 Pressure sensor (sensor)
32 Temperature sensor (sensor)
33 Flow rate sensor (sensor)
100 Abnormality sign monitoring system 200 Plant S102 Detection value acquisition step S105 First abnormality cause estimation step S107 Second abnormality cause estimation step S109 Display control step

Claims (7)

a detected value acquisition unit that acquires detected values of a plurality of state quantities regarding plant equipment from a plurality of sensors;
Based on predetermined normal data including detection values of the plurality of sensors acquired during a period when the device is known to be normal, there is an abnormality in the detection values of the plurality of sensors acquired by the detection value acquisition unit. If a detected value is included, the cause of the abnormality sign of the device is determined based on a predetermined correspondence relationship between the state quantity detected as the abnormal detected value and the cause of the abnormality sign of the device. a first abnormality cause estimation unit that estimates
In addition to the abnormal detected value, the cause of the abnormality sign of the device when it is assumed that a predetermined undetected state quantity that is not detected in the device is abnormal, and/or the undetected state quantity is normal. a second abnormality cause estimating unit that estimates the cause of the abnormality sign of the device when it is assumed that
an additional effect calculation unit that calculates an additional effect assuming the undetected state quantity based on the estimation results of the first abnormality cause estimation unit and the second abnormality cause estimation unit;
The difference between the cause of the abnormality sign of the device estimated by the first abnormality cause estimation unit and the cause of the abnormality sign of the equipment estimated by the second abnormality cause estimation unit is determined by the additional effect . An anomaly sign monitoring system comprising: a display control unit that causes a display unit to display the undetected state quantity together with the undetected state quantity. When the second abnormality cause estimation unit estimates the cause of the abnormality sign of the device, the magnitude of the detected value of at least one of the plurality of sensors and/or a predetermined command used to drive the device The abnormality sign monitoring system according to claim 1, wherein one of abnormality and normality of the undetected state quantity is assumed so as to match the magnitude of the value. The plurality of types of undetected state quantities are associated with the cause of the abnormality sign of the device, and also associated with the high sensitivity of the device to the abnormality sign,
The display control unit, when displaying the difference on the display means, also causes the display means to display information indicating the sensitivity of each of the plurality of types of undetected state quantities. The abnormality sign monitoring system according to claim 1. The display control unit, when displaying the difference on the display means, also causes the display means to display a proposal to add a predetermined sensor corresponding to the undetected state quantity. The abnormality sign monitoring system described in Item 1. The second abnormality cause estimation unit determines the cause of the abnormality sign of the device when it is assumed that the undetected state quantity is abnormal, and the cause of the abnormality sign of the device when it is assumed that the undetected state quantity is normal. Estimating both the causes of abnormal signs,
The display control unit is configured to determine the cause of the abnormality sign of the device when the undetected state quantity is assumed to be abnormal, and the cause of the abnormality sign of the device when the undetected state quantity is assumed to be normal. The abnormality sign monitoring system according to claim 1, wherein the difference between and is also displayed on the display means. a detected value acquisition step of acquiring detected values of a plurality of state quantities regarding plant equipment from a plurality of sensors;
Based on predetermined normal data including detection values of the plurality of sensors acquired during a period in which the device is known to be normal, there is no abnormality in the detection values of the plurality of sensors acquired in the detection value acquisition step. If a detected value is included, the cause of the abnormality sign of the device is determined based on a predetermined correspondence relationship between the state quantity detected as the abnormal detected value and the cause of the abnormality sign of the device. a first abnormality cause estimation step of estimating
In addition to the abnormal detected value, the cause of the abnormality sign of the device when it is assumed that a predetermined undetected state quantity that is not detected in the device is abnormal, and/or the undetected state quantity is normal. a second abnormality cause estimation step of estimating the cause of the abnormality sign of the device when it is assumed that
an additional effect calculation step of calculating an additional effect assuming the undetected state quantity based on the estimation results of the first abnormality cause estimation step and the second abnormality cause estimation step;
The difference between the cause of the abnormality sign of the device estimated in the first abnormality cause estimation step and the cause of the abnormality sign of the device estimated in the second abnormality cause estimation step is determined by the additional effect . An abnormality sign monitoring method , comprising: a display control step of displaying the undetected state amount on a display means together with the undetected state quantity. A program for causing a computer to execute the abnormality sign monitoring method according to claim 6.

Images