JP7303130B2 - Equipment diagnostic device and program - Google Patents

Description

The present invention relates to an equipment diagnosis device and program.

In plants such as power plants, equipment diagnostic technology that can be executed during operation is being introduced for the purpose of improving reliability and improving efficiency of maintenance work. As an example, Patent Document 1 below describes that "The phenomenon pattern extraction unit 104 extracts the phenomenon pattern of the past sensor signal of the facility. The related information linking unit 105 links the sensor signal based on the maintenance history information. The phenomenon pattern classification criterion creating unit 107 creates a phenomenon pattern classification criteria based on the extracted phenomenon pattern and the work keyword included in the maintenance history information linked to the sensor signal that is the basis of the phenomenon pattern. Based on the classification criteria, the phenomenon pattern classification unit 108 classifies the phenomenon patterns, and the diagnostic model creation unit 109 provides a maintenance worker with the classified phenomenon patterns and work keywords. Create a diagnostic model for estimating the working keywords to be presented.” (See abstract).

JP 2015-148867 A

By the way, with the technique described in Patent Document 1, it is possible to estimate the equipment to be maintained, but it is difficult to estimate the event to be maintained, and as a result, countermeasures for the event that caused the abnormality are postponed. There was a possibility of it getting lost.
SUMMARY OF THE INVENTION The present invention has been made in view of the circumstances described above, and it is an object of the present invention to provide a device diagnosis apparatus and a program that allow a user to appropriately estimate an event to be maintained.

In order to solve the above-mentioned problems, the device diagnosis apparatus of the present invention comprises a measured value acquiring unit for acquiring measured values of a plurality of parameter items in a plurality of devices belonging to a predetermined plant; Abnormality for calculating the degree of abnormality of the plant based on the normal measured value output unit that outputs the normal measured value, the measured value acquired by the measured value acquisition unit, and the normal measured value a degree calculation unit, an abnormality parameter item estimation unit for estimating one or more of the parameter items that contribute to the abnormality of the plant based on the degree of abnormality, and an abnormality parameter item estimation unit for calculating the parameter an abnormality determination unit that determines whether or not to execute item estimation; an event extraction unit that outputs one or more abnormal event items that are items that have caused an abnormality in the past for each of the parameter items; For each abnormal event item, an abnormal event estimation unit that calculates an event contribution rate that is the degree of contribution of the corresponding parameter item to the abnormality, and an abnormal event probability that is the probability of occurrence of each of the abnormal event items. and an event contribution ratio correction unit that calculates a modified event contribution ratio based on the event contribution ratio and the abnormality probability.

ADVANTAGE OF THE INVENTION According to this invention, a user can estimate appropriately the event which should be maintained.

1 is a block diagram of a plant equipment diagnosis system according to a preferred first embodiment; FIG. FIG. 4 is a diagram showing examples of parameter items and measurement value groups; It is a figure which shows an example of the degree of abnormality. It is a figure which shows an example of the contribution rate corresponding to each parameter item. FIG. 4 is a diagram showing an example of recorded data; It is a figure which shows an example of a device list. It is a figure which shows an example of an event list. It is a block diagram which shows an example of an abnormality occurrence probability calculation part. FIG. 5 is a diagram showing the relationship between the operating time and the amount of deterioration of equipment corresponding to an abnormal event item; It is a figure which shows an example of the temporal transition of a correction event contribution rate. It is a figure which shows an example of the display screen displayed on a display part. FIG. 11 is a block diagram of the essential parts of the second preferred embodiment;

[First embodiment]
<Configuration of the first embodiment>
FIG. 1 is a block diagram of a plant equipment diagnosis system 300 according to the first preferred embodiment.
In FIG. 1, the plant equipment diagnostic system 300 includes a plant 310, a measuring unit 320, and a plant equipment diagnostic device 330 (computer, equipment diagnostic device). The plant 310 is, for example, a nuclear power plant and comprises a plurality of (v) equipments EQ1 to EQv. The measurement unit 320 has a plurality of sensors (not shown), detects various states of the equipment EQ1 to EQv, and supplies the results to the plant equipment diagnostic device 330 as the original measured value group DG. The plant equipment diagnostic device 330 presents information that allows the user to estimate equipment to be maintained and events to be maintained based on the group of original measured values DG.

The plant equipment diagnosis device 330 includes general computer hardware such as a CPU (Central Processing Unit), RAM (Random Access Memory), ROM (Read Only Memory), and SSD (Solid State Drive). stores an OS (Operating System), application programs, various data, and the like. The OS and application programs are developed in RAM and executed by the CPU. In FIG. 1, the inside of the plant equipment diagnosis device 330 shows functions implemented by application programs and the like as blocks.

That is, the plant equipment diagnostic device 330 includes a measured value acquisition unit 2, a measured value database 4, a normal measured value database 5 (normal measured value output means, normal measured value output unit), and an abnormality degree calculation unit 6 (abnormality degree calculation means), abnormality determination section 8 (abnormality determination means), abnormal parameter item estimation section 9 (abnormal parameter item estimation means), abnormal event estimation section 11 (abnormal event estimation means), and device information database 13, an event extraction unit 16 (event extraction means), an abnormality occurrence probability calculation unit 23 (abnormality occurrence probability calculation means), an event contribution rate correction unit 25 (event contribution rate correction means), and a display unit 27. I have.

The measured value acquisition unit 2 outputs a measured value group DH based on the original measured value group DG received from the measuring unit 320 . The measured value group DH may be the result of constantly measuring the states of the devices EQ1 to EQv, ie, the original measured value group DG itself, or may be the original measured value group DG acquired during the inspection. That is, the measured value group DH is not limited by the mode of measurement such as acquisition timing.

FIG. 2 is a diagram showing an example of parameter items DN1 to DNm and measured value group DH.
Here, the parameter items DN1 to DNm are, for example, a plurality of (m) items such as "pipe temperature", "pump seal chamber pressure", "pipe flow rate", etc., as shown in FIG. Also, the measured value group DH is the measurement results corresponding to the parameter items DN1 to DNm and the time. Measured values corresponding to the individual parameter items DN1 to DNm are referred to as measured values D1 to Dm.

Returning to FIG. 1, the measured value database 4 accumulates accumulated measured values DAM. The accumulated measured values DAM are, for example, past measured values Dk (1≤k≤m) sampled at arbitrary intervals. Also, the accumulated measured values DAM may be statistical values such as the maximum value, minimum value, average value, standard deviation, etc. of the measured values Dk within a predetermined collection period. In addition, the normal measured value database 5 acquires measured values Dk of the plant 310 during normal operation in the past, and accumulates the results as normal measured values DBM.

The abnormality degree calculator 6 calculates an abnormality degree AB based on the measured value group DH and the normal measured value DBM. The degree of abnormality AB represents the degree of abnormality of the plant 310 at a certain time (that is, the degree of abnormality of the measured value group DH at a certain time). More specifically, the abnormality degree calculator 6 calculates the abnormality degree AB using the difference between the measured value Dk (1≦k≦m) belonging to the measured value group DH and the normal measured value DBM. The method for calculating the degree of anomaly AB is not particularly limited. A method of setting the degree AB can be adopted.

The abnormality determination unit 8 determines whether there is a possibility of abnormality in the plant 310 based on the abnormality degree AB calculated by the abnormality degree calculation unit 6, and outputs an abnormality determination value AJ. The abnormality determination value AJ is either "1" (high possibility of abnormality) or "0" (low possibility of abnormality). The method of determining abnormality is not particularly limited, and for example, a method of determining whether or not the degree of abnormality AB exceeds a predetermined abnormality determination threshold value DTH can be adopted.

FIG. 3 is a diagram showing an example of the degree of abnormality AB.
As illustrated, the degree of abnormality AB is calculated at each time when the measured value group DH (see FIG. 1) is input. Here, assuming that the abnormality determination threshold DTH (see FIG. 1) was "1.0", the abnormality degree AB exceeds the abnormality determination threshold DTH=1.0 from time 00:04 to 00:06, The abnormality determination unit 8 outputs "1" (high possibility of abnormality) as the abnormality determination value AJ. On the other hand, the abnormality determination unit 8 outputs “0” (low possibility of abnormality) as the abnormality determination value AJ from 00:00 to 00:03.

Returning to FIG. 1, when the abnormality determination value AJ is "1" (high possibility of abnormality), the abnormality parameter item estimation unit 9, based on the abnormality degree AB calculated by the abnormality degree calculation unit 6, the parameter An item that may be abnormal is estimated from items DN1 to DNm. The estimation method is, for example, calculating the contribution rate of each parameter item DNk (where 1 ≤ k ≤ m) to the degree of abnormality AB, and comparing them with each other to determine whether one or more parameter items DNk that may be abnormal identify. Here, the contribution ratio of each parameter item DNk to the degree of anomaly AB is called parameter contribution ratio DCRk.

Although the method of calculating the parameter contribution rate DCRk is not particularly limited, it can be calculated as follows, for example. First, the abnormal parameter item estimating unit 9 calculates a specific abnormality degree ABSk (not shown) that is the degree of abnormality of the measured values D1 to D(k−1) and D(k+1) to Dm excluding the parameter item DNk. . Then, the magnitude of the difference between the overall anomaly degree AB output from the anomaly degree calculator 6 and the specific anomaly degree ABSk is set as the parameter contribution rate DCRk of the parameter item DNk.

FIG. 4 is a diagram showing an example of contribution rates DCR1 to DCRm corresponding to each parameter item DN1 to DNm (see FIG. 2).
As shown in FIG. 4, the contribution rates DCR1 to DCRm are obtained at each time when the abnormality determination unit 8 sets the abnormality determination value AJ to "1" (high possibility of abnormality). In the example of FIG. 4, at time 00:06, the contribution rate DCR1 of the pipe temperature is 10%, the contribution rate DCR2 of the pump seal chamber pressure is 80%, and the contribution rate DCR3 of the pipe flow rate is 10%. It is estimated that the three parameter items DN1, DN2, and DN3 (see FIG. 2) are possibly abnormal parameters with a probability corresponding to .

Returning to FIG. 1 , the equipment information database 13 records record data ERD for various abnormal events that have occurred in the plant 310 .

FIG. 5 is a diagram showing an example of the recording data ERD.
In the illustrated example, the plurality of recorded data ERDs each include device name information 52, occurrence date and time information 54, abnormality overview information 56, abnormality cause information 58, response information 59 indicating the content of maintenance work, including. The abnormality overview information 56 also includes parameter item information 56a that specifies the parameter items DN1 to DNm (see FIG. 2) in which an abnormality has occurred. In the illustrated example, the parameter item information 56a is "pump seal chamber pressure".

Further, the abnormality cause information 58 includes abnormal event information 58a that caused the abnormality. In the illustrated example, the abnormal event information 58a is "a minute foreign matter is mixed in the mechanical seal". The format of the recording data ERD is not limited to that shown in FIG. 5, and various formats can be applied. However, it is preferable that the recording data ERD always include the device name information 52, the parameter item information 56a, and the abnormal event information 58a. Further, it is more preferable that the record data ERD is prepared in a report format as shown.

Returning to FIG. 1, the event extraction unit 16 generates m device lists LEQk (1≤k≤m) corresponding to the respective parameter items DN1 to DNm (see FIG. 2), and m create an event list LEVk of .

FIG. 6 is a diagram showing an example of the device list LEQk (1≦k≦m).
As illustrated, the device list LEQk is created for each parameter item DNk (1≤k≤m) such as "pump seal chamber pressure" and "piping temperature". Each device list LEQk includes a plurality of (c) device names 15-1 to 15-c, corresponding c device extraction numbers 19-1 to 19-c, and a device extraction number 19-1. and the total extraction number SUM, which is the sum of ~19-c. Here, the device names 15-1 to 15-c are lists of device name information 52 (see FIG. 5) appearing in records containing some kind of abnormality among the record data ERD.

FIG. 7 is a diagram showing an example of an event list LEVk (1≤k≤m).
Each event list LEVk (1≤k≤m) includes a plurality of (n) abnormal event items 12-1 to 12-n that have caused the abnormality and the corresponding n extracted event numbers 21-1. . Here, the abnormal event items 12-1 to 12-n list the abnormal event information 58a (see FIG. 5) appearing in records containing some kind of abnormality in the record data ERD. Note that the total extraction number SUM in the device list LEQk (see FIG. 6) for a certain parameter item DNk (1≦k≦m) matches the total extraction number SUM in the event list LEVk (see FIG. 7).

Returning to FIG. 1, the abnormal event estimation unit 11 calculates the event contribution rate Pi (1≤i≤ n) is calculated by “parameter contribution rate DCRk×number of extracted events 21-i/total number of extracted events SUM”. For example, at time 00:06 in FIG. 4, the contribution rate DCR2 of "pump seal chamber pressure" is 80%, so the event contribution rate Pi =P1 can be obtained as "contribution rate DCR2 x extracted event number 21-1/total extracted number SUM = 0.8 x 12/26 = 0.37 = 37%".

Through the above processing, the abnormal event estimator 11 identifies a plurality of abnormal event items 12-i (1≤i≤n) that may be the cause of the abnormality, and the events contributing to the parameter contribution rate DCRk. Contribution rate Pi is calculated. As a result, the user can, for example, examine maintenance measures from the abnormal event items 12-1 to 12-n having a high event contribution rate Pi, efficiently examine events that may be the cause of the abnormality, You can take action.

However, the event contribution rate Pi estimated here is calculated based on the number of extracted events 21-i (1≤i≤n) in the past. Since the states of various devices EQ1 to EQv (see Fig. 1) change due to deterioration over time, it is possible to simply apply the event contribution rate Pi based on the number of past event extractions 21-1 to 21-n. Instead, it may be preferable to reflect the current device state.
Therefore, the abnormality occurrence probability calculation unit 23 shown in FIG. 1 calculates the abnormality occurrence probability Pei (1≦i≦n) close to the current probability of causing the abnormality event item 12-i (1≦i≦n, see FIG. 7). Calculate

FIG. 8 is a block diagram showing an example of the abnormality occurrence probability calculator 23. As shown in FIG.
8, the abnormality occurrence probability calculation unit 23 includes a feature database 101 (feature data designation unit), a feature calling unit 102 (feature data designation unit), a function estimation unit 103, and a probability calculation unit 104. .

The abnormal event item 12-i described above generally progresses in deterioration such as "metal fatigue", "wear", and "corrosion" as the operating time t elapses. The progress of this deterioration can be expressed as a function gc. Here, the function gc has characteristics such as "linear function", "quadratic function", "exponential function", and "logarithmic function". Data representing these features is called feature data cs. For example, as an example of "deterioration", "metal fatigue of aluminum" can be expressed by "the number of repeated fractures". It is known that the "number of cycles to failure" of aluminum is a function that can be logarithmically approximated with respect to the stress amplitude associated with operation.

In FIG. 8, the feature database 101 stores a plurality of feature data cs such as "primary function" and "quadratic function". The feature calling unit 102 reads from the feature database 101 the feature data cs corresponding to the abnormal event item 12-i (1≤i≤n, see FIG. 7) that may be the cause of the abnormality. The function estimation unit 103 identifies the function gc based on the accumulated measurement value DAM and the feature data cs. Details thereof will be described below with reference to FIG.

FIG. 9 is a diagram showing the relationship between the operating time t and the amount of deterioration of the equipment corresponding to the abnormal event item 12-i (1≦i≦n, see FIG. 7) among the equipment EQ1 to EQv (see FIG. 1). is.
In FIG. 9, the horizontal axis represents the operating time of the equipment, and the vertical axis represents the amount of deterioration indicating the degree of deterioration of a certain abnormal event item 12-i (1≤i≤n) in the equipment. For example, if the abnormal event item 12-i is "seal surface wear", the "deterioration amount" can be "wear amount". The measurement point g is part of the accumulated measurement value DAM (see FIG. 8), which is the measurement result of the deterioration amount of the abnormal event item 12-i at or before the current operating time t1. The function estimator 103 (see FIG. 8) derives a function gc based on the feature data cs (see FIG. 8) and the measurement point g by a technique such as regression analysis.

Also, the difference between each measurement point g and the function gc is called gd. Further, the function estimating unit 103 (see FIG. 8) obtains the "dispersion" of the measurement point g such as variance and standard deviation based on the difference gd, and further based on the obtained "dispersion", the deterioration amount of the function gc Define the probability density function PDF for The type of probability density function PDF to be adopted is not particularly limited, and an appropriate probability density function may be adopted to represent the variation of the measurement points g.

In the following description, an example in which "normal distribution" is applied as the probability density function PDF will be described. When the average and standard deviation of the difference gd are determined, the probability density function PDF, which is a normal distribution, can be determined. Then, the probability density function PDF at the current operating time t1 can be represented by a curve that is the "average" of the function gc at the current operating time t1.

The probability calculator 104 (see FIG. 8) obtains an abnormality occurrence probability Pei (1≦i≦n) at the future operating time t2 based on the function gc and the probability density function PDF. Specifically, as shown in FIG. 9, the probability calculation unit 104 determines a future operating time t2 (for example, the current operating time t1 plus a predetermined operating time). Then, the probability calculation unit 104 obtains a probability density function PDF whose average value is the function gc at the future driving time t2. Then, the probability calculation unit 104 obtains an integral value in a range exceeding a predetermined abnormality threshold Th in the probability density function PDF at the future operating time t2, and the obtained integral value becomes the abnormality occurrence probability Pei. Although the method of integrating the probability density function PDF is not limited, for example, the Monte Carlo method can be applied.

Returning to FIG. 1, the event contribution rate correction unit 25 changes the abnormality occurrence probability Pei (1≤i≤n) calculated by the probability calculation unit 104 based on the following equation (1) to the event contribution rate Pi (1≤i≤n) ) to calculate the corrected event contribution rate P′i. Then, the display unit 27 displays data such as the calculated corrected event contribution rate P'i.

FIG. 10 is a diagram showing an example of temporal transition of the corrected event contribution rate P'i (1≤i≤n).
The example of FIG. 10 is an example in which the parameter item DNk is "pump seal chamber pressure" (see FIG. 2). As shown, the corrected event contribution rate P'i is obtained corresponding to time.

FIG. 11 is a diagram showing an example of a display screen 70 displayed on the display section 27 (see FIG. 1).
As shown, the display screen 70 includes a time display field 72 , anomaly degree display field 74 , a parameter contribution rate display field 76 , and a correction event contribution rate display field 78 . The degree-of-abnormality display field 74 displays the degree of abnormality AB. The time display field 72 displays the time when the degree of abnormality AB was measured. The parameter contribution rate display column 76 displays the parameter contribution rate DCRk (1≤k≤m). The corrected event contribution rate display column 78 displays the corrected event contribution rate P'i (1≤i≤n). In the example shown in FIG. 11, the corrected event contribution rate P'i is displayed in the corrected event contribution rate display column 78, but instead of this, the event contribution rate Pi may be displayed.

<Effect of the first embodiment>
As described above, the plant equipment diagnosis device 330 of the present embodiment has the abnormal event estimation unit 11 that calculates the event contribution rate Pi, which is the degree of contribution to the abnormality of the corresponding parameter item DNk, for each of the abnormal event items 12-i. , an abnormality occurrence probability calculator 23 that calculates an abnormality occurrence probability Pei that is the probability of occurrence of each of the abnormal event items 12-i, and a corrected event contribution rate P based on the event contribution rate Pi and the abnormality occurrence probability Pei and an event contribution rate correction unit 25 that calculates 'i.

As a result, according to the present embodiment, together with a plurality of abnormal event items 12-i that may be the cause of the abnormality, the event contribution rate Pi or the corrected event contribution rate P' that the event contributes to the degree of abnormality AB i can be calculated and a user of the plant equipment diagnostic system 300 can make a good estimate of the event to be serviced. For example, when the user determines that there is a possibility that an abnormality has occurred in the plant 310, for example, the necessity of maintenance and the means of maintenance can be examined in descending order of the event contribution rate Pi or the corrected event contribution rate P'i. It is possible to efficiently take countermeasures against abnormalities.

Further, the abnormality occurrence probability calculation unit 23 includes characteristic data designating units (101, 102) for designating characteristic data cs representing characteristics of progress of deterioration related to the abnormal event item 12-i due to the operation of the equipment EQ1 to EQv, and characteristic data cs , a measured value Dk, and a function estimating unit 103 that identifies a function gc of deterioration progress related to the abnormal event item 12-i by operation of the equipment EQ1 to EQv, and an abnormal event item 12-i based on the function gc It is preferable to include a probability calculation unit 104 for calculating an abnormality occurrence probability Pei, which is the probability of occurrence of As a result, the abnormality occurrence probability Pei can be appropriately calculated based on the characteristics of the progress of deterioration regarding the abnormality event item 12-i and the measured value Dk.

[Second embodiment]
FIG. 12 is a block diagram of essential parts of the second preferred embodiment.
The overall configuration of this embodiment is the same as that of the first embodiment (see FIG. 1), but instead of the abnormality occurrence probability calculator 23 shown in FIG. A calculation unit 200 (abnormality occurrence probability calculation means) is applied. In the following description, parts corresponding to those in the above-described first embodiment are denoted by the same reference numerals, and description thereof may be omitted. The abnormality occurrence probability calculation unit 200 includes a causal relational expression database 201 (causal relational expression specifying unit), a causal relational expression calling unit 202 (causal relational expression specifying unit), a parameter estimating unit 203, and a probability calculating unit 204. I have.

The causal relational expression database 201 stores a certain parameter item DNk (1≤k≤m, first parameter item) and other related parameter items (hereinafter referred to as related parameter items DNp (1≤p≤m, p≠k, (referred to as the second parameter item)) is stored in advance. Here, the causal relational expression f refers to a relational expression obtained based on a physical causal relation such as, for example, "pipe flow rate is expressed by constant×pipe passage area×pipe internal flow velocity".

The causal relational expression calling unit 202 retrieves the related parameter item DNp (1 ≦p≦m, p≠k) is read out from the causality expression database 201 . However, there may be a plurality of associated parameter items DNp corresponding to one abnormal event item 12-i. The parameter estimator 203 calculates the estimated value EV of the parameter item DNk corresponding to the abnormal event item 12-i based on the measured value Dp of the related parameter item DNp and the causal relational expression f.

Here, the estimated value EV is a value corresponding to the measured value Dk (1≤k≤m) of the parameter item DNk. is a different value. For example, when the parameter item DNk is "pipe flow rate", the measured value Dk is the measured value of the pipe flow rate, and the estimated value EV is a value calculated by "constant x pipe flow area x flow velocity in pipe" is. Here, the “constant” and the “pipe flow path area” are values recorded in advance in the causal relational expression database 201 . Also, the "pipe flow velocity" is the measured value Dp of the related parameter item DNp by a flow meter (not shown).

The probability calculation unit 204 calculates an abnormality occurrence probability Pei based on the measured value Dk of the parameter item DNk related to the abnormal event item 12-i (1≤i≤n) and the estimated value EV. The method of calculating the probability of occurrence of anomaly Pei is not particularly limited, but the probability of anomaly AP can be obtained by "Pei=|Dk−EV|/Tha" using a predetermined threshold value Tha (not shown).

<Effect of Second Embodiment>
As described above, according to the present embodiment, the user of the plant equipment diagnosis system 300 can appropriately estimate an event to be maintained, as in the first embodiment.

Furthermore, the abnormality occurrence probability calculation unit 200 calculates the difference between the first parameter item (DNk), which is the parameter item estimated to contribute to the abnormality, and the second parameter item (DNp), which is another parameter item. a causal relational expression designator (201, 202) that designates a causal relational expression f between EV), and a probability calculation unit 204 for calculating the abnormality occurrence probability Pei based on the relationship between the measured value Dk of the first parameter item (DNk) and the estimated value EV. is preferred. Thereby, the abnormality occurrence probability Pei can be calculated based on the relationship between the first parameter item (DNk) and the second parameter item (DNp).

[Modification]
The present invention is not limited to the embodiments described above, and various modifications are possible. The above-described embodiments are exemplified for easy understanding of the present invention, and are not necessarily limited to those having all the described configurations. Also, part of the configuration of one embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of one embodiment. Also, it is possible to delete part of the configuration of each embodiment, or to add or replace other configurations. Also, the control lines and information lines shown in the drawings are those considered to be necessary for explanation, and do not necessarily show all the control lines and information lines necessary on the product. In practice, it may be considered that almost all configurations are interconnected. Possible modifications to the above embodiment are, for example, the following.

(1) In each of the embodiments described above, an example in which a nuclear power plant is applied as the plant 310 has been described, but the plant 310 is not limited to a nuclear power plant, and various plants can be applied.

(2) In each of the above embodiments, the measured value database 4, the normal measured value database 5, the device information database 13, and the like are part of the plant device diagnostic device 330. FIG. However, these may be provided in separate server devices or the like (not shown), and these server devices or the like may be connected to the plant device diagnosis device 330 via a network.

(3) Since the hardware of the plant equipment diagnosis device 330 in each of the above embodiments can be realized by a general computer, the programs and the like for executing the various processes described above can be stored in a storage medium or distributed via a transmission line. may

(4) Various processes in the plant equipment diagnosis device 330 have been described as software processes using a program in the above embodiment, but some or all of them may be implemented in an ASIC (Application Specific Integrated Circuit). ), or hardware processing using an FPGA (Field Programmable Gate Array) or the like.

5 Normal measurement value database (normal measurement value output means, normal measurement value output unit)
6 Abnormality degree calculation unit (abnormality degree calculation means)
8 Abnormality determination unit (abnormality determination means)
9 Abnormal parameter item estimation unit (abnormal parameter item estimation means)
11 abnormal event estimation unit (abnormal event estimation means)
12-1 to 12-n Abnormal event item 16 Event extraction unit (event extraction means)
23 Abnormal Occurrence Probability Calculation Unit (Abnormal Occurrence Probability Calculation Means)
25 Event contribution correction unit (event contribution correction means)
101 feature database (feature data designator)
101, 102 (feature data designator)
102 feature calling unit (feature data designating unit)
103 Function estimation unit 104 Probability calculation unit 200 Abnormal occurrence probability calculation unit 201 Causal relational expression database (causal relational expression specifying unit)
201, 202 (causal relational expression specification part)
202 causal relational expression calling part (causal relational expression specifying part)
203 Parameter estimation unit 204 Probability calculation unit 310 Plant 330 Plant equipment diagnosis device (computer, equipment diagnosis device)
f Causal relational expression AB Abnormality degree Dk Measured value EV Estimated value Pi Event contribution cs Feature data gc Function DBM Normal measured value DNk Parameter item (first parameter item)
DNp related parameter item (second parameter item)
P'i Corrected event contribution rate Pei Abnormal occurrence probability EQ1 to EQv Device

Claims (4)

a measured value acquiring unit that acquires measured values of a plurality of parameter items in a plurality of devices belonging to a predetermined plant;
A normal measured value output unit that outputs a normal measured value that is the measured value during past normal operation of the plant;
an abnormality degree calculation unit that calculates the degree of abnormality of the plant based on the measured value obtained by the measured value obtaining unit and the normal measured value;
an abnormality parameter item estimating unit that estimates one or more of the parameter items that contribute to the abnormality of the plant based on the degree of abnormality;
an abnormality determining unit that determines whether or not to cause the abnormal parameter item estimating unit to execute estimation of the parameter item;
an event extraction unit that outputs one or more abnormal event items, which are items that have caused an abnormality in the past, for each of the parameter items;
an abnormal event estimator that calculates, for each of the abnormal event items, an event contribution rate that is a degree of contribution to the abnormality of the corresponding parameter item;
an anomaly occurrence probability calculation unit for calculating an anomaly occurrence probability, which is the probability of occurrence of each of the anomaly event items;
and an event contribution rate correction unit that calculates a corrected event contribution rate based on the event contribution rate and the abnormality occurrence probability. The abnormality occurrence probability calculation unit,
a feature data specifying unit that specifies feature data representing features of progress of deterioration of the abnormal event item due to operation of the equipment;
a function estimating unit that identifies a function of progress of deterioration related to the abnormal event item due to operation of the equipment based on the feature data and the measured value;
2. The apparatus diagnostic apparatus according to claim 1, further comprising a probability calculation unit that calculates an abnormality occurrence probability, which is a probability of occurrence of the abnormality event item, based on the function. The abnormality occurrence probability calculation unit calculates a causal relational expression between a first parameter item that is the parameter item estimated to contribute to the abnormality and a second parameter item that is the other parameter item. a causal relational expression specifying part to specify;
a parameter estimation unit that calculates an estimated value of the first parameter item based on the causal relational expression and the second parameter item;
The equipment diagnosis apparatus according to claim 1, further comprising a probability calculation unit that calculates the probability of occurrence of abnormality based on the relationship between the measured value of the first parameter item and the estimated value. the computer,
Measured value acquisition means for acquiring measured values of a plurality of parameter items in a plurality of devices belonging to a predetermined plant;
Normal time measured value output means for outputting a normal time measured value which is the measured value during past normal operation of the plant;
Anomaly degree calculation means for calculating an anomaly degree of the plant based on the measured value acquired by the measured value acquisition means and the normal measured value;
abnormal parameter item estimation means for estimating one or more of the parameter items that contribute to the abnormality of the plant based on the degree of abnormality;
abnormality determination means for determining whether or not to cause the abnormality parameter item estimation means to execute estimation of the parameter item;
Event extracting means for outputting one or more abnormal event items, which are items that have caused an abnormality in the past, for each of the parameter items;
Abnormal event estimating means for calculating, for each of the abnormal event items, an event contribution rate that is a degree of contribution to the abnormality of the corresponding parameter item;
anomaly occurrence probability calculation means for calculating an anomaly occurrence probability that is the probability of occurrence of each of the anomaly event items;
event contribution ratio correction means for calculating a corrected event contribution ratio based on the event contribution ratio and the abnormality occurrence probability;
A program to function as

Images