JPH0675245B2 - Plant diagnostic equipment - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial field of application] The present invention identifies an abnormal event of a plant in an online real time in order to improve the operational reliability and the operating rate of a large-scale plant. The present invention relates to a plant diagnostic device that provides aggregate information to operators.

[Conventional technology]

As a conventional device of this type, there is one shown in FIG. The figure is a block diagram showing the configuration of the plant diagnosis device. In the figure, 1 is plant data input from the plant, 2 is plant data 1 input from the plant to diagnose a plant abnormality, and details of the abnormality (cause of abnormality) , A fault location and a magnitude of the fault) and determine an operation guide for dealing with the fault, a plant fault diagnosis unit, 3 is a message processor for displaying a message to an operator after the plant fault is diagnosed, and 4 is a message processor. Message file, 5 is the important process amount determination unit (important process amount determination unit) that is started after the plant abnormality diagnosis, 6, 12 is the process trend collection unit, 7 is the parameter file for displaying the process trend, and 8 and 13 are The trend data file, 9 inputs the operation guide information from the plant abnormality diagnosis unit 2 to check the quantitativeness of the operation guide. Manipulated variable determining unit for several selected meaning,
10 and 20 are a plurality of prediction simulation units that predict plant state fluctuations in parallel for several types of manipulated variables, 11 is a data file for executing the prediction simulation, and 14 and 15 are CRT display units (display units). , 16 are the optimum physical model selection / determination units (physical model selection units) that can predict the fastest and most accurately according to the abnormality that has occurred. Reference numeral 17 denotes a normal-time plant data file that stores normal-time plant data periodically output from the plant abnormality diagnosis unit 2.

Next, the operation will be described. First, information necessary for operation is taken from the plant data 1 of the actual plant, the cause of abnormality of the plant, the abnormal place, and the magnitude of the abnormality are identified in the plant abnormality diagnosing unit 2, and the operation guide of the plant is generated. Determine the deafening event. When the message processor 3 receives the information from the plant abnormality diagnosis unit 2, the message processor 4 displays the message on the cathode ray tube display unit 14 using the message file 4. In addition, the important process amount determination unit 5 uses the plant abnormality diagnosis unit 2
It is activated by receiving a signal indicating that the cause of abnormality has been successfully identified from, and determines the amount of processes affected by the cause of abnormality, the amount of processes to be monitored during recovery operations, etc., together with the priority. Next, the first process trend collection unit 6 inputs data from the plant data regarding the process amount determined by the important process amount determination unit 5, collects process trends, and creates a process trend file 8. . Further, when the operation amount determination unit 9 receives the operation guide information from the plant abnormality diagnosis unit 2 and the current operation amount of the device corresponding to the operation guide from the plant data,
Several types of operation amounts are determined from the current operation amount and the maximum possible operation amount of the device. The physical model selection / determination unit 6 receives information such as the cause of abnormality, the location of the abnormality, and the size of the abnormality from the plant abnormality diagnosis unit 2 so that the physical model can be predicted at the highest speed and accuracy according to the information. The optimal model is selected and determined in consideration of the modeling method, the number of model nodes, the number of loops, and so on. Further, the normal-time plant data file 17 stores and updates the plant data periodically output from the plant abnormality diagnosis unit 2, and by means such as comparison with the previously stored plant data,
The plant data for normal times is always stored. The plurality of predictive simulation units 10 and 20 receive the cause of abnormality, the abnormal place, and the magnitude of the abnormality from the abnormality diagnosis unit, and select the optimum model from the physical model selection / determination unit 16 and the normal plant data file. The plant observation signal is input from 17 and the initial state of the prediction simulation is set and stored. After that, by inputting a plurality of operation signals from the operation amount determination unit, for each operation amount, the same model in the plurality of prediction simulation units 10 and 20 is simultaneously activated to determine how the plant state after an abnormality occurs. Predict if it changes in parallel and at high speed and quantitatively. The prediction simulation units 10 and 20 are executed using the data file 11. The second process trend collection unit 12 collects trend data for the process amount from the important process amount determination unit 5 by using the data file that is the result of the prediction simulation, and outputs the second trend data file 13 as a result. create. The process trend and the trend of each prediction simulation result are displayed on the same graph of the second cathode ray tube display unit 15 by using the first process trend file, the second trend data file and the trend parameter file 7 for trend display. To do.

[Problems to be solved by the invention]

Since the conventional plant diagnostic device is configured as described above, the prediction simulation unit has to perform prediction using the model of the entire plant, and if it is desired to perform prediction of certain information at higher speed, There were problems such as too much cost.

The present invention has been made to solve the above problems, and an object of the present invention is to provide a plant diagnostic device that can perform prediction for each system or for each individual device and can switch between models. To do.

[Means for solving problems]

The plant diagnostic device according to the present invention monitors the intermediate results of the simulation in the prediction simulation unit,
When the intermediate result satisfies a predetermined plant condition, the prediction simulation unit is switched to a physical model which is simplified from the physical model in use and the simulation is continued.

[Work]

In the plant diagnosis according to the present invention, when a model is selected for each system or for each device and the prediction is performed, the optimum boundary is determined from the current plant state from a plurality of boundary condition groups prepared in advance. Select conditions and make predictions. In the middle of the prediction, if the use model is simplified and the effect is small, the model is switched to further speed up.

〔Example〕

An embodiment of the present invention will be described below with reference to the drawings. In the figure, the same parts as those in FIG. 2 are shown with the same reference numerals. In FIG. 2, reference numeral 18 is a boundary condition file prepared in advance for use in carrying out the prediction, and 19 is a model switching section during the prediction.

Next, the operation will be described. In the figure, the same parts as those in FIG. 2 perform the same operations, and therefore detailed description thereof will be omitted. First, when the physical model selecting / determining unit 16 receives the cause of abnormality, the abnormal place, and the size of the abnormality from the plant abnormality diagnosing unit 2, the physical model is selected so as to make the fastest and most accurate prediction in accordance with the information. The optimum model is selected and determined in consideration of the modeling method, the number of node divisions in the model, the number of loops, the entire plant or individual devices. In the normal-time plant data file 17, the plant data periodically output from the plant abnormality diagnosis unit 2 is stored and updated, and the plant data in the normal state is always stored by means such as comparison with the previously stored plant data. It is stored. The boundary condition file 18 is prepared for performing the prediction partially at a higher speed, and is used when performing the prediction for each individual device or each system. This file has boundary conditions assuming a plurality of accidents, but a plurality of boundary conditions that are severe on the equipment or system shall be prepared for each accident.

The plurality of prediction simulation units 10 and 20 receive the cause of abnormality, the location of the abnormality, the magnitude of the abnormality, and the like from the plant abnormality diagnosis unit 2, and the physical model selection / determination unit 16
A more optimal model, a plant observation signal from the normal data file 17 and an optimal boundary condition from the boundary condition file 18 are input, and the initial state of the prediction simulation is set and stored. After that, a plurality of operation signals from the operation amount determination unit 9 are input, and a plurality of prediction simulation units are simultaneously activated for each operation amount to determine in parallel how the plant state changes after an abnormality occurs. high speed,
And make a quantitative prediction. The prediction simulation unit 1
0 and 20 are executed using the data file 111.

The model switching unit 19 monitors the main data of the data file 11, and when a state quantity that can be reconsidered as a constant value is determined by judging from the plant operating conditions, the model switching signal is predicted by the simulation unit when the model related to it can be simplified.
Send to 10,20. If there is a signal from the model switching unit, the prediction simulation units 10 and 20 switch the model and continue the prediction. The second process trend collection unit 12 collects trend data for the process amount from the important process amount determination unit 5 by using the data file that is the result of the prediction simulation, and outputs the second trend data file 13 as a result. create. Using the first trend data file 8, the second trend data file 13 and the normal plant data file 17 for displaying the trend, the process trend and each prediction are displayed on the same graph of the second cathode ray tube display unit 15. Display the trend of simulation results.

〔The invention's effect〕

As described above, according to the present invention, the intermediate result of the simulation in the predictive simulation unit is monitored, and when the intermediate result has a predetermined plant condition,
The predictive simulation unit is configured to continue the simulation by switching to a physical model that is simpler than the physical model in use, so the processing after switching the physical model is faster, and as a result, the simulation result is shorter. There is an effect that can be obtained in time.

Further, when the physical model for each system or each device is selected, the simulation result can be obtained in a shorter time.

[Brief description of drawings]

FIG. 1 is a block diagram of a plant diagnostic device according to an embodiment of the present invention, and FIG. 2 is a block diagram of a conventional plant diagnostic device. In the figure, 1 is plant data, 2 is a plant abnormality diagnosing unit, 3 is a message processor, 4 is a message file, 5 is an important process amount determining unit, 6 and 12 are first and second process trend collecting units, 7 is a trend parameter file, 8 and 13 are first and second trend data files, 9 is a manipulated variable determination unit, 10 and 20 are first and second prediction simulation units, 11 is a data file, and 14 and 15 are First and second CRT display section, 16 physical model selection / determination section, 17 normal plant data file, 18 boundary condition file, 19
Is a model switching unit. In the drawings, the same reference numerals indicate the same or corresponding parts.

Claims (2)

[Claims] 1. A plant abnormality diagnosing unit which inputs plant data from a plant to diagnose an abnormality of the plant, identifies the content of the abnormality, and determines an operation guide for dealing with the abnormality, and the plant abnormality diagnosing unit. The important process amount determination unit that determines the important process amount based on the content of the abnormality identified by the, and the plant data related to the important process amount determined by the important process amount determination unit of the plant data is collected. A first process trend collection unit that generates a trend data file, an operation amount determination unit that determines an operation amount of a device corresponding to the operation guide determined by the plant abnormality diagnosis unit, and the plant abnormality diagnosis unit. A physical model selection unit that selects a physical model to be used for simulation based on the content of the identified abnormality , A prediction simulation unit that simulates a state change of the plant with respect to the manipulated variable determined by the manipulated variable determiner according to the physical model selected by the physical model selection unit, and an intermediate result of the simulation in the predictive simulation unit is monitored. If the intermediate result has a predetermined plant condition, a model switching unit for switching the prediction simulation unit to a physical model simplified from the physical model and continuing the simulation, and a prediction simulation unit A second process trend collection unit that collects the simulation result related to the important process amount among the simulation results and generates a trend data file, and the trend data generated by the first and second process trend collection units. Plant diagnosis apparatus having a display unit for displaying the trend graph on the basis of the datafile. 2. The physical model selection unit selects a physical model for each system or each device based on the details of the abnormality identified by the plant abnormality diagnosis unit when selecting a physical model used for simulation. The plant diagnostic device according to claim 1, wherein