JPS6277612A - Abnormality diagnosing method for plant - Google Patents

Abstract

PURPOSE:To analyze abnormality speedily without evaluating all logical expressions on every sampling operation by evaluating only logical expressions regarding variation of the process signal of a plant selectively. CONSTITUTION:The processing signal obtained from the plant at constant intervals of time is inputted to a signal input device 6. The input signal is supplied to an input signal processor 7 to check whether its value varies or not and the evaluation result is sent to an input signal storage device 8. The input signal storage device 8 and an evaluation order storage device 9 are used to evaluate the logical expressions regarding the input signal which has the variation. The last evaluation result is substituted in other logical ex pressions as it is. Consequently, the output signal which contains abnormality diagnostic logic generated previously by an abnormality diagnostic logic storage device 10 is inputted to an abnormality diagnostic processor 11, whose diagnostic result is displayed on a diagnostic result display device 12 to an operator.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to a plant abnormality diagnosis method for diagnosing plant abnormalities using a computer system.

[Technical background of the invention and its problems]

A conventional plant abnormality diagnosis device will be explained with reference to the block diagram shown in FIG. In the figure, 1 is a signal input device that inputs a plant process signal, 2 is an input signal processing device that compares the value of the input signal with a preset threshold value and determines the state of the signal, and 3 is a signal input device that inputs a plant process signal. An abnormality diagnosis logic storage device 4 stores abnormality diagnosis logic created in advance, and 4 calculates the abnormality diagnosis logic according to the state of the input signal, identifies the cause of the abnormality, and extracts operations for restoring the plant state. The abnormality diagnosis processing device 5 is a diagnosis result display device that displays diagnosis results.

Here, the logic stored in the abnormality diagnosis logic storage device 13 may be expressed as an individual logical formula, or may have a tree structure such as a cause-and-effect tree.

In other words, when the abnormality diagnosis logic is expressed as a logical formula, the abnormality diagnosis is performed using a logical operation using the input signal referenced by the logical formula and the evaluation result of another # ring formula, and the abnormality diagnosis logic is established. Determine whether or not. Furthermore, when a logical expression refers to the evaluation result of another logical expression, the referenced logical expression must be evaluated first, and the evaluation order is uniquely determined. Furthermore, when no other logical expressions are referenced, the evaluation order is mechanically evaluated from the top. However, in either case, all logical expressions are evaluated for each sampling, so when the number of logical expressions increases, there is a problem in that it takes a lot of analysis time and makes it impossible to perform appropriate evaluation.

Next, a case where the abnormality diagnosis logic is a cause-and-effect tree will be described. Figure 5 is an example of a portion of a cause-and-effect tree.
In the figure, N7. N13 is a diagnostic message, Gl, G2
．． G4. G5 is an AND gate, G3. G6 is an OR gate, and N1 to N13 are nodes representing signal states.

Execution processing of the cause-effect tree is performed by the abnormality diagnosis processing device based on information from the input signal processing device and the abnormality cause diagnosis storage device that stores the cause-effect tree, and the following processing is executed to identify the cause of the abnormality. Ru.

First, for a node where a signal whose state has changed is used, check whether the condition for that node to become negative holds true.For example, in Figure 5, if node N7 becomes true, then N7 is true. The condition for this is that one of N5, N1, N2, N6, N3, and N4 holds true. This set that is true is called a minimal cut set.

Also, for node Nil to be true, N8・N7・(N5・N1・N2+N6・N3・N4)
=N8・N7・N5・N1・N2・tenN8・N7・N6
・N3 and N4 must be satisfied. At this time, there are two minimal cut sets: N8, N7, N5, N1, N2, N3, N7, N6, N3, and N4. Therefore, when the state of a signal used at a node changes, the upstream side of that node (5th
In the figure, the state of all nodes located at the bottom is upstream and the top is downstream) is investigated to determine whether any minimal cut set is established. Then, if a certain minimal cut set is established, it is assumed that the cause identification has been successful, and if there is a diagnostic message for that node, the diagnostic message is displayed.

This type of processing is performed in the cause-and-effect tree, and since the nodes that become true are analyzed, it can be said that the processing is selective to some extent.

However, in a cause-effect tree, it is essential to establish a minimal cut set in order to identify the cause.

In addition, the minimal cut set is composed of all the nodes upstream of the node of interest, so when the event progresses and the node of interest comes downstream in the cause-and-effect tree, that node is placed at the top. The amount of minimal cut sets required becomes enormous, and the analysis time increases. Furthermore, if the order of changes in events changes due to a time delay in events or the way thresholds are given, the minimal cut set will not hold, and the cause of the abnormality will fail to be identified.

[Purpose of the invention]

The present invention has been made in response to the above-mentioned demand for technical solutions, and its purpose is to provide a plant abnormality diagnosis method that efficiently diagnoses plant abnormalities using logical expression expressions.

[Summary of the invention]

In order to achieve the above object, the present invention inputs a process signal from a plant, performs logical operation processing on the input signal to identify the cause of abnormality in the plant, and detects the abnormal state according to the operating status of the plant. In a plant abnormality diagnosis method that provides operators with operations to prevent the spread of anomaly, a change in the input signal is used as a trigger to selectively process logical operations and to set conditions on the evaluation of changes in the input signal, thereby preventing erroneous judgments. This improves analysis efficiency by preventing this from happening.

[Embodiments of the invention]

Embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram of an embodiment of the present invention. In the figure, 6 is a signal input device that inputs plant process signals, 7 is an input signal processing device that compares the value of the input signal with a predetermined value and evaluates the state of the signal, and 8
9 is an input signal storage device that stores the evaluation results in a prepared first table; 9 is an evaluation order storage device that stores the evaluation order in a second table corresponding to the first table; and 10 is an abnormality diagnosis logic prepared in advance. 11 is an abnormality diagnosis processing device that calculates abnormality diagnosis logic according to the state of the input signal, identifies the cause of the abnormality, and selects an operation for restoring the plant state; 12 is a diagnosis result; This is a diagnostic result display device that displays.

Next, the operation of this embodiment will be explained. First, process signals obtained from the plant at regular intervals are input to the signal input device 6. The input signal is.

In the case of an analog signal, the input signal processing device 17 compares the value of the input signal with a predetermined threshold value to check whether there is a change in the state, and if it is a digital signal, checks whether the value has changed and evaluates the value. The result is sent to the input signal storage [8.

In the input signal storage device [8, a first table as shown in FIG. 2 is prepared. Each item in Figure 2 means the following.

No, : Number of the input signal Value : If the input signal is an analog signal, the truth evaluation result of whether it exceeds the threshold, if it is a digital signal, the true/false value Flagl : The value of the signal is false → Flag value that is set when the signal value changes from true to false Flag2: Value of the flag that is set when the signal value changes from true to false Flag3: The value of the flag that becomes the logical expression analysis trigger when Flagl or Flag2 becomes true. Values Ref, Buf, No,: Buffer No. of the logical expression to which the signal is referenced.

Corresponding to this first table, a second table as shown in FIG. 3 is prepared in the evaluation order storage device [9].Each item in Figure 3 means the following.

Buffer No.: Number indicating the evaluation order of the logical expression F1ag4: Value of the flag used to determine whether to evaluate the logical expression TD: Count value of the time delay operator included in the logical expression Logic No.: Number of the logical expression Number Value:
Evaluation result of logical expression Ref, Buf, No,: BofforNo of the logical expression to which the logical expression is referred.

Here, Buffer No. is a number indicating the evaluation order when evaluating a logical expression at all points. The logical expression selection processing method will be explained below.

First, when the input signal changes, Flag 3 of that signal becomes true, and accordingly, F of the logical formula that refers to that signal.
lag4 becomes true. If there are multiple changed signals, set the flag of the related logical expression for all the changed signals.
Make 4 true.

Next, it is determined whether Buffer No. and Flag 4 of the logical formula are true or false, and if true, Flag 4 of the logical formula that refers to the logical formula is set to true. Thereafter, Flag4 is determined in the same manner up to the last Buffer No.

After that, if the logical expression for which Flag 4 is true is evaluated, the evaluation of all logical expressions related to the changed input signal is completed.

There is also a time delay operator as another analysis trigger for logical expressions. Therefore, a time delay counter is prepared for logical expressions that include time delay operators. This counter is filled with the delay time N when it becomes necessary to perform a time delay operation, that is, when the logical operation that applies the time delay changes. When N is entered in this counter, the count value is sequentially decreased in accordance with the sampling period, and when the count value becomes 0 or more, Flag 4 of the logical formula is set to true. This allows a change due to time delay to be used as a trigger for analysis, similar to a change in the input signal.

Note that for logical expressions that have not undergone selection processing, the previous evaluation result is entered as is.

Then, the output signals from the input signal storage device 8 and the evaluation order storage device 9 stored as described above and the output signals storing the abnormality diagnosis logic created in advance in the abnormality diagnosis processing device [10] is input to the abnormality diagnosis processing device 11, which calculates abnormality diagnosis logic according to the state of the input signal, identifies the cause of the abnormality,
An operation for restoring the plant condition is selected, and the selection result is displayed to the operator on the diagnostic result display device 12.

As described above, according to the plant abnormality diagnosis method of this embodiment, the selection process of abnormality diagnosis logic can be performed without using a tree ellipse such as a cause-and-effect tree, and it is expected that the analysis time will be shortened. In addition, the change in processing time is small even when the scale of the abnormality diagnosis logic increases, and there is a possibility that the processing time greatly depends on the scale of the abnormality diagnosis logic, as in the conventional all-point processing method or cause-effect tree processing method. There are almost no

In addition, by setting conditions for evaluating input signal changes, it is possible to exclude any incorrect signal changes due to detector failure from being analyzed, which also prevents redundant analysis and erroneous judgments. It is valid.

〔Effect of the invention〕

As explained above, according to the present invention, by selectively evaluating logical expressions using changed signals as triggers, abnormalities can be quickly analyzed and diagnostic information can be used to prevent the spread of abnormality propagation. can be provided to members.

[Brief explanation of drawings]

FIG. 1 is a block diagram of an embodiment of the present invention, FIG. 2 is a diagram for explaining the input signal storage device in FIG. 1, and FIG. 3 is a diagram for explaining the evaluation order storage device in FIG. 1. Figure 4
The figure is a block diagram of a conventional plant abnormality diagnosis device, and FIG. 5 is a diagram for explaining a conventional cause-effect tree. 6...Signal input device 7...Input signal processing device 8...Input signal storage device 9...Evaluation order storage device 10...Abnormality diagnosis logical storage device 11...
Abnormality diagnosis processing device 12...Diagnosis result display device (8733) Representative Patent attorney Yoshiaki Inomata (and 1 others)
name) Figure 1

Claims (1)

[Claims] By inputting process signals from the plant and processing the input signals with logical operations, it identifies the cause of abnormality in the plant, and provides operators with operations to prevent the abnormality from expanding depending on the operating status of the plant. In the plant abnormality diagnosis method that uses input signal changes as a trigger, logical operations are selectively processed, and conditions are attached to the evaluation of input signal changes to prevent erroneous judgments and improve analysis efficiency. A plant abnormality diagnosis method characterized by: