JPH08263135A - Plant control system maintenance support system - Google Patents

Abstract

PURPOSE: To enable even a general service engineer who does not understand a control system well to perform operation as a service engineer who knows the whole control system well does. CONSTITUTION: When the control system becomes abnormal, the service engineer 6 inputs the contents of an abnormal event to a maintenance support device main body 1 from an input/output device 5. An inference engine 3 in the maintenance support device main body 1 knows cause which possibly causes the said inputted abnormal event by referring to an abnormal event frame 2a in a knowledge base 2 on the basis of the inputted abnormal event. Then the inference engine 3 refers to a cause frame 2b in the knowledge base 2 and displays messages asking the service engineer 6 questions about whether or not respective events are derived from respective causes on the input/output device 5. When the service engineer 6 answers the questions, the inference engine 3 refers to the cause frame 2b according to the answers and displays which measure should be taken against the current abnormal event on the input/output device 5 to teach the service engineer 6.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a plant control system maintenance support system for diagnosing the cause of an abnormality in a control system used in, for example, a power plant and providing a countermeasure to an operator.

[0002]

2. Description of the Related Art In recent years, for example, in a power plant or the like, the entire plant is monitored and operated in a central cab, and its control system is becoming larger and more complicated. Further, the control device is also shifting from the conventional analog control system to a digital control system using a microcomputer, and the inside of the control system is becoming difficult to understand from the outside.

In such a situation, if a failure occurs at a certain place in the control system, an abnormal event will appear on the table over a wide range in the control system. However, in the past,
When an abnormality occurs in the control system, the cause of the abnormal event is investigated by relying on the knowledge and feeling of a person who is familiar with the entire control system, and what kind of countermeasure should be taken.

[0004]

As described above, conventionally, when an abnormality occurs in the control system, it is necessary to find out the cause of the abnormal event and the countermeasures against it, based on the knowledge of the individual who is familiar with the entire control system. There were many places that depended on my feelings, and there were many difficult places. Further, recently, since a complicated and large-scale control system is manufactured by a plurality of companies and a large number of engineers, the number of people who are familiar with the entire control system is almost zero. Under these circumstances, there is currently a demand for a maintenance support system that can immediately provide the following support. (1) When a malfunction occurs in a complicated large-scale control system, the cause of the malfunction is determined, and the maintenance staff is instructed (advised) to take corrective action. (2) When an abnormality occurs in a device (part) having a control system, a maintenance worker learns what kind of phenomenon occurs in the entire plant.

The present invention has been made in view of the above circumstances,
It is an object of the present invention to provide a plant control system maintenance support system that allows a general maintenance person who does not fully understand a control system to perform the same work as a maintenance person who is familiar with the entire control system.

[0006]

A plant control system maintenance support system according to the present invention includes an input / output device having a data input function and a display function, and a question, an answer, and a corresponding action output to the input / output device. The control means for controlling the above, etc., and the abnormal event frame that describes the cause for the abnormal event that may occur on the target control system, and the countermeasures against the derived event and the cause that may occur when the cause occurs are described. Based on the knowledge base having a cause frame and the above-mentioned knowledge base based on the abnormal event or cause input from the input / output device, a countermeasure against the input abnormal event or a derivative event for the input cause And an inference engine for displaying on the input / output device.

Further, the plant control system maintenance support system according to the present invention controls an input / output device having a data input function and a display function, and controls the question, answer, and corresponding treatment output to the input / output device. Control means to perform,
A knowledge base having an abnormal event frame in which the cause for an abnormal event that may occur on the target control system is described and a cause frame in which countermeasures for derivative events and causes that can occur when the cause occurs are described, For the abnormal event input by the input / output device, search the cause that may be abnormal by referring to the abnormal event frame of the knowledge base, refer to the cause frame based on this cause, and check the occurrence event for each cause. An inference engine for checking the presence / absence and displaying a countermeasure against an abnormal event on the input / output device is provided.

[0008]

When the control system is abnormal, the maintenance staff inputs the content of the abnormal event into the maintenance support system. When an abnormal event is input, the inference engine in the maintenance support system refers to the abnormal event frame of the knowledge base to obtain the cause that can be the abnormal event this time. Next, the inference engine refers to the cause frame in the knowledge base and displays a message on the input / output device asking the maintenance staff whether or not each derived event of each cause is derived. When the maintenance staff answers every question whether or not the derived event is derived, the inference engine refers to the cause frame and tells the input / output device what is the cause of this abnormal event and what kind of countermeasure should be taken. Display and teach maintenance personnel.

As described above, even a maintenance worker who does not fully understand the control system investigates whether or not the derived event indicated by the maintenance support system is derived, and responds to it, so that the inference engine outputs a countermeasure to be taken this time. Moreover, it becomes possible for the maintenance staff to solve a complicated abnormal event by taking the countermeasures.

In addition, even when no abnormality has occurred, when it is desired to know what kind of event occurs when a certain device fails, the maintenance staff inputs the device failure, that is, the cause. The engine refers to the cause frame and displays all derivative events for this cause. This allows the maintenance personnel to know what kind of event will occur due to the failure of a certain device.

[0011]

An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is an overall schematic configuration diagram of a plant control system maintenance support system according to an embodiment of the present invention.

In FIG. 1, reference numeral 1 is a main body of a maintenance support device constructed by using a computer, which comprises a knowledge base 2, an inference engine 3, and a maintenance staff conversation program 4, to which an input / output device 5 is connected. The input / output device 5 has a data input function and a data display function (CRT). The knowledge base 2 has an abnormal event frame 2a and a cause frame 2b in a divided manner. The abnormal event frame 2a describes the first cause to the Nth cause that can be abnormal events, and the cause frame 2b describes the derived events 1 to M of each cause and countermeasures for each cause.

That is, the cause for each abnormal event that can occur on the control system that is the support target, the derived event (when the cause for the abnormal event occurs, a possible event, for example, FAIL lamp lighting of the CPU card, plant trip, etc.) And the countermeasures against each cause are shown in Fig. 2 FT (Fa
(ult Tree) diagram format, and this FT diagram is divided into abnormal event frames and cause frames and registered in the knowledge base 2 as shown in FIG. In addition, the above-mentioned maintenance staff conversation program 4 instructs the maintenance staff 6 through the input / output device 5 to
Controls questions, answers, response measures output, etc.

When an abnormality occurs in the control system, the maintenance staff 6 inputs the content of the abnormal event into the maintenance support device main body 1, and the inference engine 3 in the maintenance support device main body 1 causes the abnormal event frame 2a in the knowledge base 2 to enter. To obtain the cause that may be the abnormal event this time. Next, the inference engine 3 refers to the cause frame 2b in the knowledge base 2 and displays a message on the input / output device 5 to ask the maintenance staff 6 whether or not each derived event of each cause is derived. When the maintenance staff replies whether or not the derived event in question is derived, the inference engine 3
By referring to the cause frame 2b, the cause of the abnormal event this time is displayed on the input / output device 5 and what countermeasure should be taken, and the maintenance staff 6 is instructed.

The inference engine 3 makes inferences using, for example, the MYCIN method, and it is not necessary for the maintenance staff to answer all the questions, and if any of the derived events has occurred completely, That is the cause of this abnormal event.
Therefore, the questions of the derived events are arranged in the order in which the maintenance staff 6 can easily answer. The above-mentioned MYCIN is a product name which was developed by Shortliffe of Stanford University in 1976 and realized the problem of medical diagnosis on a computer. Recently, it has been widely used not only for medical problems but also for abnormality diagnosis of plants. Has been applied. The MYCIN method will be briefly described below.

The normal event is FT as shown in FIG.
(Fault Tree) Can be represented by a diagram. For each cause and derivative event shown in this FT diagram, the weighting factor R (0.0
~ 1.0) is determined empirically and the event evaluation value V
Calculate (0.0-1.0).

The weighting factor Ri (R1, R2, ...) Means the strength of the relationship between the cause and the abnormal event. That is, when the cause event occurs, if this abnormal event always occurs, R
i has a large value.

The weighting factors Rij (R11, R12, ...) Mean the strength of the relationship between the cause and the derived event. In other words, if this derived event always occurs when this cause event occurs,
Rij has a large value.

The event evaluation value V (0.0 to 1.0) means the degree of occurrence of a derivative event, and is usually calculated using the threshold function shown in FIG. Then, when an abnormal event occurs, the certainty factor CF value of each cause is obtained by the following calculation formula, and the cause with the highest certainty factor CF is diagnosed as the cause of this abnormal event.

CFi = Ri (1-Ri1Vi1) (1-Ri2
Vi2) ... (1-RijVij) Next, the control system which is the object of the present invention will be described. FIG. 5 is a block diagram showing a configuration example of a control system for a combined plant, which is a target of this embodiment.

This control system comprises a gas turbine main body 11, a boiler main body 12, a steam turbine main body 13, which are the main machines of a plant, a gas turbine control unit 14, a boiler control unit 15, and a steam turbine which control each of these main units 11 to 13. Control device 16, a plant computer (DAS: Data AcquisitionSy) that collects and manages plant data
stem) 17 and a central control operation panel 18 for an operator to monitor and operate the plant, and exchanges data via a network 19.

The gas turbine controller 14, the boiler controller 15, and the steam turbine controller 16 are respectively D
DC (Direct Digital Controller: Digital control device used) CPU card 20 is provided.
The C CPU card 20 sends the gas turbine main body 1 through I / O based on instructions from the central control operation panel 18 respectively.
1. Control of the boiler body 12 and the steam turbine body 13 is performed. The DDC CPU card 20 has a self-diagnosis function, and when the DDC fails to function (CP
U error, communication error with I / O, etc.), DDC FAIL
Generate an alarm.

Then, the process data of the plant is measured by a sensor and sent to each of the control devices 14 to 16 via I / O, and at the same time, a plurality of remote (RE
MOTE) I / Os 21a to 21n and the HUB 22 are sent to the plant computer 17. The sensor has a fixed measurable range, and when a value outside the measurable range is input from the sensor to the plant computer 17, a RANGE OVER alarm is issued. This R
The cause of the ANGE OVER alarm may be the trouble of the sensor itself or the disconnection of the cable.

The remote I / Os 21a to 21n perform input processing for a large amount of process data in a distributed manner and connect them by the HUB 22 to connect them to the plant computer 1.
It is transmitted to 7. As described above, remote I / O 21a
The load on the plant computer 17 is reduced by performing the process data input processing by 21n. The plant computer 17 receives the captured process data via the network 19 from the central control panel 1
Transfer to 8.

As shown in FIGS. 6 and 7, the central control operation panel 18 has a plurality of control systems including a keyboard 25, a CRT 26, a CRT controller 27, a display controller 28, and the like. FIG. 6 is an external perspective view of the central control operation panel 18, and FIG. 7 is an internal configuration diagram of the central control operation panel 18.

FIG. 8 shows a system configuration of the central control operation panel 18. The CRT controller 27 is a C
The CRT controller CPU card 271 is composed of an RT controller CPU card 271 and a base screen store expansion memory 272.
A color image is displayed on the CRT 26 based on the control command and the image information.

The display controller 28 comprises a display controller CPU card 281, a RAM disk 282 for storing image information, and a communication card 283. The communication card 283 communicates with the plant computer 17 and the control devices 14-16. From the display controller CPU card 281 to the CRT controller CPU card 27 of the CRT controller 27
The control command and the image information are sent to 1.

The central control operation panel 18 is a CRT 26,
A CRT OP. Combining a CRT controller 27 and a display controller 28. It is equipped with a plurality of computers, each of which can independently display the CRT screen requested by the operator. That is, one CRT O
P. Even if the computer becomes abnormal, another CRT O
P. It can be operated on a computer and has no effect on plant operation. Further, the CRT OP. The computer is a control device 14-1
6 or the process data transmitted from the plant computer 17 is displayed on the CRT screen at regular intervals.
This allows the operator to monitor the latest process data on the CRT 26 of the central control panel 18.

In a normal plant, the central control panel 18
Software alarm displayed on CRT26 of
There is a hardware alarm that goes out to the Hardware Annunciator. Normally, in the case of a hardware alarm, when an alarm is generated, first, an alarm sound sounds in the annunciator, and the operator looks at the CRT 26 to confirm the detailed content of the alarm.

In the flow of process data as described above,
An FT diagram of an abnormal event in the control system of the maintenance support object, for example, an abnormal event "accurate process data is not displayed on the CRT screen of the central control panel" is shown in FIG. 9 and FIG.
As shown in 0.

The maintenance support system in this embodiment is
As shown in FIG. 1, the FT diagram relating to the abnormal event is divided into an abnormal event frame 2a and a cause frame 2b as a knowledge base 2 and is held, and the "on the CRT screen of the central control panel" shown in FIGS. Fig. 11 shows an FT diagram related to an abnormal event in which the accurate process data is not displayed on the screen.
Registered in the knowledge base 2 using the abnormal event frame shown in FIG. 10 and the cause frame shown in FIG.

When the abnormal event "the accurate process data is not displayed on the CRT screen of the central control panel" is input to the maintenance support system, the maintenance support system performs the following processing as shown in the flowchart of FIG. Then, the cause and countermeasure are found out, displayed on the input / output device 5 of FIG. 1, and the maintenance staff 6 is instructed.

(1) A maintenance worker inputs an abnormal event occurring in the control system into the maintenance support system (step A
1). (2) The maintenance support system enumerates possible causes by referring to the abnormal event frame relating to the abnormal event input by the maintenance staff (step A2).

(3) The maintenance support system refers to the cause frame from the listed causes and asks the maintenance staff whether or not a derivative event for the cause has occurred (step A).
3).

(4) The maintenance support system finds the cause of this abnormal event after diagnosing the cause based on the reply from the maintenance staff. (5) The maintenance support system refers to the cause frame,
Find out the corrective action related to the cause determined as the cause of the abnormal event this time, and teach the maintenance personnel.

First, a question is asked, "Is it displayed correctly on another CRT and nothing is displayed on this CRT?" (Step A4). If the answer of the maintenance staff is NO to this question, for example, "is it displayed correctly on the other CRT and is the FAIL lamp of the CRT controller (27) of this CRT lit?" (Step A5),
Questions such as "The plant has tripped and the DDCFAIL alarm has been issued" (step A6) are sequentially asked.

If the answer to the question at step A4 is YES, the cause is "abnormal CRT or C
The RT controller is abnormal. Please replace the CRT and CRT controller. "Is displayed on the CRT 26 (step A7). If the answer to the question in step A5 is YES, the message "Cause is CRT controller abnormality, replace CRT controller" is displayed on the CRT 26 (step A8).
If the answer to the question in step A6 is YES, the cause is "DDC CPU card (2
0) is abnormal, please replace the DDC CPU card "is displayed on the CRT 26 (step A9). If all the answers to the questions in steps A4 to A6 are NO, "the cause of the abnormal event this time cannot be estimated" is displayed on the CRT 26 (step A1).
0).

[C] in the above target control system
The flow of software for finding a derivative event and a countermeasure that occur when "RT abnormality" and "DDC CPU card abnormality" occur is as shown in FIG.

When the failure event is "CRT abnormality", the following processing is performed as shown in FIG. (1) The maintenance staff inputs "CRT abnormality" as a failure event to be investigated (step B1).

(2) The maintenance support system searches and refers to the cause frame relating to the failure event "abnormality of CRT" input by the maintenance staff, enumerates and displays the derived events that may occur due to this failure, and responds to this failure. The measure is taught to the maintenance staff (step B2).

As the above-mentioned possible derivative event, for example, "the other CRT is correctly displayed and nothing is displayed for this CRT" is displayed on the CRT 26, and as a countermeasure against the "CRT abnormality", "CRT replacement" is displayed. Is displayed.

When the failure event is "abnormality of DDC CPU", the following processing is performed as shown in FIG. (1) "DDC C
PU card abnormality "is input (step C1).

(2) The maintenance support system searches and refers to the cause frame related to the failure event "DDC CPU card abnormality" input by the maintenance staff, and the failure causes "plant trip, DD
"C FAIL alarm is output" is displayed, and "DDC CPU
As a countermeasure against "card abnormality", "DDC CP
"Replace U card" is displayed (step C2).

In the above embodiment, the case of finding a countermeasure against an abnormal event "process data is not accurately displayed on the central control panel CRT screen" in the combined plant control system has been described. As for the above, a countermeasure can be obtained in the same manner as in the above embodiment.

[0045]

As described in detail above, according to the present invention, when a malfunction occurs in a complex large-scale control system, the cause of the malfunction is determined by inputting the content of the abnormal event into the maintenance support system. On top of that, the corrective action for the cause is displayed, so even maintenance personnel who do not fully understand the control system can investigate the cause of the abnormal event and what countermeasure should be taken for the abnormal event. I understand. In addition, if you want to know what kind of event will occur when a certain device fails even when no abnormality has occurred in the control system, enter the device failure (cause) and enter the derived event for that cause. Are all displayed. This allows the maintenance personnel to learn what kind of event will occur when a certain device of the control system fails, which can be useful for improving the knowledge of the maintenance personnel.

[Brief description of drawings]

FIG. 1 is a system configuration diagram of a plant control system maintenance support system according to an embodiment of the present invention.

FIG. 2 is an FT diagram regarding an abnormal event in the example.

FIG. 3 is a diagram showing a description method of a knowledge base in the embodiment.

FIG. 4 is a diagram showing a threshold value function when an event evaluation value is calculated in the example.

FIG. 5 is a configuration diagram of a combined plant control system to be maintained.

FIG. 6 is a perspective view showing an external configuration of a central control operation panel in the control system.

FIG. 7 is a diagram showing an internal configuration of a central control operation panel.

FIG. 8 is a system configuration diagram of a central control operation panel.

FIG. 9 is a diagram showing a part of an abnormal event FT diagram for a control system.

FIG. 10 is a diagram showing another part of the abnormal event FT diagram for the control system.

FIG. 11 is a diagram showing an example of an abnormal event frame included in the maintenance support system of the present invention.

FIG. 12 is a diagram showing an example of a cause frame of each cause of the maintenance support system.

FIG. 13 is a flowchart showing processing for finding a cause and a countermeasure when an abnormal event occurs.

FIG. 14 is a flow chart showing a process for finding out what kind of derivative event occurs due to this failure when a failure occurs.

[Explanation of symbols]

 1 Maintenance Support Device Main Body 2 Knowledge Base 3 Inference Engine 4 Maintenance Worker Conversation Program 5 Input / Output Device 6 Maintenance Worker 11 Gas Turbine Main Body 12 Boiler Main Body 13 Steam Turbine Main Body 14 Gas Turbine Control Device 15 Boiler Control Device 16 Steam Turbine Control Device 17 Plant computer 18 Central control operation panel 19 Network 20 DDC CPU card 25 Keyboard 26 CRT 27 CRT controller 28 Display controller

Claims (2)

[Claims] 1. An input / output device having a data input function and a display function, control means for controlling the input / output device with respect to questions, answers, corresponding treatment output, etc., and a target control system. A knowledge base having an abnormal event frame in which causes for possible abnormal events are described and a cause frame in which countermeasures for possible derived events and causes when the causes occur are described, and input from the input / output device. Based on the abnormal event or cause, the knowledge base is referred to, and the inference engine for displaying the countermeasure against the input abnormal event or the derivative event for the input cause on the input / output device is provided. A characteristic plant control system maintenance support system. 2. An input / output device having a data input function and a display function, control means for controlling the input / output device with regard to questions, answers, corresponding treatment output, etc., and a target control system. A knowledge base having an abnormal event frame in which causes for possible abnormal events are described and a cause frame in which countermeasures for possible derived events and causes when the causes occur are described, and input by the input / output device. For the abnormal event, refer to the abnormal event frame of the knowledge base to search the cause that may be abnormal, refer to the above cause frame based on this cause, and check the existence of the occurrence event for each cause to check the abnormal event. A maintenance support system for a plant control system, comprising: an inference engine that displays countermeasures against the above in the input / output device.