JP2770228B2 - Knowledge-based plant information processing system - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a knowledge-based plant information processing system for a knowledge-based plant control device which is a combination of a (real-time) knowledge processing system and a control system based on a procedural program. More particularly, the present invention relates to a knowledge-based plant information processing method suitable for use in a plant control device that realizes more intelligent functions such as optimal control of a plant, control at the time of abnormality, predictive control, and plant diagnosis.

(Prior Art) Inferring or identifying the cause of abnormality in various plants, etc., using a knowledge base obtained by accumulating, classifying, and organizing the knowledge and experience of experts in the relevant field is disclosed in, for example, 60
No. -14303. Here, a cause / effect rule group, a result / cause rule group, and a cause estimation rule group are prepared as a knowledge base, and when an abnormal sign is detected in the collected plant information, the aforementioned rule group is set. The cause of the abnormality of various plants is inferred or identified based on the above.

A procedural programming control system that performs control using a conventional procedural programming language using the above-described knowledge base, equipment diagnosis, abnormality detection, operator guidance (instruction to the operator) at the time of abnormality, and / or automatic control It is conceivable that a knowledge processing system having an inference function for performing a protection operation is organically coupled to quickly and accurately control an abnormal situation in a plant or the like.

As described above, the knowledge base expresses procedures and methods for knowledge processing by software based on the knowledge of experts in the field and empirical know-how.

In other words, this knowledge base organizes the knowledge about the prediction and detection of abnormalities and the method of processing abnormalities obtained through the experience of the expert in the past in the form of If…, then… and registers as rules. In addition, the data required by the rule is registered in the form of a frame.

The inference function is for inferring the cause or countermeasure for the abnormal symptom based on the knowledge base.

The inference result can be used in various ways, such as being used as a CRT output to an operator or a print output to a typewriter, or transferred to another system as data. In any case, the result obtained by the inference function can be handled as information by another system or human.

When a knowledge processing system having such a knowledge base and an inference function is used for online control, the following configuration or method can be considered.

That is, data obtained from a process input / output device of a plant or a control device of another system is taken into the knowledge processing system, and abnormality or a sign of abnormality of the plant is detected based on the knowledge described by the rules in the knowledge base.

When an abnormality or an abnormal sign is detected, an inference function for inferring how to deal with the abnormality or the abnormal sign based on rules and frames in the knowledge base is activated.

The conclusion obtained by the inference function is displayed as an instruction to an operator, or transmitted as it is to the procedural programming control system or another control system, and used as a command for online automatic control.

An outline of such a knowledge type plant information processing apparatus will be briefly described with reference to Japanese Patent Application No. 62-109176 proposed earlier by the present inventors. FIG. 5 is a schematic block diagram thereof.

This plant information processing apparatus includes a plant control system unit 10 and a knowledge processing system unit 20 using a conventional procedural programming system.

The plant control unit 10 according to the conventional procedure normally uses the terminal 36
-Machine function unit 11 that provides information to the user (operator) through the data processing and data collection / processing unit that collects information from the target plant 34 via the process input / output device 32
12, a plant database 13 for storing collected data, a plant database creating unit for organizing and editing data stored in the plant database 13.
14 and an SCC (Supervisory Computer Control) that controls these units.

Further, the plant control unit 10 includes a monitoring function unit 17.

The monitoring function unit 17 converts, based on the expert knowledge and know-how obtained empirically in the target plant, knowledge on signs that may cause a plant abnormality, knowledge and procedures for determining an abnormal state into a procedural program. It is rewritten and stored, and based on the plant data obtained by the data collection / processing unit 12, the occurrence of a plant abnormality or abnormal sign is detected.

On the other hand, the knowledge processing system unit 20 includes knowledge, that is, a knowledge base 21 for storing the above-described rules and frames, an inference function unit 22 for inferring a measure or means for avoiding or recovering a plant abnormality using the knowledge, and an inference result. Or, it is composed of the man-machine function unit 23 which provides information related thereto to the user via the intelligent terminal 38.

Next, the function and operation of each unit will be described while following the flow of information.

Information on the target plant 34 is converted into data by the process input / output device 32, and stored and managed in the data collection / processing unit 12. Specifically, analog information and digital information of the plant are digitized in the process input / output device 32,
It is temporarily managed in the data collection / processing unit 12.

Some information is provided to the user by the man-machine function unit 11. There are various forms of provision in this case,
A typical one is to display the information on a CRT usually arranged in the terminal 36 or to print out the information using a typewriter or the like.

Some information is stored in the plant database 13. The processing of the stored data is performed by the plant database creating unit 14, and is stored in an auxiliary memory (such as a fixed disk) as historical data or the like.

Furthermore, some information is used as plant control data, is output to the process input / output device 32 via the controller of the SCC 15, and controls the target plant.

Next, the monitoring function unit 17 applies the plant data (frames) collected by the data collection / processing unit 12 to the knowledge and the procedure (rule) for determining the occurrence of a plant abnormality or an abnormal sign stored in the plant in advance. By applying this, the occurrence of abnormalities or abnormal signs of the plant is detected.

For example, if the knowledge indicating the sign of abnormality is that "if the number of NO.1 pumps is 3600 rpm or more, there is a great possibility of adversely affecting the plant", the latest information on the number of revolutions of the NO.1 pump should be updated periodically. Data collection and processing unit 12
Read out from the management table (not shown) in it and check whether the data is 3600 rpm or more.

If the number of revolutions of the NO.1 pump indicates 3600 rpm or more for a longer period of time than the predetermined (specified) time, the monitoring function unit 17 outputs a detection signal of "abnormal number of revolutions", that is, an event. Issue

The above-mentioned event for notifying that “NO. 1 pump speed is abnormal” is supplied to the inference function unit 22 of the knowledge processing system unit 20. The inference function unit 22 selects knowledge (rules) related to the number of revolutions of the No. 1 pump in accordance with this event, and executes inference on them.

The results obtained by this inference are used as operation guidance for the operator, and
The information is displayed on the high-function terminal 38 via the 23 (CRT display or typewriter printing), or supplied to the data collection / processing unit 12 as a control output to the target plant 34.

In the above specific example, the result of the inference is “stop the operation of the NO.1 pump”, and “respond simultaneously with the NO.2 pump”.

This instructs the plant control unit 10 to output an output signal for stopping the operation of the NO.1 pump and, at the same time, to output an output signal for starting the NO.2 pump, in accordance with the rule description method. You.

Further, according to the rule description method, it is possible to make the inference result a guidance to the operator, and it is also possible to perform both of them at the same time.

The above-described monitoring function is realized by creating a monitoring program for each facility and each device, and activating and executing each program at a predetermined cycle.

And, as described above, when an abnormality or an occurrence of an abnormal sign is detected for any of the facilities or devices,
Requests the knowledge processing system to execute inference for each facility and equipment.

For this reason, a plurality of inference execution requests to the inference function unit 22 may occur at the same time, but the inference function unit 22 can also infer the complex detection processing by comprehensively viewing these requests. .

For this purpose, a rule that is described in the form of If ..., then ... includes in advance an If statement that includes a combination of abnormalities or abnormal signs of each facility or device for the complex detection. It is natural that they must be.

By the way, in the procedural programming control system,
As is well known, it is necessary to define a desired control operation or procedure using a predetermined programming language as a continuous and fixed series of contexts. In the system, the rules and the files that represent the relationship between cause and effect in the form of If…, then… and the data for that are arranged independently and randomly. Completely different. The expression format of data (information) in each system is also different. For this reason, conventionally, the construction of each of these systems has been performed completely independently.

(Problems to be Solved by the Invention) As described above, in a knowledge-type plant control apparatus in which a real-time knowledge processing system and a control system based on a procedural program are organically combined and coexist, these two are used.
In order to construct two systems, there is a problem that the construction methods and data expression formats are different from each other, so that these must be performed independently.

In other words, these two systems have the purpose of performing control, monitoring, abnormality processing, preliminary detection, diagnosis, etc. on the same control target, and providing information relating to them to the user or other systems. The data (plant data base and knowledge base) required to perform the process are the same, but separate systems suitable for each system (knowledge processing system and procedural programming control system). It was necessary to give these data in the form of expression.

As described above, although the data required by the knowledge-based plant control device is common, there is a difference in the expression format of the data in each system configuring the knowledge-based plant control device. The man-hours, time, and man-hours required for construction of all systems are the same as when two systems (a knowledge processing system and a control system) are individually constructed, and the man-hours, time, and man-hours are extremely large. .

Furthermore, since the representation format of substantially the same data is changed and given to each system, humans intervene there, and the reliability of processing and management of data is necessarily reduced. There was a problem.

An object of the present invention is to provide a knowledge-based plant information processing system capable of preventing information reliability from being reduced by reducing the number of man-hours, time and labor required for system construction, and reliably managing information (data to be used). Is to do.

(Means for Solving the Problems) In the present invention, substantially the same data is converted into two systems (a knowledge processing system and a procedural programming control system) constituting a knowledge-type plant control device by changing the expression form. In order to minimize the double effort of separately applying the information to each knowledge base and plant data base, the number of construction steps is reduced by the following means.

That is, in the present invention, the information used (defined) used in the procedural programming control system is used as it is, and this information is automatically changed to a representation format according to the grammar of the knowledge processing system. (Hereinafter referred to as a knowledge base automatic generation function), the man-hours, labor, and time required to construct the two systems are reduced as much as possible.

In this knowledge base automatic generation function, the range of the generation scale of the knowledge base can be determined, such as the whole or a part (for example, a skeleton part) of the knowledge base, according to the designation of the user.

(Operation) When constructing a procedural programming control system,
The user usually defines the plant data by a blank entry method (FIG. 2) of the TAG (process variable) processing specification. That is, the input capture function, optical value conversion function, monitoring function, control calculation / control output function, recording / guidance function, and the like are determined in detail.

Therefore, the user decides whether or not these data are data (= TAG) used in the knowledge processing system in a unit of TAG by this blank entry method.

And for the TAG selected for this,
The expression format is changed by the knowledge base automatic generation function, and the data is automatically constructed as a frame in the knowledge base handling data in the knowledge processing system.

Like the above TAG information, information determined in a certain format can be used for knowledge base frames (including slots and slot values) by the knowledge base automatic generation function if it is determined that it is necessary for knowledge processing. And a hierarchical representation.

Embodiment FIG. 1 shows a schematic configuration of an embodiment of the present invention.

1 includes a procedural programming control system 10 including a control computer (for example, HIDIC V90 / 25) equipped with a process control system (for example, HIDACS), and a knowledge processing software (for example, EUREK).
A) and a knowledge processing system 20 including an engineering work station (cpu: for example, ES 330) or a control computer (HIDICV90 / 25). In FIG. 1, the same reference numerals as those in FIG. 5 represent the same or equivalent parts.

In the procedural programming control system 10, the information of the target plant 34 is
(Process variable). With this registration,
The plant information is stored and managed in the plant database 13.

The registration of the TAG to the control system is performed by using a standard screen (illustrated in FIG. 2) provided by the control computer in the control system 10 and filling in the required information (hereinafter referred to as TA
G processing specification definition) is input from an input device (keyboard) in the man-machine function unit 11.

More specifically, the TAG processing specification definition for constructing the procedural programming control system is defined in the second section.
This is performed by inputting necessary information into blanks (lower dotted line in the figure) of the TAG screen illustrated in the figure.

In this example, tag name, engineering unit definition, digital input / output status definition (open, close, run, stop
), Scale conversion type specification (square root conversion, linear conversion, linear approximation conversion, thermocouple conversion, etc.), instrument check limit value (input signal range, upper limit, lower limit), zero adjustment value, output specification definition (output Over several tens of items such as conversion type, output change rate limit value, output lower limit value, output upper limit value, etc., etc., and detailed specifications for each are defined. The left half of FIG. 3 is an example of a TAG for which definition processing has been performed, which will be described later.

At this time, whether the registered TAG is information necessary for inference for control, monitoring, abnormality processing, preliminary detection, and diagnosis performed in the knowledge processing system 20, that is, as a knowledge base 21 of the knowledge processing system. At the same time, whether to be registered is designated and input in the same blank entry method (although the designated blank for this is not shown in FIG. 2).

TAG specified to be registered in Knowledge Base 21
Is transferred to the knowledge base automatic generation function unit 24.

Based on these TAGs (information), the knowledge base automatic generation function unit 24 changes the TAG information into an expression format according to the grammar of the knowledge processing system (see the right half of FIG. 3),
TAG names are expressed in units called frames.

A frame is a representation of a certain entity or concept, a logical model created to represent a complex object from various perspectives, and is used as one method of expressing knowledge in a knowledge processing system. Is what it is.

Further, the frame has a slot and a slot value in order to represent in detail the object (concept or entity) defined by the name of the frame. Slots represent attributes or characteristics,
This is a more detailed representation of the frame name. The slot value expresses a value representing a slot or a state.

FIG. 3 is a diagram showing a difference or a correspondence relationship between representation formats of data (plant information) in the procedural programming control system 10 and the knowledge processing system 20. The left half 101 of FIG. 6 is an example of TAG processing specification definition data to be performed.

In the example of FIG. 3, TAG (name process variable PV) is T101A
It is. The "T" at the beginning of T101A indicates that this TAG is related to temperature.

The engineering unit (ENG) in this example specifies ° C and specifies (EX
P) is 0, in engineering units of integrated data, this TAG does not perform integration, there is an alarm suppressor (SA), the sampling period (SP) is 1 second, and the zero adjustment value (ZRO) is 3.

The scale conversion type (STYP) is resistance temperature detector conversion (SRC), and the input signal lower limit (VL) is defined as 0 ° C, the input signal upper limit (VH) is defined as 50 ° C, and so on. I will do it.

The user determines that the TAG information defined in this way is necessary for the knowledge processing system, and uses the screen shown in FIG. In 24, as shown by the reference numeral 201 in the right half of FIG.

In the example of FIG. 3, the frame name is T101A. In addition, the attributes and characteristics of the slot, ie, T101A, include engineering units, exponents, integrated data engineering units, alarm suppression, sampling period, scale conversion type, input signal lower limit, input signal upper limit, and zero adjustment.

Next, a slot value, that is, an attribute value indicating an attribute, a state value indicating a characteristic, or the like is expressed as shown in 201 in the right half of FIG.

Used when performing inference of knowledge processing systems,
The data common to the procedural programming control system is the current value (instantaneous value) of the plant, historical data (past value) stored in the plant database, and numerical analysis result data analyzed based on the historical data. These data are stored in the knowledge base 21 as a frame for each TAG, in a different expression form by the same method as described above.

Incidentally, the frame of the knowledge base 21 of the knowledge processing system 20 can basically have a hierarchical structure.

As mentioned earlier, the first character (symbol) of the TAG name
Is, for example, the temperature is T,
Basic pressure is P, flow rate is F, etc.
Is determined by

Therefore, the knowledge base automatic generation function unit 24 classifies TAGs into levels such as temperature, pressure, and flow rate by judging the TAG name, and further generates a frame that is an actual lower level in a hierarchical structure. Can be.

As the hierarchical structure of the frame goes from the upper level to the lower level, it expands from "abstract concept to concrete concept" and "from whole to detailed". By providing such a hierarchical structure, maintainability and expandability can be improved.

As described above, a process (operation) of importing the concept of TAG, which has been conventionally handled by the procedural programming control system 10 into the knowledge processing system 20, and re-editing it into a usable expression form (operation) By the automatic generation function unit 24 of the knowledge base, the operation Know-How of the operator involved in the operation of the plant, the abnormality diagnosis Know-How of the plant, the preliminary detection Know-How, etc., using the concept of the TAG, It becomes possible to arrange and describe rules in the form If…, then….

At the same time, the plant data in the procedural programming control system 10 can be easily referred to in the knowledge processing system 20.

As an example of rule creation, If (the current value of T101A is 120 or more and the plant trend is increasing) and (the current value of F102 is 50 or less), then (the setting value of F102 is 60) (Set the valve opening of F102 large). This example uses the knowledge base T101A TAG
A new set value is set to the TAG of F102 according to the data included in the frame and the state of the TAG frame of F102.

The “plant tendency” in the If statement is the T101A T
This is to analyze the historical data (past values) of AG and see if the slope of the primary regression is positive.

Then, the inference function unit 22 displays the conclusion “increase the opening of the F102 valve” described in the section of then on the screen of the man-machine function unit 23 via the abnormality detection / process diagnosis unit 25. At the same time, the data is transferred to the SCC 15 of the control system 10 to execute necessary control (change of set values).

FIG. 4 is an example of a flowchart showing the operation of the knowledge base automatic generation function unit 24.

First, it is confirmed whether or not the registration TAG of the target is designated to be registered in the knowledge processing system 20 from the plant information definition table of the procedural programming control system 10 (401).

If registration is specified, read the TAG name (4
02), a new frame is created in the knowledge base of the knowledge processing system 20 using the currently read TAG name as a frame name (403).

Subsequently, data is read from the table in which the information is stored, changed to the expression format of the knowledge processing system 20, and all necessary storage information is defined one after another in the frame (405, 406).

By executing such a process for all the registered TAGs, the whole or a part of the knowledge base of the knowledge processing system 20 is generated.

As described above, TAG names are defined in JIS by the basics of naming, so imported TAGs can be roughly classified by referring to the initials of TAG names. Can be used to determine the hierarchical structure of the frame as shown in the right half of FIG.

(Effect of the Invention) According to the present invention, the provision of the knowledge base automatic generation function unit enables the TAG information and the plant information generated and defined for the construction of the procedural programming control system to be transmitted without human intervention. Since it can be automatically converted to TAG information for building a knowledge processing system, it is possible to maintain the reliability of the definition information in the knowledge base, and further reduce the man-hour and time required for building the knowledge processing system. .

[Brief description of the drawings]

FIG. 1 is a schematic block diagram of one embodiment of the present invention. FIG. 2 is an example of a blank entry method standard screen for plant information definition (TAG definition) for constructing a procedural programming control system. FIG. 3 is a diagram showing a comparison between a representation format of plant information of the procedural programming control system and a representation format of plant information of the knowledge processing system. FIG. 4 shows a flowchart of the knowledge base automatic generation function. FIG. 5 is a block diagram of a knowledge-type plant control device proposed earlier by the present inventors. 10: Plant control unit, 11, 23: Man-machine function unit, 12: Data collection / processing unit, 13: Plant database, 14: Plant database creation unit, 15: SC
C, 20: Knowledge processing system section, 21: Knowledge base, 22
…… Inference function part, 24 …… Knowledge base automatic generation function part, 25
…… Abnormality detection / process diagnosis section, 32 …… Process input / output device, 34 …… Target plant, 36 …… Normal terminal, 38 …… Advanced terminal

Continuation of front page (58) Fields investigated (Int.Cl. ⁶ , DB name) G05B 23/02 G06F 17/00 G06F 9/44

Claims (6)

(57) [Claims] 1. A data collection / processing unit for collecting and processing data obtained from a target plant, a plant data base storing the collected data, and a state in the target plant based on the data. And a knowledge base storing knowledge about the target plant, based on a request from the monitoring function unit, and using the knowledge in the knowledge base to detect the target plant. A knowledge-type plant information processing method for a plant control apparatus including a knowledge processing system unit including a reasoning function unit that infers a countermeasure for a given state and outputs the inference result to at least one of a terminal and a plant control unit The plant information stored in the plant database is
When the plant information is also stored in the knowledge base when inputting in the G format, an instruction to that effect is further input, and the plant control unit transmits the input plant information to the knowledge base.
The information is stored in the plant database in the TAG format, and the knowledge processing system unit converts the plant information instructed to be stored into an expression format in which the TAG name is a frame name according to the grammar of the knowledge processing system. A knowledge-based plant information processing method characterized by being stored in a base. 2. The knowledge-type plant information processing method according to claim 1, wherein the TAG names in the plant database and the knowledge base are the same for the same plant information. method. 3. The knowledge stored in the knowledge base is If, then
2. A knowledge-based plant information processing system according to claim 1, comprising a rule described in a format of ... and a frame for storing data applied to said rule. 4. A knowledge type plant information processing system according to claim 1, wherein the inference result is displayed on a terminal via a man-machine function unit. . 5. The knowledge according to any one of claims 1 to 3, wherein the inference result is supplied to a plant control unit and used for automatically controlling a target plant. Type plant information processing system. 6. The method according to claim 1, wherein the knowledge stored in the knowledge base is a measure corresponding to a detection state of the target plant, obtained based on expert knowledge and experience in the field. Item 6. The knowledge-based plant information processing method according to any one of Items 1 to 5.