JPH08152906A - Information processor and control program design supporting device - Google Patents

Abstract

PURPOSE: To provide an information processor, which is equipped with a function for automatically reading the intension of an operation on a screen and executing the intended operation, and a control program design supporting device which can be easily used. CONSTITUTION: This device is provided with input means 21 and 22 equipped with a screen to display an object capable of operating the screen, knowledge holding means 10 equipped with an ontology 12 on this object to be displayed on the screen and grammar 13 for user intension analysis as knowledge described by using vocabulary of the ontology 12, analyzing means 14 for analyzing the intension of the operation to the screen by using the grammar 13 for user intension analysis, and action executing means 15 for executing the operation corresponding to the analyzed result of this analyzing means 14.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an information processing apparatus for processing various kinds of information in accordance with visual instructions and a control program design support apparatus for automatically generating control specifications in accordance with visual instructions.

[0002]

2. Description of the Related Art A programmable controller for controlling various plants requires a control execution program for effectively operating it.

In recent years, it has become possible to automatically generate a control execution program. As a result, the design specification design work, which is an input to the control execution program automatic generation device, is attracting attention as a support target in the program creation process.

As a method of designing control specifications, a method of indicating the flow of the operation of the equipment in the order of execution is generally used in the form of a flow chart. In addition to this method, in order to interpolate the flow chart type representation at the time of input and the image of the device operation, the target device and the processing material are schematically represented on the display by using a figure, and the figure is operated. Therefore, a method of visually describing control specifications has also been proposed.

As a visual instruction method, there are an instruction method of an operation level of a level similar to a flow chart format and an instruction method of a state instruction level in which an instruction level is advanced by using planning with a background of a state space model. Proposed. In the method using a state space model as a background, as shown in Japanese Patent Application No. 4-328997, the state of the entire equipment to be controlled and the state of the equipment constituting the equipment are included, and the equipment between each state is A model that represents the state space of equipment is used by associating each state of equipment with the secondary state transition that expresses the behavior of the equipment and the interaction between the equipment and the connection relationship with the state of the constituent equipment.

The former method facilitates the instruction of the operation of the controlled object and the state in which the operation ends, and the latter method facilitates the setting work of the initial state and the target state in particular. It greatly affects the ease of use of the entire device.

The most common visual instruction method is to use a button or menu. On the other hand, as disclosed in JP-A-5-7325,
There is also proposed a method of giving a visual instruction by using a direct operation for a target graphic, such as drag and drop, without using a menu. In the latter method, the behavior of the system depending on the combination of figures to be operated is collectively defined in a tabular form.

However, in these conventional visual instruction methods, for example, methods using buttons, menus, etc.,
There is a problem that the selection work of a menu and the like becomes a heavy burden on the user when the number of operations increases. On the other hand, the visual instruction method based on the direct operation on the figure has a relatively smaller operation burden on the user than the case of using the menu or the button. However, since the meaning of similar operations often differs depending on the designer, and the behavior of the system associated with the operations also differs depending on the user, there is a problem in that the degree of ease of use varies significantly among users.

Further, regarding a rule for connecting an operation and a behavior, a graphic user interface (GU) is used.
The designer of I) enumerates all combinations of figures and the like in the GUI and collectively defines the corresponding behaviors in a tabular form. This work is gradually designed as the GUI becomes complicated. Is a heavy burden on the person.

Further, in the case of performing an operation using a graphic,
How the system (control program design support device) interprets the operation performed by the user cannot be understood by the user without looking at the behavior of the system. In this way, since the operation error of the user becomes clear after the behavior of the system, it is easy to cancel or redo the behavior, and there is also a psychological effect on the user due to the incorrect operation, thus improving work efficiency. It was the cause of the decline.

On the other hand, Japanese Patent Application No. 4-328997 incorporates the concept of a scope in order to improve the efficiency of planning. The scope has a basic scope and a search scope. The basic scope is expressed by describing which equipment participates in each work for some work purpose of the controlled object, and has a tree structure. The search scope indicates which device participates in the control to reach the target state designated by the designer, and is represented by a list of devices. It is created from the basic scope when performing the state search in the state space.

According to Japanese Patent Application No. 4-328997, one or a small number of works is performed when there are not many devices that participate in a plurality of works in duplicate and when there is not much change in the physical configuration in control. A solution can be efficiently obtained to generate a control specification that fits in units. However, when there are many devices that participate in many tasks, or when the control from the initial state to the target state includes many operations of multiple devices, that is, the designer wants control that spans several tasks. Is required, it is difficult to narrow down the search range efficiently.

Further, when the designer requests a control that spans a number of tasks, there is a possibility that all the tasks between the initial state and the target state may not match, and the devices that should participate in the control are not always required. There was a possibility that all could not be selected.

Further, there is no way to inform the designer whether or not a huge search range is required to guide the control desired by the designer. In addition, in order to prevent an infinite search from being performed endlessly if the search process fails, a method of limiting the time or the length of the searched path is generally used. Depending on the setting of, there was a problem that the search was aborted even if it did not fail, or it took a very long time or computer resources to notice the failure.

[0015]

As described above, the target device or the processing material is schematically represented on the display by using a figure, and the control specification can be visually described by operating the figure. In the conventional control program design support device, there is a problem in that the ease of use is extremely deteriorated as the number of operations increases, which imposes a heavy burden on the user.

Further, in the control program design support apparatus which employs the above-mentioned visual instruction method against the background of the state space model, the control from the initial state to the target state includes many operations of a plurality of devices. In some cases, it has been difficult to efficiently obtain a solution for control specifications.

Therefore, a first object of the present invention is to provide an information processing apparatus having a function of automatically reading an intention of an operation on a screen and executing an intended operation. Further, the present invention provides a control program design support device which applies the above information processing device to automatically generate an initial state and a target state of a display target and generates a control specification based on the initial state and the target state. The second purpose is to do so.

Further, the present invention provides a control program design support device which can generate a control specification based on a state space model as a background, for example, based on an operation on a screen and can efficiently obtain a solution of the control specification. This is the third purpose.

[0019]

In order to achieve the first object, the invention according to claim 1 comprises a screen for displaying an object, an input means capable of operating the screen, and the screen. Knowledge holding means having knowledge of an ontology related to the displayed object and a user intention analysis grammar described using the vocabulary of the ontology, and an intention of an operation on the screen are analyzed using the user intention analysis grammar. The gist is an information processing apparatus including an analysis unit and a behavior execution unit that executes an operation according to an analysis result of the analysis unit.

In order to achieve the above first object,
The invention according to claim 2 includes a screen for displaying an object, an input means capable of operating the screen, an ontology for the object displayed on the screen, and a user intention analysis described using the vocabulary of the ontology. A first knowledge holding means having knowledge of a working grammar as knowledge, an analyzing means for analyzing an intention of an operation on the screen by using the user intention analyzing grammar, and an operation according to an analysis result of the analyzing means. The behavior executing means, the second knowledge holding means having knowledge as a confirmation sentence generating grammar described using the vocabulary of the ontology for generating the analysis result explanation sentence of the analyzing means, and the behavior executing means Before executing the above, a description sentence of a matter related to the behavior of the behavior executing unit is generated using the grammar of the second knowledge holding unit, and the generated explanation sentence is described above. The information processing apparatus having a behavior confirmation means for notifying via the force means is a spirit.

It is preferable that the ontology related to the target also includes the ontology in the state of the target.

In order to achieve the second object,
The invention according to claim 4 comprises a screen for displaying a controlled object or a processing material and an input means capable of operating the screen, an ontology relating to the controlled object or the processing material displayed on the screen, and a vocabulary of the ontology. A knowledge holding unit having knowledge of a user intention analysis grammar described by using, and an intention of an operation on the screen are analyzed using the user intention analysis grammar, and an initial state of a display target is displayed based on the analysis result. Means for generating a goal state and
The gist is a control program design support device provided with a means for generating a control specification based on the initial state and the target state obtained by this means.

Further, in order to achieve the above second object,
The invention according to claim 5 comprises a screen for displaying a controlled object or a processing material, an input means capable of operating the screen, an ontology relating to the controlled object or the processing material displayed on the screen, and a vocabulary of the ontology. First knowledge holding means having knowledge of the user intention analysis grammar described by using the user intention analysis grammar, the intention of the operation on the screen is analyzed using the user intention analysis grammar, and the display target of the display target is analyzed based on the analysis result. A means for generating an initial state and a target state, a means for generating a control specification based on the initial state and the target state obtained by this means, and an explanation sentence as a result of analyzing the intention of the operation on the screen. To generate a confirmation sentence generating grammar described using the ontology vocabulary as knowledge, and before generating the control specification. A control program design support device comprising: a confirmation means for generating an explanation sentence of a matter related to the analyzed result using the grammar of the second knowledge holding means and for notifying the generated explanation sentence via the input means. It is a summary.

It is preferable that the ontology related to the controlled object or the processing material also includes the ontology of the state of the controlled object or the processing material.

Further, in order to achieve the third object,
According to a seventh aspect of the present invention, in a control program design support device that obtains a control procedure by performing a search on a model of a state space of a control target, the devices in the control target are grouped and the A knowledge holding means having knowledge as a matching condition for determining whether a device group is effective, and a scope for limiting a search range from a group of effective devices by using an initial state and a target state of a control target. And a scope determining means for determining.

Further, in order to achieve the third object,
According to an eighth aspect of the present invention, in a control program design support device that obtains a control procedure by performing a search on a model of a state space of a control target, the devices in the control target are grouped, and depending on the state of the control target, Knowledge holding means provided with matching conditions for determining whether the device group is effective, means for determining the order in which the device group becomes effective by using the initial state and the target state of the controlled object, A scope determining / switching unit is provided that switches a group of valid devices from the searched control target state, determines a scope that limits the search range based on the group of valid devices, and switches the scope.

It is preferable that the apparatus further comprises an estimation means for estimating the calculation amount required for the search from the factors that determine the calculation amount for the search before starting the search.

Further, it is preferable that the apparatus further comprises a prediction means for predicting a failure of the search process from the order in which the group of devices determined by using the initial state and the target state becomes effective.

[0029]

Action Ontology, in philosophy, means ontology, but in artificial intelligence it means a set of minimum units (primitives) that are the basis for expressing things. Furthermore, in the field of artificial intelligence knowledge processing, it is usually used as a vocabulary for describing knowledge in a knowledge base.
Each vocabulary included in the ontology corresponds to the concept included in the target domain or task. Just as a concept has a hierarchy based on upper and lower relationships, the ontology vocabulary also has a hierarchy based on upper and lower relationships.

The specific actors that define the ontology are usually the classes, relations, functions, and functions incorporated in the concepts of object-oriented and frame systems.
You can give things like objects and constants. In the embodiment described later, the ontology is defined as a class based on the concept of object-oriented programming and its hierarchy, or an object-oriented database (OO).
Although the description will be given assuming that it is expressed using the schema of (DB), it is not always necessary to stick to the concept of object-oriented programming as long as it achieves the same purpose.

Since the information processing apparatus according to claims 1 and 2 has the grammar using the ontology and the vocabulary of the ontology regarding the display target, the GUI designer is limited as long as the display target is the one in the ontology category. Is G
A desired GUI can be created only by designating which concept of the ontology the display target belongs to, without creating a rule for connecting the graphic operation and the behavior each time the UI is created. Even if the user interface designer or user is different, the meaning of the similar operation is interpreted and executed in the same manner, so that the connection between the operation and the behavior is easily accepted by any user, and even if the number of operations is large. The operation does not burden the user.

Further, in the control program design support apparatus according to the fourth and fifth aspects, a rule connecting the graphic operation and the state change of the control target intended by the user is created every time a screen for displaying the control target is created. Even if it is not created, a desired GUI can be created simply by designating which concept of the ontology the controlled object represented by the graphic belongs to, and thus the creation efficiency of the GUI can be improved. In addition, even if the number of operations is large, the operation does not burden the user, and the meaning of the similar operation is the same even if the designer or the user is different, and the relationship between the graphic operation and the state change of the controlled object is not determined. It will be easily accepted by users.

Further, in the control program design support apparatus according to claims 7 and 8, when there are many devices that participate in a number of tasks, or when the designer requests control over a number of tasks. In addition, since the search range can be narrowed down sufficiently, an efficient search can be performed. Further, even when the user demands a control over a number of tasks, the device that should participate in the control is selected without leaking, so that the search failure can be prevented. In addition, it is possible to inform the user whether or not a huge search range is required to guide the control desired by the user before starting the search, so that careless large-scale search is prevented. be able to.

[0034]

Embodiments will be described below with reference to the drawings.

FIG. 1 is a block diagram of the information processing apparatus according to the first embodiment of the present invention, which operates by interpreting a visual operation of a user.

Further, FIG. 2 shows a block diagram of a control program design support apparatus according to the second embodiment to which the information processing apparatus shown in FIG. 1 is applied. The device shown in FIG. 2 is a device related to a device for supporting the specification creation of a control program by planning as dealt with in Japanese Patent Application No. 4-328997, and the target state at the time of planning is visually checked. It is a device that is designed to be obtained from a manual operation.

As will be understood from the description given later, the knowledge base provided in the apparatus shown in FIG. 2 has the same basic configuration as the knowledge base provided in the apparatus shown in FIG. The elements are specialized for the purpose of obtaining the target state from the visual operation when planning. Further, the portion of the apparatus shown in FIG. 1 that performs the processing has basically the same configuration as the portion of the apparatus that performs the processing of the apparatus shown in FIG. It is a specialized device for obtaining visual operations.

As can be seen from this, by applying the device shown in FIGS. 1 and 2, it is possible to construct a device for widely interpreting and operating the visual operation of the user. In particular, a device that handles object knowledge about the operation of a system that has an object (or object, agent) attribute and its behavior and state that can be expressed in a clear form that can be handled by concepts such as object-oriented or agent-oriented. Can be easily applied to.

The information processing apparatus shown in FIG. 1 which operates by interpreting the visual operation of the user is roughly divided into a knowledge base 10 and a processing system 20.

The knowledge base 10 includes at least display data 11, ontology 12, user intention analysis grammar 13,
Behavior dictionary 14, confirmation sentence generation grammar 15, display target state 1
6. Have knowledge of display target 17. The knowledge held by the knowledge base 10 includes an input / output control unit 22, a user intention analysis unit 23, a user intention confirmation unit 24, which will be described later, in the processing system 20,
It is referred to by the behavior execution unit 25 and the like.

The display data 11 is data such as graphic data to be displayed which the apparatus of this embodiment handles. At least a part of the display data 11 is expressed using the vocabulary of the ontology related to the display target and the state of the display target, and the input / output control unit 2 which will be described later along with the change of the display target and the state thereof.
2 can switch the display data.

The ontology 12 is an ontologyized knowledge about the display target and its state. The ontology 12 is used to represent the display data 11, the user intention analysis grammar 13, the confirmation sentence generation grammar 15, the behavior dictionary 14, the display target state 16, the display target 17, and the like, and the user intention analysis unit described later. 23 and the user intention confirmation unit 24.

The user intention interpretation grammar 13 is grammar knowledge for analyzing the user's intention from an instruction given by the user via the input / output device 21 described later. When the grammatical rule applied to the interpretation is determined, the grammar 15 for generating the confirmation sentence and the behavior dictionary 14 can be searched using the rule.

The behavior dictionary 14 is knowledge of behaviors to be executed by the behavior execution unit 25, which will be described later, based on the result of interpreting the user's instruction.

The confirmation sentence generation grammar 15 is grammatical knowledge used for generating a sentence for confirming to the user the result of interpreting the user's instruction.

The display target state 16 is knowledge about the state of the display target, and is an input / output control unit 22, a user intention analysis unit 23, a user intention confirmation unit 24, and a behavior execution unit 25 which will be described later.
Etc.

The display object 17 is knowledge about the display object, and includes an input / output control unit 22 and a user intention analysis unit 2 which will be described later.
3, referred to by the user intention confirmation unit 24, the behavior execution unit 25, and the like.

On the other hand, the processing system 20 includes the input / output device 2
1, an input / output control unit 22, a user intention analysis unit 23, a user intention confirmation unit 24, and a behavior execution unit 25.

The input / output device 21 receives an instruction from the user and transmits it to the input / output control unit 22, and the input / output control unit 22
The display screen is output to the user or character data, voice data, or the like is output under the control of.

The input / output control unit 22 controls the input / output device 21 with reference to the knowledge of the display target 17, the display target state 16, the display data 11, and the like.

The user intention analysis unit 23 uses the ontology 1
2. The user's intention is analyzed by referring to the knowledge of the user intention analysis grammar 13, the display target 17, the display target state 16, and the like.

The user intention confirmation unit 24 uses the ontology 1
2, referring to the knowledge of the confirmation sentence generation grammar 15, the display target 17, and the display target state 16 to generate a sentence for asking the user what the device interprets as the user's intention, and input / output this It serves to pass it to the control unit 22.

The behavior execution unit 25 refers to the behavior to be executed based on the result of interpreting the user's instruction from the behavior dictionary 14, and executes the behavior by referring to the knowledge about the target and the knowledge about the state of the target.

As described above, the apparatus shown in FIG. 2 is an application of the information processing apparatus shown in FIG.
A device related to a device for supporting specification of a control program by planning as dealt with in No. 328997, which is a control program design supporting device for obtaining a target state from a visual operation during planning. is there.

Therefore, the structure of the control program design support device 30 shown in FIG. 2 will be described here, and then the operation of the device 30 shown in FIG. 2 will be described as a representative.

The control program design support device 30 shown in FIG. 2 is a control program for controlling each device of plant equipment having a line for winding a steel plate to produce a coil and a line for transporting the produced coil. It is intended for the design support of the control specifications necessary for the creation of the, and is roughly divided into a knowledge base 40 and a processing system 50.

The knowledge base 40 includes at least equipment / processing material display data 41, ontology 42, user intention analysis grammar 43, target state generation procedure 44, confirmation sentence generation grammar 45, equipment / processing material / sensor state 46, equipment.・ Knowledge of processing material / sensor 47. Further, the knowledge of the knowledge base 40 is referred to by the input / output control unit 52, the user intention analysis 53, the user intention confirmation unit 54, the target state generation unit 55, and the like of the processing system 50, which will be described later.

The equipment / working material display data 41 is display data knowledge such as graphic data of equipment / working material handled by the apparatus of this embodiment. At least part of this data is expressed using the vocabulary of the ontology related to the equipment / processed material and the state of the equipment / processed material. The display data can be switched.

The ontology 42 is an ontologyized knowledge about the equipment, the processing material, the sensor, and the state thereof. This ontology 42 is device / processing material / sensor display data 41, user intention analysis grammar 43, confirmation sentence generation grammar 45, target state generation procedure 44, device / processing material / sensor 47, device / processing material / sensor state. 46 and the like, and is referred to by the user intention analysis unit 53 described later.

The user intention interpretation grammar 43 is grammatical knowledge for analyzing the user's intention from an instruction given by the user via the input / output device 51 described later. When the grammatical rule applied to the interpretation is determined, the grammar 45 for generating the confirmation sentence and the target state generating procedure 44 can be searched using the rule.

The target state generation procedure 44 is knowledge as a procedure for generating the target state based on the result of interpreting the user's instruction. In this portion, procedural knowledge as shown in FIGS. 30 and 32 is stored.

The confirmation sentence generating grammar 45 is grammatical knowledge used for generating a sentence for confirming to the user the result of interpreting the user's instruction.

The equipment / working material / sensor state 46 is knowledge about the state of the equipment / working material / sensor. This knowledge is stored in the input / output control unit 52 and the user intention analysis unit 5 described later.
3, the user intention confirmation unit 54, the target state generation unit 55, and the like.

The equipment / working material / sensor 47 is knowledge about the equipment / working material / sensor. This knowledge is referred to by the input / output control unit 52, the user intention analysis unit 53, the user intention confirmation unit 54, the target state generation unit 55, etc., which will be described later.

On the other hand, the processing section 50 uses the input / output device 5
1, an input / output control unit 52, a user intention analysis unit 53, a user intention confirmation unit 54, a target state generation unit 55, an initial state setting unit 56, and a control specification generation unit 57.

The input / output device 51 receives an instruction from the user and informs the input / output control unit 52 of the input / output control unit 52.
11 is a device for outputting the screens shown in FIGS. 11 to 14 or the like, or outputting character data, voice data, or the like to the user under the control of FIG.

The input / output control unit 52 controls the input / output device 51 with reference to the knowledge of the equipment / working material / sensor 47, the equipment / working material / sensor state 46, the display data 41 and the like.

The user intention analysis unit 53 uses the ontology 4
2. The user's intention is analyzed with reference to the user intention analysis grammar 43, the knowledge about the target, the knowledge about the state of the target, and the like.

The user intention confirmation unit 54 uses the ontology 4
2, referring to the confirmation sentence generation grammar 45, knowledge about the equipment / working material / sensor, and knowledge about the state of the equipment / working material, a sentence for confirming with the user what the device interprets as the user's intention It is generated and passed to the input / output control unit 52.

The target state generation unit 55 refers to the target state generation procedure to be referred to based on the result of interpreting the user's instruction, and further has knowledge about the equipment / machining material / sensor and the status of the equipment / machining material / sensor. Execute the goal state generation procedure with reference to the knowledge about.

The initial state setting unit 56 refers to the target state generation procedure to be referred to based on the result of interpreting the user's instruction, and further has knowledge about the equipment / working material / sensor and the state of the equipment / working material / sensor. Set the initial state by referring to the knowledge about.

The control specification generation unit 57 is the target state generation unit 5
The control specification is generated based on the target state as shown in FIGS.

Then, the generated control specification is given to the control program generating device 60. The control program generation device 60 generates a control program for a control target 61 such as a device or a processing material based on the received control specifications.

The control program design support device 3
The target state output by 0 can also be used to set the initial state of the controlled object 61.

3 and 4 show an example of the conceptual hierarchy of the vocabulary of the ontology used in the apparatus of this embodiment.

In the figure, the coil car, the thin plate, the detection of the stop position 1 and the like, which are subordinate to the equipment, the processing material and the sensor, are part of the vocabulary of the ontology related to the equipment, the processing material and the sensor, and are under the equipment state. Loading and supporting is part of the ontology vocabulary regarding the state of equipment, processed materials, and sensors.

The vocabulary of the ontology and its structure are constructed, for example, based on a consensus document (steel handbook, etc.) in the control field, and the grammar shown in FIGS. 6 and 7 described later is created based on this ontology. As a result, the operation rule can be widely standardized and can be generally used in the control field regardless of the designer or the user.

FIG. 5 shows an example of the definition of the coil car, which is a word included in the ontology vocabulary.

In the figure, 71 is an attribute item relating to a name, a physical configuration, a state figure, a processing material handled by the coil car, and the like used when the coil car describes other knowledge, and 72 is an attribute value. Into the attribute value, words in the vocabulary of the ontology such as a thin plate and the state of the ascending / descending state, constants, and some data are registered.

In the implementation, the class attribute or the type of the attribute value of the instance (object) may be used.

FIGS. 6 and 7 show examples of knowledge of the user intention analysis grammar 43 used in the apparatus of this embodiment.

In this grammar, when the device, processing material, or sensor selected by the user is included in the concept represented by the ontology word on the left side, the behavior on the right side is interpreted as intended by the user. The knowledge of the confirmation sentence generation grammar 45 and the target state generation procedure (behavior dictionary) 44 can be searched using the term on the right side.

8 to 10, a confirmation sentence generating grammar 45 is shown.
An example of knowledge of is shown.

In the apparatus of this embodiment, case grammar is used for description. In the figure, 101, 111 and 121 represent cases of each item of these grammatical rules. 102, 112 and 122
In, the concept applicable to each term is expressed by the words of the ontology enclosed in parentheses, and the ancillary words added to each term when the sentence is generated using this grammar are described after the words of the ontology. 103, 113 and 123
Is an example of a sentence generated by applying these grammars.

11 to 13 show examples of screens displaying cases (states) in which the state of the equipment, the processing material, and the sensor of the plant equipment targeted by this apparatus are different. This equipment is called a delivery equipment in this embodiment.

The outline of the operation of this equipment is as follows. First, the tension reel winds a thin iron plate into a coil. The thin plates wound in a coil shape are the coil 1, the coil 2, and the coil 3 in the figure. Coils 1-3
No. The coils are sequentially conveyed by the 1-output side coil car and placed on the saddle. In these figures, for example saddle 1
Is displayed. After being mounted on the saddle 1, the coils 1 to 3 are carried to another facility together with the saddle 1 by another carrier device. Until the coiled thin plate placed on the saddle 1 is carried away, a new coil cannot be placed on another saddle, so the coil is once placed on the skid 1 and the skid 2 to measure the timing. In the figure, in order to display the position of the device in an easy-to-understand manner, a graphic showing the place is displayed as necessary. Here, this figure is called a sensor.

FIG. 14 shows an example of a screen displaying a scene of another place in the plant equipment. This equipment is called a transportation equipment in this embodiment.

The outline of the operation of this equipment is as follows. First, a coiled thin plate (coil 4 or coil 5) is carried on a saddle. In this figure, the saddle 5 is displayed. The coil 5 placed on the saddle 5 is No. It is received by the 6 spool tongs, conveyed, and placed on the skid 5 or the skid 6. The coils on the skid are received by other equipment and transported to other equipment. In the figure, in order to display the position of the device in an easy-to-understand manner, a graphic showing the place is displayed as necessary. Here, this figure is called a sensor.

Even when the screen of the equipment using different types of equipment is displayed and operated as shown in FIGS. 11 and 14, FIG.
Further, the grammars of FIGS. 6 to 10 described based on the vocabulary of the ontology shown in FIG. 4 are effectively used, and it is not necessary to newly create an operation rule by a combination of figures.

FIG. 15 to FIG. 18 show examples of knowledge about equipment and processing materials.

In the figure, 131, 141, 151, 161
Is an attribute item related to an object name, a word in the vocabulary of the ontology to which the object corresponds, a physical configuration, a state when a device / processing material is displayed, a state item of the device, a primitive state that the state item can take, a figure, etc. is there. In the figure, 132, 142, 152 and 162 are attribute values. The graphic data of the attribute value is connected to the display data.

19 to 22 show examples of knowledge about the state of the equipment / processed material. These figures show primitive states corresponding to the signals representing the states of the sensors on the programmable controller, the logical combination of the signals, and the like.

In the figure, 171, 181, 191, 201
Is the object name of each state, the words in the ontology vocabulary that the object corresponds to, the device that has this state,
Attribute items related to a state item to which this state belongs, and 172, 182, 192, and 202 represent attribute values.

23 and 24 show examples of knowledge representing the state of the entire device by combining the primitive states.

In the figure, 211 and 221 represent object names, words in the vocabulary of the ontology to which the objects correspond, and attribute items related to combinations of primitive states, and 212 and 222 represent their attribute values.

FIG. 25 shows an example of knowledge regarding the processing material.

In the figure, reference numeral 231 is an attribute item relating to a word in the vocabulary of the ontology to which the object corresponds, the state of the processing material or the processing material of the processing material, the state of the processing material, and the like, and 232 is the attribute value thereof.

In the present embodiment, the object name is treated like an object identifier, but the object identifier or another method may be used as long as knowledge can be connected.

26 to 28 show examples of target states of the control program design support apparatus of this embodiment. In this embodiment, the target state is treated as a condition that the target state should satisfy, rather than listing the states of the devices as shown in FIG. That is, it is a condition that "the primitive state registered in the state item b of the device a is the primitive state c" using the combination of (device a, state item b, primitive state c). In this way, generation of the target state is simplified by only specifying a part of the original target state. Therefore, the target state is also referred to as a state search condition.

A means for generating and setting a target state or an initial state of a controlled object from a simple graphic operation will be described with reference to FIGS. 29 to 33. Further, it will be explained here that different equipments as shown in FIG. 11 and FIG. 14 can be commonly processed by using the same grammar and procedure.

FIG. 29 shows a processing procedure for receiving a visual instruction from a user and performing a behavior intended by the user, using an analytic grammar described by using a vocabulary of an ontology related to equipments and processed materials in equipment of a plant. It is shown.

At step 301, the user selects a device graphic from the screen (FIGS. 11 and 14) displaying the graphic of the device or the processing material. For example, from FIG.
No. 1 from the output side coil car and coil 1 on the skid 2, from FIG. 6 Select the coil 5 on the spool tongs and saddle 5. You can select the figures one by one using a pointing device such as a mouse, or drag the figure of the equipment or the figure of the processing material.
You can do it by dropping and dropping.

In step 302, step 301
Search for words in the corresponding ontology vocabulary for the equipment and processed material selected in. In this embodiment, since the words in the ontology vocabulary are defined by using the object-oriented programming class, the attribute item instance-of of the object representing the device or the processing material is defined.
Search for words in the ontology vocabulary from. For example, No. No. 1 coil coil from the output side coil car, thin plates from coil 1 and coil 5, No. The spool tongs are searched from the 6 spool tongs.

In step 303, step 302
A grammar rule satisfying a combination of words in the ontology vocabulary retrieved in step 1 is retrieved from the user intent analysis grammar in FIG. For example, a coil car is a subordinate concept of a trolley-type carrier device, a carrier device, and a device because of the ontology vocabulary hierarchy of FIG. 82, 83)
Will be searched. Since the thin plate is a subordinate concept of the steel material and the processed material, the grammatical rule (81, 82) in which the thin plate, the steel material and the processed material are the elements on the left side is searched. Similarly,
The spool tongs are searched for the grammar rule (82, 83).

In step 304, step 303
The grammatical rule to be applied is determined from among the grammatical rules searched in. For example, in the process starting from the instruction of FIG. 11, the intersection of the grammar rules retrieved by the coil car and the thin plate is taken,
Grammar rule 82 is selected. In the process starting from the instruction in FIG. 14, the product set of the grammar rules retrieved by the spool tongs and the thin plate is taken, and the grammar rule 82 is selected. Here, when there are a plurality of elements of the intersection of grammar rules, a grammar rule having a higher specificity is selected, that is, a grammar rule described in a lower concept is selected in the hierarchy of the ontology vocabulary shown in FIGS. 3 and 4. To do.

In step 305, step 304
The behavior to be executed is selected from the behavior dictionary according to the behavior name on the right side of the grammar rule selected in. In the processing example starting from the instruction in FIGS. 11 and 14, the processing (FIG. 30) corresponding to the right side of the grammar rule 82: transfer 2 (receive) is searched.

In step 306, step 305
Execute the process found in.

FIG. 30 shows a processing procedure corresponding to transfer 2 (receive).

In step 401, a device registered in the attribute "loaded-by" of the processing material object is searched. For example, in the process starting from the instruction of FIG. 11, the skid 2 is searched for from the object of FIG.

In step 402, a state belonging to the subordinate concept of "state without processing material" is retrieved from the retrieved state items ("mounted state" or "support state") of the equipment. For example, with respect to the skid 2, an instance of “no coil support”, which is an instance of the subordinate concept “no support” of the state without a processing material, is searched from the attribute item “support state” of the object of the skid 2 in FIG.

In step 403, a state belonging to the subordinate concept of "state with processing material" is searched from the state item "installed state" of the carrier device. For example, in the process starting from the instruction of FIG. Figure 1 for the 1-side coil car
No. 5 1 From the attribute item “installed state” of the object of the output side coil car, “in-coil installed”, which is an instance of the subordinate concept “installed” of the state with processing material, is searched.

In the processing starting from the instruction of FIG.
No. 6 spool tongs shown in FIG. From the attribute item "installed state" of the 6-spool tong object, "coil installed" which is an instance of the subordinate concept "installed" of the state with processing material is searched.

In step 404, a list of state search conditions is created and output. The output format is, for example,
A list format such as ((conveying device “mounted state” with processing material) (searched device “mounted state” or “support state” without processing material)) may be used, or a table format may be used. When the state search conditions in the process starting from the instructions of FIGS. 11 and 14 are expressed in a table format, they are shown in FIGS. 26 and 2, respectively.
It becomes like 7.

At step 405, a sentence is created by applying the object name of the instance object to the confirmation sentence generation grammar “(transport device) receives (processing material) of (searched device).” In FIG. To do. For example, in the process starting from the instruction of FIG.
The first output coil car, the skid 2 on the searched device, and the coil 1 on the processing material are applied, and the sentence "No. 1 output coil car receives the coil 1 on the skid 2." is created.

In the process starting from the instruction of FIG. 6 spool tongs, the saddle 5 is applied to the searched equipment, and the coil 5 is applied to the processing material, and the sentence "No. 6 spool tongs receives the coil 5 on the saddle 5" is created.

As a method of outputting, it may be output as a character string as it is, or voice data may be held for each word or object in the vocabulary of the ontology, and in addition to this, voice data of particles and directives may also be held. May output by voice. It is also possible to provide a means for storing and referring to the history of output sentences.

By repeating the processing of steps 401 to 405, the condition of (device status item primitive status) of the status search condition may be increased while confirming the operation.

FIG. 31 shows a visual instruction given by the user using the analysis grammar described using the vocabulary of the ontology related to the state of the equipment and the processing material in addition to the ontology related to the equipment and the processing material in the plant equipment. The processing procedure until the user receives the desired behavior is shown.

In step 501, the user selects the device graphic from the screen displaying the graphic of the device or the processing material. For example, from FIG. 1 Select the output side coil car and saddle 1. As for the selection method, the figures may be selected one by one using a pointing device such as a mouse, or may be performed by dragging and dropping the figure of the device or the figure of the processing material.

In step 502, step 501
Search for words in the corresponding ontology vocabulary for the equipment and processed material selected in. In this embodiment, since the words in the ontology vocabulary are defined by using the object-oriented programming class, the attribute item instance-of of the object representing the device or the processing material is defined.
Search for words in the ontology vocabulary from. For example, No. 1 From the output side coil car, coil car, saddle 1
Will search for saddles.

In step 503, step 502
A grammatical rule satisfying the combination of words in the ontology vocabulary retrieved in step 1 is searched from the user intention analysis grammar 43 shown in FIGS. 6 and 7.

For example, the coil car is a subordinate concept of the trolley type carrier device, the carrier device and the device due to the hierarchy of the ontology vocabulary shown in FIG. 3. Therefore, the coil car, the carrier type carrier device, the carrier device and the device are on the left side element. Grammar rules (82,
83, 91-97). Since the saddle is a subordinate concept of the processing material supporting device and the device, the grammatical rule (83,
91-97) are searched.

In step 504, step 503
With respect to the grammar restricted by the vocabulary of the ontology of states in the grammar searched in step 1, the grammar is narrowed down by comparing it with the state of the device and the processing material selected in step 501 at the selected time. .

In this embodiment, the grammar rules 91 to 97 are restricted by the vocabulary of the state ontology. Among these, for example, the grammar rule 91 is restricted by the words in the ontology vocabulary that the carrier device is installed. In the process starting from the instruction of FIG. 1 The output side coil car corresponds. Therefore,
No. 1 Search the object (FIG. 23) representing the state of the coil car from the attribute item current-state of the object (FIG. 15) of the output side coil car. Furthermore, the coil item in the coil car state is searched for from the attribute item mounting state of the object (FIG. 15). Since it is searched for from the attribute item instance-of that the coil is installed, No. 1 For output coil car, grammar rule 91
Remains as a candidate for applicable grammar rules. If the value of the attribute item mounting state of the object in the coil car state (FIG. 23) is no coil mounted, the grammar rule 91 is deleted from the candidates of the applicable grammar rule. Similarly,
As a result of narrowing down other grammars, No. Grammar rules (82, 83, 91,
94, 96) remain candidates. For saddle 1, grammar rules (83, 91, 93, 96) remain candidates.

In step 505, step 504
Determine the grammar rule to apply from among the grammar rules narrowed down in. For example, in the process starting from the instruction in FIG. 12, the product set of the grammar rules retrieved by the coil car and the saddle is taken, and the grammar rule (83, 91, 96) is selected. Of these, the grammatical rule 83 is the remaining two grammatical rules (91, 9) because the device is a superordinate concept of the processing material supporting device.
Less specific than 6). Further, in the grammar rule 96, the left side is the rotating device, but the device given by the user's instruction is the carrying device, and thus the grammar is less satisfied. Therefore, the grammar rule 91 is selected.

As described above, when there are a plurality of elements of the intersection of grammar rules, the grammar rule having high specificity and sufficiency is given priority.

In step 506, step 505
The behavior to be executed is selected from the behavior dictionary according to the behavior name on the right side of the grammar rule selected in. In the processing example starting from the instruction in FIG. 12, the processing (FIG. 32) corresponding to the right side of the grammar rule 91: transfer 3 (load) is searched.

In step 507, step 506
Execute the process found in.

FIG. 32 shows a processing procedure corresponding to transfer 3 (loading).

In step 601, the state item “supporting state” of the processing material supporting device is changed to “state with processing material”.
Search for states that belong to the subordinate concept of. For example, in the process starting from the instruction of FIG.
From the attribute item “supporting state” of the object of the saddle 1 of “1”, “coil supporting”, which is an instance of the subordinate concept “supporting” of the state with processing material, is searched.

In step 602, a state belonging to the subordinate concept of the "state of no processing material" is searched from the state item "installed state" of the carrier device. For example, in the process starting from the instruction of FIG. Figure 1 shows the coil car on the output side
No. 5 From the attribute item "mounting state" of the object of the first output side coil car, "no coil mounting", which is an instance of the subordinate concept "no mounting" of the state without processing material, is searched.

In step 603, a list of state search conditions is created and output. The output format may be a list format such as ((conveying device “mounted state” no processing material state) (processing material supporting device “support state” processing material present state)) or a tabular format. FIG. 28 shows the state search conditions in the process starting from the instruction of FIG. 12 in a table format.

In step 604, the processing material mounted on the carrying device is retrieved from the attribute item having-material of the object of the carrying device. For example, No. 1
Coil 1 is retrieved from the object on the outgoing side coil car.

At step 605, a sentence is created by applying the object name of the instance object to the confirmation sentence generation grammar of FIG. 9: "(transporting device) places (processing material) on (processing material supporting device)." Output. For example, in the process starting from the instruction of FIG.
o. 1 output side coil car, saddle 1 for processing material support equipment,
The coil 1 is applied to the processing material, and the sentence "No. 1 output side coil car puts the coil 1 on the saddle 1." is prepared.

As the output method, the character string may be output as it is, or voice data may be held for each word or object in the vocabulary of the ontology, and in addition to this, voice data of particles and directives may also be held. , May be output by voice. It is also possible to provide a means for storing and referring to the history of output sentences.

By repeating the processing of steps 601 to 605, the condition of (device status item primitive status) of the status search condition may be increased while confirming the operation.

FIG. 33 shows a procedure for setting the initial state by diverting the process corresponding to "transfer 3 (place)" in FIG.

Steps 701 to 703 are shown in FIG.
The process is almost the same as 2.

In step 701, the state item "supporting state" of the processing material supporting device is changed to "processing material present state".
Search for states that belong to the subordinate concept of. For example, in the process starting from the instruction of FIG.
From the attribute item “supporting state” of the object of the saddle 1 of “1”, “coil supporting”, which is an instance of the subordinate concept “supporting” of the state with processing material, is searched.

In step 702, a state belonging to the subordinate concept of the "state of no processing material" is searched from the state item "installed state" of the carrier device. For example, in the process starting from the instruction of FIG. Regarding the output side coil car, No. 1 in FIG. From the attribute item "mounting state" of the object of the first output side coil car, "no coil mounting", which is an instance of the subordinate concept "no mounting" of the state without processing material, is searched.

In step 703, a list of state search conditions is created. The format to be created may be a list format such as ((conveying device “mounted state” no processing material state) (processing material supporting device “support state” processing material present state)) or a table format. FIG. 28 shows the state search conditions in the process starting from the instruction of FIG. 12 in a table format.

In step 704, the machine state in which the state item "support state" of the machine material supporting machine is replaced with the machine material present state is searched for, and the machine state in which the "equipment state" of the carrier machine is changed to the state without machine material is conveyed. Search the device status. For example, in the case of the process starting from the instruction of FIG. 12, for example, in the case of the coil car, the coil car state α of FIG. 24 is searched by the instruction from the coil car state α of the initial state of FIG.

In step 705, a message for redisplaying the display screen according to the retrieved device status is sent to the input / output control unit 52. Thus, for example, in the process starting from the instruction of FIG. 12, from the first screen (FIG. 12),
The screen of FIG. 13 is displayed by the instruction.

By repeating the processing of steps 701 to 705, while confirming the operation, some conditions may be set for the initial state in the same way as the condition of the (device state item primitive state) of the state search condition is increased. You may perform the operation.

Furthermore, after this processing, a confirmation sentence may be generated in the same manner as in the analysis of the target state.

As described above, since the ontology relating to the display object and the grammar using the vocabulary of the ontology are provided, as long as the object which is in the ontology category is the display object, the GUI designer creates a GUI every time the GUI is created. A desired GUI can be created by simply specifying which word in the vocabulary of the ontology corresponds to the display target without creating a rule that connects the behavior with the GUI.
It is possible to improve the creation efficiency of.

Even if the designer of the user interface or the user is different, the meaning of the similar operation is similarly interpreted and executed, and the connection between the operation and the behavior is easily accepted by any user. It is possible to concentrate on the purpose, and by clearly telling the user how the user's operations are interpreted by the system, the user can work accurately and with peace of mind. Can be done.

Further, even when applied to the field for supporting the generation of control specifications, if the grammar utilizing the ontology and the vocabulary of the ontology regarding the controlled object to be displayed is provided,
As long as the control target in the ontology category is the display target, every time you create a screen that schematically displays the control target,
A desired GUI can be created simply by specifying which word in the vocabulary of the ontology the control object represented by the figure corresponds to, without creating a rule connecting the figure operation and the state change of the control object intended by the user. Therefore, the GUI creation efficiency can be improved.

Even if the user interface designer or user is different, the meaning of the similar operation is similarly interpreted and executed, so that the connection between the graphic operation and the state change of the controlled object is accepted by any user. This makes it easier for the user to concentrate on the control flow of the original control target. Moreover, by clearly informing the user of how the user's operation is interpreted on the system side, the user can safely and accurately perform the work, and the design work of the control program can be performed more efficiently. .

FIG. 34 is a block diagram of a control program design support device according to the third embodiment of the present invention.

Further, FIG. 35 shows a block diagram of a control program design support apparatus according to the fourth embodiment of the present invention.

The difference between the device shown in FIG. 34 and the device shown in FIG. 35 lies in the control specification creating procedure. That is,
As will be described later, the apparatus shown in FIG. 34 enhances the search efficiency by using the work knowledge having the matching condition to create one scope that limits the search range for one search of the state space. Further, the device shown in FIG. 35 creates several scopes for one search of the state space by using work knowledge having the same matching condition, and performs the search while switching the scopes according to the phase of the search. By doing so, the search efficiency is improved.

First, the structure of the control program design support apparatus in the third embodiment shown in FIG. 34 will be described.

The control program design support device is roughly divided into a knowledge base 800 and a processing unit 900.

The knowledge base 800 stores the target knowledge 801, the state space model 802, the control specification 803, etc., and provides the knowledge used in the processing performed by the processing unit 900.

The target knowledge 801 includes knowledge about the control target such as the graphic knowledge 804, the state configuration knowledge 805, the work knowledge 806, the motion knowledge 807, and the secondary state transition 808. These pieces of knowledge are used for each processing in the processing unit 900.

The graphic knowledge 804 is knowledge of how to display a facility or a device in the facility by using a graphic or the like according to the state thereof.

The state configuration knowledge 805 is knowledge relating to the states of equipment and devices. In this embodiment, this knowledge is divided into equipment level, equipment level, and primitive level.

The work knowledge 806 is knowledge about work in the facility. In this embodiment, the matching condition of the work knowledge 806 is one key point. Matching conditions are expressed using logical combinations of primitive level states.

The operation knowledge 807 is knowledge about the operation of the equipment in the facility. Each operation has its own weight, and this weight is (can be treated like a weighted directed graph).
Used as a weight for the state space model.

The subsidiary state transition 808 is knowledge of the subsidiary state transition that occurs in the device operated by the operation of the device and other devices.

The state space model 802 is a model used in the state transition path search.

The control specification 803 is an output of this control program design support device.

On the other hand, the processing section 900 uses the input / output device 9
01, input / output control unit 902, initial state setting unit 903, initial state information storage unit 904, target state (state search condition) definition unit 905, target state (state search condition) information storage unit 90
6, scope determination unit 907, scope information storage unit 90
8, search calculation amount estimation unit 909, state transition route search unit 9
10, state transition route information storage unit 911, state transition route-
The control specification conversion unit 912 is provided.

The input / output device 901 is provided with a display screen, a keyboard and the like, and performs input / output processing between the user and this control program design support device.

The input / output control unit 902 uses the state configuration knowledge 80
5 and the graphic knowledge 804 and the like, control is performed to display the equipment in the facility by the input / output device 901 and interaction with the user is controlled.

The initial state setting unit 903 uses the graphic knowledge 804.
The initial state of the control target (equipment) is set based on the user's instruction using the state configuration knowledge 805 and the like.

The initial state information storage unit 904 stores initial state information indicating the state of the control target (equipment) at the beginning when creating the control specifications.

Target state (state search condition) definition unit 905
Defines a target state (state search condition) of a control target (equipment) from a user's instruction using the graphic knowledge 804 and the state configuration knowledge 805.

Target state (state search condition) information storage unit 90
Reference numeral 6 stores target state (state search condition) information representing a state that a controlled object (equipment) reaches by device operation.

The scope determining unit 907 determines the state configuration knowledge 8
The scope is determined from 05, work knowledge 806, and the like.

The scope information storage unit 908 stores information that restricts the search space in the search of the state space, that is, represents the device of interest when searching the state transition path.

The search calculation amount estimation unit 909 uses the work knowledge 8
06, motion knowledge 807, scope information, state space model 802, etc. are used to estimate the amount of calculation at the time of search.

The state transition path searching unit 910 generates a state transition path from the initial state to the target state by using the state configuration knowledge 805, the operation knowledge 807, the subsidiary state transition 808, the state space model 802 and the like.

The state transition path information storage unit 911 stores the state transition path information indicating the operation of the device from the initial state to the target state and the state transition accompanying the operation.

The state transition path-control specification conversion unit 912 is
A state transition path is converted into a control specification by using the state configuration knowledge 805, the operation knowledge 807, and the like.

In the figure, reference numeral 913 indicates a control program generation device for generating a control program from given control specifications, and reference numeral 914 indicates a control program executed on a factory computer for controlling a controlled object (equipment). is there.

Next, the structure of the control program design support apparatus in the fourth embodiment shown in FIG. 35 will be described.

This control program design support device is also roughly divided into a knowledge base 1000 and a processing unit 1100.

The knowledge base 1000 is the target knowledge 100.
1, the state space model 1002, the control specification 1003, etc. are stored, and the knowledge used in the processing performed by the processing unit 1100 is provided.

The target knowledge 1001 is the figure knowledge 1004,
State configuration knowledge 1005, work knowledge 1006, motion knowledge 1
007, knowledge of controlled objects such as secondary state transition 1008 is provided. These pieces of knowledge are used for each processing in the processing unit 1100.

The graphic knowledge 1004 is knowledge of how to display a facility or a device in the facility by using a graphic or the like according to its state.

The state configuration knowledge 1005 is knowledge relating to the states of equipment and devices. In this embodiment, this knowledge is divided into equipment level, equipment level, and primitive level.

The work knowledge 1006 is knowledge about work in the facility. In this embodiment, the matching condition of the work knowledge 1006 is one key point. Matching conditions are expressed using logical combinations of primitive level states.

The operation knowledge 1007 is knowledge about the operation of the equipment in the facility. Each action has its own weight, and this weight is used as the weight of the state space model (which can be treated like a weighted directed graph).

The subsidiary state transition 1008 is knowledge of the subsidiary state transition that occurs in the device operated by the operation of the device and other devices.

The state space model 1002 is a model used in the state transition path search.

The control specification 1003 is an output of this control program design support device.

On the other hand, the processing section 1100 has the following components.

The input / output device 1101 has a display screen, a keyboard and the like, and performs input / output processing between the user and this control program design support apparatus.

The input / output control unit 1102 uses the state configuration knowledge 1
005 and the graphic knowledge 1004 are used to control the display of the equipment in the facility on the display of the input / output device 1101 and control the interaction with the user.

The initial state setting unit 1103 uses the graphic knowledge 10
04 and the state configuration knowledge 1005, etc., the initial state of the controlled object (equipment) is set from the user's instruction, and the initial-operation is set from the set initial state.

The initial state information storage unit 1104 stores information indicating the state of the control target (equipment) at the beginning when creating the control specifications.

Initial-operation information storage unit 1105
Is the work knowledge that the initial state satisfies the matching condition, and stores the initial-operation information used to create the first scope in the state route search.

Target state (state search condition) definition unit 1106
Defines the target state (state search condition) of the controlled object (equipment) from the user's instructions using the figure knowledge 1004 and the state configuration knowledge 1005, and g
Define oal-operation. Target state (state search condition)
The information storage unit 1108 stores target state (state search condition) information indicating a state that a control target (equipment) reaches by the operation of a device.

The goal-operation information storage unit 1109
The work knowledge that the target state (state search condition) satisfies the matching condition is stored as information that also serves as a criterion for success / failure of the search.

The work sequence creating section 1110 uses the work knowledge 100
Create a sequence of work knowledge (work sequence) between 6 and initial-operation and goal-operation.

The work sequence information storage unit 1111 stores the initial-op
Store work sequence information that represents the sequence of work knowledge between the eration and goal-operation.

The search calculation amount estimation unit 1112 estimates the calculation amount at the time of search using the work sequence and the like. In the block diagram of FIG. 35 and the procedure of FIGS. 67 and 68, only the work sequence is used, but as in the block diagram of FIG. 34 and the procedure of FIG. 66, work knowledge 1006, motion knowledge 1007, scope, The state space model 1002 or the like may be used.

The state transition path search unit 1113 has the state configuration knowledge 1005, the operation knowledge 1007, and the subsidiary state transition 10
08, the state space model 1002, the work sequence, and the like are used to generate a state transition path from the initial state to the target state. The state transition route search unit 1113 mainly determines the scope and
The switching unit 1114, the state transition destination determination unit 1115, the target state determination unit 1116, the search failure determination unit 1117 and the like are included.

The scope determination / switching unit 1114 determines the search phase from the matching condition of the state configuration knowledge 1005 and the work knowledge 1006, the work sequence, etc., switches the work knowledge used for the scope determination, and switches the work knowledge to the scope. To decide.

The scope information storage unit 1118 limits the search space in the search of the state space, that is, stores the scope information indicating the device of interest when searching the state transition path.

The state transition destination determining unit 1115 has the state configuration knowledge 1005, the operation knowledge 1006, and the subsidiary state transition 100.
8. The state space model 1002 and the like are used to successively determine the transition destinations of the equipment states.

The state transition path information storage unit 1119 stores information indicating the operation of the device from the initial state to the target state and the state transition accompanying the operation.

The target state determination unit 1116 determines whether the equipment state searched by using the state configuration knowledge 1005, the state search conditions, etc. is the target state.

The search failure determination unit 1117 determines whether or not the state search route deviates from the work sequence by using the state configuration knowledge 1005, goal-operation and the like.

State transition path-control specification conversion unit 1120
Converts the state transition path into a control specification using the state configuration knowledge 1005, the operation knowledge 1006, and the like.

In the figure, reference numeral 1121 indicates a control program generating device for generating a control program from the generated control specifications, and 1122 indicates a control program executed on the factory computer for controlling the controlled object (equipment). ing.

[0209] Fig. 36 shows an example of a screen displaying the front of the equipment and processing materials in the plant equipment targeted by the device of this embodiment, and Fig. 37 shows an example of the screen displaying its side. ing. Further, FIG. 38 shows an example of a screen displaying the front of different scenes (states) in the plant equipment, and FIG. 39 shows an example of the screen displaying the side surface thereof. In this embodiment, this equipment is called the output equipment.

The outline of the operation of this equipment is as follows.

First, in the screens from the front of FIGS. 36 and 38, the iron plate is passed through a device called a shear for cutting an iron plate, a device called a deflector pinch roll for bending an iron plate downward, and a device called a table for guiding an iron plate. Is sent to the tension reel. The tension reel winds the iron plate into a coil while being assisted by the upper / lower tail end press rolls, belt wrappers, and the like. The illustrated outboard, mandrel, etc. are part of the tension reel. The mandrel is a shaft of varying thickness for winding the iron plate, and the outboard is a part that supports the mandrel from the outside (the side opposite to the main body of the tension reel).

Further, in the screens from the side surfaces of FIGS. 37 and 39, the tension reel winds the iron plate into a coil shape, which is the coil in the figure. The coil is No. It is pulled out from the tension reel by the 1-output side coil car, sequentially conveyed, and placed on the skids 1 to 3.

After the coil is placed on the skid 3, it is carried to another facility by another carrying device. Until the coil placed on the skid 3 is carried away, another coil cannot be placed anew, so the coil is once placed on the skid 1 and the skid 2 to measure the timing.

In the figure, in order to display the position of the device in an easy-to-understand manner, a graphic showing the place is displayed as necessary. Here, this figure is called a sensor.

FIG. 40 shows an example of knowledge about a coil car which is a device in the facility. In the figure, 1201
Is an attribute item related to a knowledge name, a physical configuration, a state item of a device, a primitive state that the state item can have, a figure, and the like. 1202 is the attribute value.

FIG. 41 shows an example of the structure showing the state inside the equipment. In the embodiment shown in FIG. 34 and FIG. 35, state configuration knowledge 805 (1
005).

The state configuration knowledge 805 (1005) is an operation that describes the operation of the device, a side effect (secondary state transition) by the operation of the machine itself or another machine, and the operation or state constraint of the device. It is used as a vocabulary base for expressing knowledge.

In the embodiment shown in FIGS. 34 and 35, the state is divided into three levels. The lowest level state is the primitive state of a device measured by one sensor or set of sensors, or PC (Prog
This is a primitive state of a device that directly corresponds to the signal on the ramable controller, and is mainly called the primitive state. In the primitive state, not only the device but also the primitive state such as the sensor and the processing material may be handled.

The device state is a direct product of state items made up of a set of device primitive states. Furthermore, the equipment state is a direct product of a set of equipment states.

The primitive state has an adjacency relation (pre-state, next-stat) with another primitive state. For example, the next-state of the coil carrying height has a rising limit, and the pre-state of the coil carrying height has a falling limit. Also, when the expression “state that is next to the lower limit” is used, it refers to the primitive state such as the coil carrying height and the upper limit. On the contrary, when the expression "state that is in pre rather than ascending limit" is used, it refers to the primitive state such as the coil carrying height and the descending limit.

FIG. 42 shows an example of the structure of the work knowledge 806 (1006).

In the embodiments shown in FIGS. 34 and 35, the knowledge expressing which equipment participates in the work or function in the facility is called work knowledge. Work knowledge 806 (1
The structure of (006) has a tree structure based on the upper and lower relations. The devices that commonly participate in the lower work are registered in the upper work of the tree structure. For example, a coil car that participates in both the coil extracting work and the coil carrying work is registered in the coil payout work. Further, in the knowledge of the coil car, there are registered works for the coil car to participate such as a coil payout work, a tail end opening work, and a tail end pressing work.

In the embodiment shown in FIG. 34 and FIG. 35, a composite device constituted by combining several devices such as a tension reel is also registered in the work knowledge as one device. However, it is often the case that one part of the composite device does not participate in the work of another part. For example, in the work of both embodiments, in the mandrel and the outboard that are part of the tension reel, the mandrel participates in the coil extracting work, but the outboard does not participate in the work knowledge. In this case, the device registered in the scope information storage unit 908 (1118) can be a partial device of the composite device. In these cases,
In addition to restricting the state search by the device, it is also possible to limit the search to a partial operation (or function) of the composite device.

Further, in both embodiments, the work knowledge is used to determine the scope, but the scope is not limited to the work knowledge, and any device capable of grouping devices such as physical configurations can be used. For example, in the physical configuration, a physical configuration can be used for the upper / lower relation, and a physical or functional adjacency relation can be used for the order relation.

Further, the upper / lower relation is not always indispensable, and if the relation between the equipment and the work is clear,
It does not have to be limited to a tree structure. The order relation may not be essential depending on the processing.

FIG. 43 shows a description example of the work knowledge 806 (1006).

In the figure, reference numeral 1301 is an upper / lower relationship, order relationship, matching conditions, attribute items relating to devices participating in work, and 1302 is an attribute value.

The work knowledge 806 (1006) has a structure according to the upper-lower relation of FIG. 42 and also has an order relation.
The order relation is represented by the attribute items “previous work” and “immediate work”. When enumerating the work between a certain work and another work, basically this relationship is followed.

The matching condition is represented by a logical connection using symbols as shown in FIG.

In the attribute item "matching condition" of FIG. 43,
nil is registered, which means that it will match unconditionally when searched from the knowledge of the coil car. 45 to 49 show examples of matching conditions.

FIG. 44 shows an example of symbols that can be used for the matching condition and their meanings. For example, the description of (more-next coil car lower limit) is true if the state of the coil car in the equipment state includes the coil transport height and the upper limit.

FIG. 50 shows the state space model 802 (100
2) An example of the configuration is shown.

The model of the entire plant facility can be treated as a weighted directed graph as a whole. The structure of this model is composed of three states: a state space model of equipment, a scope branch, and a plant state group.

That is, the state space model 802 (100
2) has two levels: an equipment level state space and a plant equipment level state space including the equipment level state space. The link between the equipment status and the status of the corresponding plant equipment is called a scope branch. In addition, in the transition of the equipment state, the state space model of the device is temporarily passed. The state transition in the state space model of the equipment dynamically changes depending on the operation mode and the plant equipment state of the transition source.

The state of the plant equipment is represented as shown in FIG. 51 and FIG. 52, which is incrementally generated when the state route is searched.

There are two factors that cause the device status to transition: operation and subsidiary state transition. The operation represents a direct state transition due to an operation performed by the device itself. The secondary state transition is
It represents a secondary state transition (so-called side effect) that occurs due to the operation of the device itself or the operation / state of another device.

The equipment in the plant facility includes an active equipment which has its own operation and a passive equipment which does not have its own operation and whose state transits by a side effect accompanying the operation of the active equipment. There is. Accompanying this, there are two types of state transitions for active devices, one is due to operation and the other is due to the action of secondary state transition, and the state transition for passive devices is due to secondary state transition. There is only one.

Device state space model 802 (1002)
Is a weighted directed graph composed of device state nodes, directed edges of actions, and directed edges of secondary state transitions. The "device state node" is a node representing the device state and is the state α in FIG.
55, and is expressed as shown in FIG. 55.
The "directional branch of motion" is a directional branch as indicated by a ₁ in FIG. 50, and is represented by the part relating to motion in FIG. The “directed branch of the secondary state transition” is a directed branch such as ε ₁ in FIG. 50, and is represented by the portion related to the secondary state transition in FIG. 62.

53 and 54 show an example of description of the target state (state search condition) stored in the target state (state search condition) information storage unit 906 (1108).

For example, in FIG. 53, the primitive state of the state item "coil support" of skid 2 is in coil support, and No. 1 Indicates that the primitive state of the state item "coil mounted" of the output side coil car is satisfied when the coil is not mounted.

The target state determined before performing the state route search represents a condition to be satisfied by the target state, as added to the state search condition in parentheses. Therefore, at the start of the search, the search is started toward a set of equipment states that satisfy the state search condition. The final target state is the one that is closest to the initial state, that is, the one that has the smallest total weight of the path of the state transition. This process
It can be performed using an algorithm based on a method for solving the shortest path problem of a weighted directed graph such as the Dijkstra algorithm.

However, it is appropriate to see that even if it is simply applied to an object with a large state space such as equipment of a plant, it is generally unusable.

In both embodiments, a scope is used to solve this problem. Regarding the scope, Japanese Patent Application No. 4-
As mentioned in Japanese Patent No. 328997, in both examples, the method of Japanese Patent Application No. 4-328997 is made more efficient.

FIG. 51 and FIG. 52 show an example of the description of the equipment state of the plant.

The equipment state is represented by a combination of equipment states of equipment in the equipment. In the figure, 1401 and 1501 are attribute items concerning the state of each device in the facility, and 1402 and 1
502 is the attribute value.

55 to 60 show an example of description of the state of the device.

55 and 56 show the state of the coil car, FIGS. 57 to 59 show the state of the tension reel,
FIG. 60 shows the state of the upper tail presser roll.

In the figure, 1601, 1701, 1801,1
Numerals 901, 2001 and 2101 are status items of each device and attribute items related to directed edges.
02, 1902, 2002 and 2102 are the attribute values.

FIG. 61 and FIG. 62 show an example of the description of the directional branch coming out of the device state.

FIG. 61 shows the directional branch coming out of the equipment state of the tension reel, and FIG. 62 shows the directional branch coming out of the equipment state of the coil car.

In the figure, 2201 and 2301 are attribute items related to the operation of the device and the device state of the state transition destination due to the secondary state transition, and 2202 and 2302 are the attribute values thereof.

63 and 64 show one notation example of the control specifications.

In the figure, reference numerals 2401 and 2402 denote device operating units. The device operating unit is described using the signal name sent to the actuator that starts the operation of the device. In the figure, 2501,
2502 is a transition condition part. The transition condition part is a PC (Programmable Cont
roler) is described using the signal name above.

FIG. 65 shows a procedure for generating control specifications from the initial state to the target state after the user inputs the initial state and the target state of the controlled object.

The difference between this procedure and the control specification creating procedure in Japanese Patent Application No. 4-328997 is that the matching condition is applied to the work knowledge and the device registered in the scope is restricted using this condition. .

As a method of designating the initial state and the target state,
It is also possible to display a graphic of a device or the like on the screen and perform it by visual operation as in Japanese Patent Application No. 4-328997, or to express an initial state or a target state (state search condition). It is also possible to prepare a description format, describe the initial state and the target state in advance, and interpret them at the time of setting the initial state and the target state. Alternatively, it may be directly entered from the command line using the description format.

In step 2601, the initial state of equipment is set. The equipment state as shown in FIG. 51 is designated as the equipment state when the control is started. Examples of displaying this equipment state on the screen are shown in FIGS. 36 and 37.

When creating continuous control specifications, the target state obtained by the control specification creating process performed immediately before can be used as it is as the initial state. In this case, the process of setting the initial state can be omitted.

At step 2602, the target state (state search condition) of the equipment state is set. As shown in FIG. 37, in the example of the visual operation using the device figure, the processing material on the coil car is selected with the pointing device as indicated by arrow 1, and the skid 2 is further selected with the pointing device as indicated by arrow 2. Do the operation. Then, the state search condition (condition that the target state should meet) as shown in FIG. 53 is obtained from the knowledge of the equipment and the processing material.

The state search condition may be expressed in the following list format.

((Skid 2 supported state, coil supported) (No. 1 coil car mounted state, coil not mounted)) In step 2603, the attribute of knowledge of the device (for example, coil car and skid 2) registered in the state search condition ""Work" retrieves knowledge about the work of the facility in which each device can participate. For example, coil car knowledge (Fig. 4
0) is searched for (tail end pressing work, tail end opening work, coil payout work: FIG. 43). In addition, when the retrieved work has a subordinate work, the subordinate works (coil extracting work, coil carrying work) are added to the retrieved work.

Work knowledge (coil carrying work) is retrieved from the skid 2. As a result of these, the retrieved work is (tail end pressing work, tail end opening work, coil paying out work, coil extracting work, coil carrying work).

At step 2604, the matching condition of each work knowledge is evaluated. From the retrieved work knowledge, work knowledge whose initial state of the equipment satisfies the logical combination of matching conditions is picked up. The equipment state of FIG. 51 is configured by equipment states such as the coil car state f of FIG. 56, the tension reel state g of FIG. 59, and the upper tail end pressing roll state s of FIG. The primitive state registered in the state item of these device states is the step 2
Matching conditions (FIG. 45, FIG. 47, FIG. 4) of work knowledge (tail end pressing work, tail end opening work, coil payout work, coil extracting work, coil carrying work) retrieved in 603.
It is checked whether the attribute “matching condition” of FIG. 3, FIG. 48, FIG. 49) is satisfied. In FIG. 45, it is satisfied only when the tension reel state includes the state immediately after cutting and the tail end pressing roll state includes the open state. The tension reel state of FIG. 59 is not satisfied because it does not include the state immediately after cutting. Similarly, when the attribute “matching condition” of FIGS. 47 and 43 and the matching condition of FIGS. 48 and 49 are examined, the matching condition of the coil payout work of the attribute “matching condition” of FIG. 43 and the coil transport work of FIG. 49 is found. It can be seen that the equipment state of FIG. 51 is satisfied.

In step 2605, in the tree structure of work knowledge, the coil paying out work, the coil transporting work, and the work knowledge of the work knowledge satisfying the matching conditions in step 2604.
Create a minimal subtree containing root. In this example, (delivery-side equipment work, coil payout work, coil transportation work) constitutes the minimum tree.

In step 2606, a device which is a subordinate concept of the active device is selected from the devices registered in the attribute "working device" of the work knowledge forming the minimum tree. In this example, the coil car is the only device that enters the scope.

In this example, a complicated example is used as compared with the example used in Japanese Patent Application No. 4-328997.
If the matching condition is not used like No. 328897, the minimum tree of work knowledge is (output side equipment work, winding end work, tail end pressing work, tail end opening work, coil paying out work, coil extracting work, Coil transportation work). Therefore, the devices registered in the scope are the upper tail end presser roll, the lower tail end presser roll, the tension reel, and the coil car. The scope can be narrowed down to, for example, four devices out of eleven devices. On the other hand, if the matching conditions used in this embodiment are used, it is possible to further narrow down to one coil car.

At step 2607, the time and effort required for creating the control specifications are estimated. The procedure of the processing performed here will be described later with reference to FIG. 66.

At step 2608, control specifications are generated. For example, the control specification (FIG. 63) can be generated by performing the same processing as that after determining the scope in Japanese Patent Application No. 4-328997. The control specification targeted in the present embodiment is composed of at least a device operation part that describes the operation of the device and a transition condition part that describes the condition for ending the operation of the device.

The contents described in the control specification of FIG. 63 will be described first. No. 1 output side coil car no. 1 Until the signal "deceleration position 2" of the sensor measuring the position of the output side coil car is turned on. After that, the No. This time, the device called the output side coil car is now in the No. 1 low speed backward movement. 1 Until the signal "stop position 2" of the sensor measuring the position of the output side coil car turns ON. Furthermore, No.
The device called the output coil car No. 1 is now in the No. 1 operation. 1 Repeat until the signal "Descent limit" of the sensor that measures the position of the output side coil car turns ON.

FIG. 66 shows a procedure for estimating the time and effort required for creating the control specifications.

In the present embodiment, as an index for the effort estimation in the generation of the control specification, the number of scope devices, the number of device states in the search range, the number of operations of the scope devices, and the initial state to the target state. The number of working knowledge up to is used. In addition to this, it may be used as an index as long as it affects the amount of calculation of the search.

As an estimation method, all of the indexes mentioned above may be used, or some of them may be used. Further, the method of evaluation may be performed by adding weights depending on the object or by using logical combination or an evaluation function.

In the example shown in FIG. 66, the number of devices in the scope,
Boundary values (number of devices Border, number of device states Border, number of device operations Border, respectively) for the number of equipment states in the search range and the number of operations of scope devices are set in advance,
Check if any of the values are exceeded before performing the search.

If the user's instruction requires control that exceeds any one of the values, a warning is issued.

This means that the following logical combination is evaluated,
If it is true, it represents a process of issuing a warning.

(Number of Scope Devices> Number of Devices Border) (Π (Number of Equipment States in Search Range)> Number of Device States Border) (Σ (Number of Operations of Scope Devices)> Number of Device Operations Border) Step 2701 In, check the number of devices registered in the scope and compare with the "number of devices Border". If the number of devices is less than or equal to the “number of devices Border”, step 27
02, otherwise the process proceeds to step 2704.

In step 2702, the product of the number of states of all the devices registered in the scope is calculated, and the "number of device states" is calculated.
Border ". If the number of devices is "number of device states Bor
der ”or less, the process proceeds to step 2703, otherwise proceeds to step 2704.

At step 2703, the sum of the operation numbers of all the devices registered in the scope is calculated, and the "device operation number" is calculated.
Border ". If the number of equipment operations is "the number of equipment operations
Border "or less, step 2608 of FIG.
Otherwise, the process proceeds to step 2704.

At step 2704, it is informed that the generation of the control specification may take a time that the user does not allow. At the same time, the user may be notified of which of the device number Border, the device state number Border, and the device operation number Border is exceeded.

In Step 2705, Step 27
Upon receiving the warning in 04, the user decides whether or not to execute the control specification generation work. When performing the generation work,
The process proceeds to step 2608 in FIG. If not, the process ends. When finished, start over from the initial or target state settings.

In the processing described here, one boundary value is set for one index, but several modes are provided for the boundary value to switch or handle the boundary value to be applied according to the user's intention. The boundary value to be applied may be switched depending on the equipment or the like.

FIG. 67 and FIG. 68 show a procedure for generating a control specification from the initial state to the target state from the input of the initial state and the target state of the controlled object by the user.

The main difference from the procedure shown in FIG. 65 is that the devices registered in the scope are switched during the search.

From among the work knowledge, the work knowledge whose initial state satisfies the matching condition is initial-operation, and the work knowledge whose state search condition satisfies the matching condition.
Goal-operation.

[0285] Furthermore, from initial-operation to goal-op
Create a sequence of work knowledge (work sequence) between erations.

First, a search is performed by creating a scope from the minimum tree including the work knowledge that became the initial-operation and the root of the work knowledge. If the equipment state generated during the search does not satisfy the matching condition, initial-ope
Search the next work knowledge from the "immediate work" or work sequence of the work knowledge of ration. Work knowledge searched (curren
Create a new scope from root of t-operation) and work knowledge. Similarly, the equipment status generated during the search is
When the matching condition of current-operation is not satisfied, current-operation is switched to the next work.

The procedure shown in FIG. 67 will be described below.

At step 2801, the initial state of the equipment state is set. The equipment state as shown in FIG. 52 is designated as the equipment state when the control is started. 38 and 39 show examples in which this equipment state is displayed on the screen.

In step 2802, the target state (state search condition) of the equipment state is set. 38 and 3
In the example of visual operation using the device figure shown in 9, the processing material being wound with the tension reel is selected with the pointing device as indicated by arrow 1, and the skid 2 is selected with the pointing device as indicated by arrow 2. Perform the operation. Then, from the knowledge of the equipment and the processing material, the state search condition as shown in FIG. 54 (condition that the target state should satisfy)
Get.

The state search condition may be expressed in the following list format.

((Skid 2 supported state, coil supported) (No. 1 tension reel winding state, no coil)) In step 2803, the initial state of the equipment (see FIG. 5).
2) searches for work that satisfies the matching conditions and ini
tial-operation and first current-operation. In the case of the equipment state of FIG. 52, the matching condition of the tail end stop work of FIG. 46 is satisfied, and this work is initial-o.
peration and the first current-operation.

When a plurality of work knowledge matching conditions are satisfied, the most upstream work may be set as initial-operation, or the user may be asked which work to make.

From the beginning, which work knowledge is given to the user i
One way is to inquire whether to use nitial-operation.

When continuously creating control specifications,
The target state obtained in the immediately preceding control specification creating process is often used as the initial state in the next control specification creating process. In this case, the work knowledge at the time of reaching the target state is given as initial-ope.
By setting ration, the initial-operation determination process can be omitted.

Even when the target state obtained in the immediately preceding control specification creating process is not used as it is in the initial state of the next control specification creating process, the equipment state far from the target state of the immediately preceding process is controlled as the initial state. Few specifications are created. Therefore, first, in addition to the work knowledge that the target state of the immediately preceding process satisfied the matching condition, only the work knowledge (knowledge of the adjacent work) registered in the attributes "previous work" and "immediate work" of this work By evaluating the matching conditions, the initial-operation determination process can be made more efficient.

At step 2804, the device knowledge registered in the target state (state search condition) and the work knowledge of which the state satisfies the matching condition are searched and set as goal-operation.

In the case of FIG. 54, the equipment registered in the state search condition is (tension reel and skid 2). As performed in step 2803, knowledge about the work of the equipment in which each device can participate is searched from the attribute “work” of the knowledge of these devices.

In this example, the tension reel and the skid 2 are searched for (tail end opening work, coil extracting work, coil carrying work, ...). Similar to the procedure shown in FIG. 65, when the retrieved work has a lower work, the lower work is also added to the retrieved work.

Next, the matching condition of the retrieved work knowledge is evaluated. At this time, if the state of the device registered in the target state (state search condition) is not constrained by the work matching condition, it is determined that the device matches. In the case of FIG. 54, since the state of the state item “winding state” of the tension reel is “no coil state”, (tail end opening work, coil removing work, ...) Does not match. After all, since the condition of the tension reel and the state of the skid 2 are not restricted by the matching condition of the coil carrying work, the coil carrying work is goa.
l-operation.

At this time, if a plurality of works match, the most downstream work among the matched works is goal-op.
eration.

At step 2805, the initial-oper
Search all work knowledge between ation and goal-operation and create a work knowledge sequence (work sequence). 52 is an initial state and FIG. 54 is a state search condition, initial-op
Since the eration is the tail end stopping work and the goal-operation is the coil carrying work, the work sequence: (tail end stopping work, tail opening work, coil extracting work, coil carrying work) is obtained in this example.

In this case, the work knowledge at the end of the tree structure of work knowledge, that is, leaves, is collected. If the work knowledge that is not a leaf is initial-operation or goal-operation, the lower-level work is searched for and the most upstream work among the work knowledge that is a leaf is ini.
tial-operation and the most downstream work is goal-operat
Ion.

At step 2806, the time and effort required for creating the control specifications described with reference to FIG. 66 are estimated. Here, the number of work knowledge in the work sequence is used as an index. A boundary value (work knowledge number Border) is set in advance, and it is checked whether the number of work knowledge items in the work sequence exceeds the work knowledge number Border before executing the search. When FIG. 52 is the initial state and FIG. 54 is the state search condition, the number of works is four.

[0304] If the number of work knowledge is the number of work knowledge Border
If so, the process proceeds to step 2814, and if not, the process proceeds to step 2807.

In step 2807, current-oper
Select a scope device from the equipment registered in all the work knowledge between ation and the work knowledge that is the root of the work knowledge.

Work line: (tail end stop work, tail end open work,
In the example of (coil removal work, coil transportation work), when each work is current-operation, the following equipment is selected.

Tail end stop work: (Tension reel, deflector pinch roll) Tail end release work: (Tension reel, coil car, upper tail end press roll, lower tail end press roll) Coil removal work: (tension reel, coil car) Coil transportation Work: (Coil car) In the case of this example, when the procedure shown in FIG. 65 is used, an initial state and a target state across four works are set. In such cases, depending on how the work knowledge is written,
It is not always possible to set the scope accurately. Even if the scope can be set accurately, in the case of the above example, five devices are always searched. In the case of this example, the number can be narrowed down to 5 out of 11, so that the search efficiency can be improved by this alone.

On the other hand, in the case of the procedure shown in FIG. 68, the scope is switched, so in this example, the tail end opening work is
At the time of current-operation, four devices are temporarily in the search range, but other work is current-ope.
At the time of ration, it can be narrowed down to one or two. This makes it possible to further improve the search efficiency.

At step 2807, the equipment state to be transited and the equipment state at the transition destination (new equipment state) are obtained based on the Dijkstra algorithm while being bound by the scope. However, it is not necessary to stick to the Dijkstra algorithm as long as it is an algorithm that solves the shortest path problem of the weighted directed graph.

When the equipment state of FIG. 52 is set to the initial state,
In the processing up to step 2807, the scope first applied is (tension reel, deflector pinch roll).

When the tension reel is selected from the scope, the tension reel state e (FIG. 57) is searched from the equipment state of FIG. Then, from the tension reel state e to the directional branch of the tension reel state e (FIG. 61)
To search. Further, an effective operation of the tension reel (tail end stop control) that can be performed from the tension reel state e is selected from the directed branches of the tension reel state e. The transition destination (FIG. 58) of the device state due to the selected operation is searched.

After considering the state transitions of other devices due to the state transitions of the tension reels, new equipment states are created in a manner that reflects the state of the transitioned equipments in the original equipment state.
At this time, the weight of this operation is searched from the attribute item “weight” of the knowledge of the operation of the tension reel: tail end control, and the equipment state of FIG. 52 is changed to the new equipment state.

Similarly, all equipment states that can transit from the equipment state shown in FIG. 52 are searched for within the scope-constrained range. As a result of the search, the equipment states are further changed in the order closer to the initial state.

[0314] In step 2809, it is checked whether or not the newly generated equipment state satisfies the state search condition. If so, the process proceeds to step 2816. Otherwise, the process proceeds to step 2810.

[0315] In step 2810, it is checked whether the new equipment state satisfies the matching condition of current-operation. If yes, step 2
The process proceeds to 808. Otherwise, step 28
The process proceeds to 11.

In the control specifications of FIG. 64, 2503 and 250
When the equipment state used when generating 4, 2505 is reached, the new equipment state does not satisfy current-operation.

For example, in the equipment state used for producing 2503, the deflector pinch roll is not in the rolled state, and the tension reel is in the winding completed state. Therefore, the matching condition of current-operation: tail end stop work at that time is not satisfied. Therefore, the process proceeds to step 2811 to switch to current-operation: tail end opening work.

At step 2811, the current-oper
If the equipment state that satisfies the target state cannot be searched even if the ation passes the goal-operation, the search process ends. In the procedure shown in FIG. 68, a so-called goal-operation, that is, a rough prediction of the target state is performed. If it deviates from this prediction range, it is determined that the search has failed. Of course, the search, especially the breadth-first search such as Dijkstra's method, searches multiple muscle routes, so if the state at the beginning of the search all deviates from the range of the work queue, a warning is issued. You can give a warning if any one of them deviates. There are many ways to make this judgment, as is done in estimation.

At step 2812, the work immediately after is executed.
Current-operation. Work sequence: (tail end stop work, tail end open work, coil extraction work, coil transportation work) in the example, from tail end stop work to tail end open work, and further from tail end open work to coil extraction work, and so on. , The work knowledge is switched until the equipment state satisfying the state search condition is found.

In step 2813, current-oper
If the goal state cannot be found even after the ation has passed the goal-operation, the user is notified with a warning and the process ends.

In step 2814, the user is warned that the work knowledge of the work sequence exceeds the boundary value (work knowledge number Border).

[0322] In step 2815, if the user desires to continue the control specification creating process, the process advances to step 2807 to continue the process. Otherwise, the process ends.

In step 2816, the operation of the equipment from the initial state of the equipment to the target state and the equipment state of the eye for each operation are determined.

In the path of the equipment operation and the equipment state from the initial state to the target state, the same operation of the same equipment may continue even if the equipment state transits. For example, both control specifications 2506 and 2507 in FIG. 64 are signal names of sensors that represent the position of the coil car in the front-rear direction. The deceleration position 1, the stop position 1 and other positions of the coil car exist between the two. Those states are also searched for in the route of the equipment state.

In this case, it is usually not reflected in the control specifications. Therefore, the equipment state when the operation of the equipment ends is determined by following the route of the equipment state.

At step 2817, a control specification is generated from the operation and the device state of the eye instead of each operation. By performing conversion from the device state as in Japanese Patent Application No. 4-328997, it is possible to obtain a control specification as shown in FIG.

As described above, even when a relatively long control procedure for a complicated controlled object is created, the search range can be narrowed down sufficiently and without omission, and a large scale is carelessly performed. Since it is possible to prevent the search from being performed, and to prevent the search processing from being pointed out accurately and prevent the useless search from being performed, it is possible to contribute to the efficiency improvement of the control program design work.

[0328]

As described above, the information processing apparatus according to the present invention has the grammar using the ontology related to the display target and the vocabulary of the ontology. Therefore, as long as the display target is the one in the ontology category, the GUI designer It is not necessary to create a rule for connecting the graphic operation and the behavior each time the GUI is created, and the desired GUI can be created by only specifying which word in the ontology vocabulary the display target corresponds to. Therefore, the GUI creation efficiency can be improved.

Further, according to the control program design support apparatus of the present invention, since the ontology relating to the controlled object to be displayed and the grammar using the vocabulary of the ontology are provided, the controlled object within the ontology category is to be displayed. As long as each time a screen that displays a controlled object is created, even if a rule that connects the figure operation and the state change of the controlled object intended by the user is not created, which word in the ontology vocabulary the controlled object represented by the figure is Since the target state can be generated only by designating whether or not to correspond to, it is possible to improve the efficiency of creating the control specification.

Further, according to the control program design support apparatus of the present invention, even when a complicated and relatively long control procedure is created, the search range can be narrowed down sufficiently and without any omission. Moreover, it is possible to prevent a large-scale search from being done carelessly, and also to prevent an unnecessary search by accurately pointing out the failure of the search processing, thus contributing to the efficiency improvement of the control program design work. it can.

[Brief description of drawings]

FIG. 1 is a block configuration diagram of an information processing apparatus according to a first embodiment of the present invention.

FIG. 2 is a block configuration diagram of a control program design support device according to a second embodiment of the present invention.

FIG. 3 is a diagram showing a conceptual hierarchy of ontology vocabulary.

FIG. 4 is a diagram showing a conceptual hierarchy of ontology vocabulary.

FIG. 5 is a diagram showing an example of a definition of a word in an ontology vocabulary.

FIG. 6 is a diagram showing a grammar for user intention analysis.

FIG. 7 is a diagram showing a grammar for user intention analysis.

FIG. 8 is a diagram showing a confirmation sentence generation grammar.

FIG. 9 is a diagram showing a grammar for confirmation sentence generation.

FIG. 10 is a diagram showing a grammar for confirmation sentence generation.

FIG. 11 is a diagram showing a screen displaying a certain scene (state) of equipment, processing material, and sensor of plant equipment.

FIG. 12 is a diagram showing a screen displaying a scene (state) of equipment, processing materials, and sensors of plant equipment.

FIG. 13 is a diagram showing a screen displaying a scene (state) of equipment, processing materials, and sensors of plant equipment.

FIG. 14 is a diagram showing a screen displaying a certain situation (state) of plant equipment, processing materials, and sensors.

FIG. 15 is a diagram showing knowledge about devices.

FIG. 16 is a diagram showing knowledge about equipment.

FIG. 17 is a diagram showing knowledge about devices.

FIG. 18 is a diagram showing knowledge about devices.

FIG. 19 is a diagram showing knowledge about the state of a device.

FIG. 20 is a diagram showing knowledge about the state of the device.

FIG. 21 is a diagram showing knowledge about the state of the device.

FIG. 22 is a diagram showing knowledge about the state of the device.

FIG. 23 is a diagram showing knowledge about the state of the device.

FIG. 24 is a diagram showing knowledge about the state of the device.

FIG. 25 is a diagram showing knowledge about processing materials.

FIG. 26 is a diagram showing a target state.

FIG. 27 is a diagram showing a target state

FIG. 28 is a diagram showing a target state.

FIG. 29 shows a processing procedure for receiving a visual instruction from a user and performing a behavior intended by the user, using an analysis grammar described by using a vocabulary of an ontology related to equipment and processed materials in equipment of a plant. Figure

FIG. 30 is a diagram showing a processing procedure corresponding to transfer 2 (receive).

FIG. 31 is a diagram showing a processing procedure until a behavior intended by a user is received upon receiving a visual instruction from the user.

FIG. 32 is a diagram showing a processing procedure corresponding to transfer 2 (receive).

FIG. 33 is a diagram showing a procedure for setting an initial state.

FIG. 34 is a block configuration diagram of a control program design support device according to a third embodiment of the present invention.

FIG. 35 is a block configuration diagram of a control program design support device according to a fourth embodiment of the present invention.

FIG. 36 is a diagram showing an example of a screen displaying devices and processing materials.

FIG. 37 is a diagram showing an example of a screen displaying devices and processing materials.

FIG. 38 is a diagram showing an example of a screen displaying devices and processing materials.

FIG. 39 is a diagram showing an example of a screen displaying devices and processing materials.

FIG. 40 is a diagram showing an example of knowledge about equipment coil cars in equipment.

FIG. 41 is a diagram showing an example of the configuration of the state inside the equipment.

FIG. 42 is a diagram showing an example of the structure of work knowledge.

FIG. 43 is a diagram showing an example of description of work knowledge.

FIG. 44 is a diagram showing an example of symbols that can be used as matching conditions and their meanings.

FIG. 45 is a diagram showing an example of matching conditions.

FIG. 46 is a diagram showing an example of matching conditions.

FIG. 47 is a diagram showing an example of matching conditions.

FIG. 48 is a diagram showing an example of matching conditions.

FIG. 49 is a diagram showing an example of matching conditions.

FIG. 50 is a diagram showing a configuration example of a state space model.

[Fig. 51] Fig. 51 is a diagram showing a description example of the equipment state of the plant.

FIG. 52 is a diagram showing a description example of the equipment state of the plant.

FIG. 53 is a diagram showing a description example of a target state (state search condition).

FIG. 54 is a diagram showing a description example of a target state (state search condition).

FIG. 55 is a diagram showing a description example of the state of the device.

FIG. 56 is a diagram showing a description example of the state of the device.

[Fig. 57] Fig. 57 is a diagram showing a description example of the state of a device.

FIG. 58 is a diagram showing an example of a description of a device state.

[Fig. 59] Fig. 59 is a diagram showing a description example of a device state.

FIG. 60 is a diagram showing an example of a description of a device state.

[Fig. 61] Fig. 61 is a diagram showing a description example of a directional branch emerging from the device state.

[Fig. 62] Fig. 62 is a diagram showing a description example of a directional branch coming out of the device state.

FIG. 63 is a diagram showing one notation example of control specifications.

FIG. 64 is a diagram showing one notation example of control specifications.

FIG. 65 is a view showing the input of the initial state and target state of the controlled object,
Diagram showing the procedure for generating control specifications from the initial state to the target state

FIG. 66 is a diagram showing a procedure for estimating the effort required to generate control specifications.

[FIG. 67] From the input of the initial state and the target state of the controlled object,
Diagram showing the procedure for generating control specifications from the initial state to the target state

[FIG. 68] From the input of the initial state and the target state of the controlled object,
Diagram showing the procedure for generating control specifications from the initial state to the target state

[Explanation of symbols]

 10, 40, 800, 100 ... Knowledge base 20, 50, 900, 1100 ... Processing unit 11 ... Display data 12, 42 ... Ontology 13, 43 ... User intention analysis grammar 14 ... Behavior dictionary 15, 45 ... Confirmation sentence generation Grammar 16 ... Display target state 17 ... Display target 21, 51 ... Input / output device 22, 52 ... Input / output control unit 23, 54 ... User intention analysis unit 24 ... User intention confirmation unit 25 ... Behavior execution unit 41 ... Equipment / processing material -Sensor display data 44 ... Target state generation procedure 46 ... Equipment / machining material / sensor state 47 ... Equipment / machining material / sensor 55 ... Target state generation section 56 ... Initial state setting section 57 ... Control specification generation section 60 ... Control program generation Part 801, 1001 ... Target knowledge 802, 1002 ... State space model 804, 1004 ... Graphic knowledge 805, 1005 ... State structure Knowledge 806, 1006 ... Work knowledge 807, 1007 ... Operation knowledge 808, 1008 ... Secondary state transition 901, 1101 ... Input / output device 902, 1102 ... Input / output control unit 903, 1103 ... Initial state setting unit 905, 1106 ... Target state (state search condition) definition unit 907 ... Scope determination unit 909, 1112 ... Search calculation amount estimation unit 910, 1113 ... State transition route search unit 912, 1120 ... State transition route-control specification conversion unit 913 ... Control program generation device 1110 ... Work queue creation unit

Claims (10)

[Claims] 1. A user intent analysis grammar comprising a screen for displaying an object and capable of operating the screen, an ontology for the object displayed on the screen, and a vocabulary of the ontology. Knowledge holding means having knowledge as a knowledge, an analysis means for analyzing an intention of an operation on the screen using the user intention analysis grammar, and a behavior execution means for executing an operation according to an analysis result of the analysis means. An information processing device characterized by the following. 2. An input means having a screen for displaying an object and capable of operating the screen, an ontology for the object displayed on the screen, and a grammar for user intention analysis described using the vocabulary of the ontology. First knowledge holding means having knowledge as a knowledge, an analysis means for analyzing an intention of an operation on the screen using the user intention analysis grammar, and a behavior execution means for executing an operation according to an analysis result of the analysis means. And a second knowledge holding means having knowledge as a confirmation sentence generation grammar described using the vocabulary of the ontology for generating the analysis result explanation sentence of the analysis means, and causing the behavior execution means to execute the operation. Before, using the grammar of the second knowledge holding means, an explanation sentence of a matter related to the behavior of the behavior executing means is generated, and this explanation sentence is inputted via the input means. The information processing apparatus characterized by comprising; and a behavior confirmation means for notifying. 3. The information processing apparatus according to claim 1, wherein the ontology related to the target also includes an ontology of the state of the target. 4. An input means comprising a screen for displaying a controlled object and a processing material and capable of operating the screen.
Knowledge holding means having knowledge as an ontology related to the controlled object or processed material displayed on the screen and a grammar for user intention analysis described using the vocabulary of the ontology, and the user intention analysis of the intention of the operation on the screen And means for generating an initial state and a target state of a display target from the analysis result, and means for generating a control specification based on the initial state and the target state obtained by this means. A control program design support device characterized by the following. 5. An input device having a screen for displaying a controlled object and a processing material and capable of operating the screen.
A first knowledge holding unit having knowledge as an ontology related to the controlled object or processed material displayed on the screen and a grammar for user intention analysis described using the vocabulary of the ontology, and an intention of an operation on the screen. Means for generating an initial state and a target state to be displayed from the analysis result by analyzing using the user intention analysis grammar, and means for generating a control specification based on the initial state and the target state obtained by this means A second knowledge holding means having knowledge as a confirmation sentence generation grammar described using the vocabulary of the ontology for generating an explanation sentence as a result of analyzing the intention of the operation on the screen; and the control specification. Before the generation, the grammar of the second knowledge holding means is used to generate an explanation sentence of a matter related to the analyzed result, and the generated explanation sentence is passed through the input means. By comprising; and a confirmation means for informing Te control program design support apparatus according to claim. 6. The control program design support apparatus according to claim 5, wherein the ontology related to the controlled object or the processing material includes an ontology of the state of the controlled object or the processing material. 7. A control program design support device for determining a control procedure by performing a search on a model of a state space of a control target, wherein devices in the control target are grouped, and which device group is selected according to the state of the control target. A knowledge holding means having knowledge as a matching condition for determining whether or not is effective, and a scope for determining a scope for limiting a search range from a group of effective devices by using an initial state and a target state of a controlled object. A control program design support apparatus comprising: a determining unit. 8. A control program design support apparatus for determining a control procedure by performing a search on a model of a state space of a control target, wherein devices in the control target are grouped, and which device group is selected depending on the state of the control target. Knowledge holding means having knowledge as a matching condition for determining whether or not is valid, means for determining the order in which the group of devices becomes valid by using the initial state and the target state of the controlled object, A control characterized by comprising a scope determining / switching means for switching a group of valid devices from a state of a control target, determining a scope for limiting a search range based on the group of valid devices, and switching the scope. Program design support device. 9. The control program design according to claim 7, further comprising an estimation means for estimating the calculation amount required for the search from a factor that determines the calculation amount for the search before starting the search. Support device. 10. A predicting means for predicting a failure of a search process from an order in which a group of devices determined by using the initial state and the target state state becomes effective. 8. The control program design support device described in 8.