Expert system development support system and expert system environment utilizing a frame processing tecnique

• US 5,359,701 A
• Filed: 02/17/1993
• Issued: 10/25/1994
• Est. Priority Date: 04/12/1990
• Status: Expired due to Fees

• Current Assignee: Hitachi, Ltd.
• Original Assignee: Hitachi, Ltd.
• Inventors: Watanabe, Dai, Harada, Yasushi, Amano, Masahiko, Kudo, Hiroyuki, Fukui, Chihiro
• Primary Examiner(s): Downs, Robert W.
• Application Number:
  US08/018,581
• Time in Patent Office: 615 Days
• Field of Search: 395/62, 395/903, 395/906, 395/907, 395/50
• US Class Current: 706/53
• CPC Class Codes: G01N 35/0099 comprising robots or similar manipulators robots per se B25J
• CPC Class Codes: G06N 5/027 Frames
• CPC Class Codes: Y10S 706/919 Designing, planning, programming, CAD, CASE
• CPC Class Codes: Y10S 706/921 Layout, e.g. circuit, construction

Abstract

An expert system in which a plurality of knowledge descriptive layers are provided in a flame system. Each descriptive layer is composed of several flames. The flames of the lower knowledge descriptive layer correspond to the elements of a knowledge processing object system. The content of frames of the upper knowledge descriptive layers is determined as an inference made with reference to the frames of the lower layer. This determination is initiated and executed when the content of slots of the flames is updated.

Claims

1. An expert system for processing knowledge of an object system that has a plurality of elements, comprising:

a) a knowledge base including a plurality of knowledge elements each of which is a framework for representing knowledge, and comprising a plurality of logical layers including;

i) a unit knowledge element descriptive layer comprising a portion of said plurality of knowledge elements each of which contains knowledge for representing a portion of a first model of the object system, said portion of the first model of the object system corresponding to each one of the plurality of elements of the object system, respectively; and

(ii) a summarized knowledge element descriptive layer being hierarchically arranged above said unit knowledge element descriptive layer, and comprising a second portion of said plurality of knowledge elements each of which contains knowledge for representing a portion of a second more abstract model of the object system than said first model, said knowledge elements of said second portion corresponding to knowledge of at least one of the knowledge elements of said first portion; and

b) a processor for generating knowledge from the knowledge including updated knowledge contained in the knowledge elements included in the unit knowledge element descriptive layer, to represent the second model of the object system, and storing the generated knowledge in the knowledge element included in the summarized knowledge element descriptive layer, said knowledge element in which the generated knowledge is to be stored corresponding to the knowledge element containing the updated knowledge, said updating being based on a change of status in the object system or on an explicit instruction;

wherein each of said knowledge elements is a frame having a plurality of slots each of which contains a slot value, said plurality of slots being comprised of different types of slots, and the expert system further comprising an indexer which includes;

(i) a plurality of frame managing tables each of which corresponds to each of the types of slot and manages said plurality of frames according to an ordering by the evaluated slot values resulting from an evaluation function executed on the slot values;

(ii) a slot type managing table for managing said plurality of frame managing tables according to an ordering by the types of slots corresponding to the frame managing table;

(iii) a reorganizer reorganizing said frame managing table in response to updating of said slot values, wherein said reorganizing is performed by executing said evaluation function on the updated slot values and changing the content of the frame managing table according to the evaluated slot values resulting from the evaluation function executed on the updated slot values; and

(iv) a retriever retrieving the frame managing table corresponding to the designated slot type by referring to said slot type table according to the designated slot type, retrieving a set of frames according to the designated evaluated slot value by referring to the retrieved frame managing table according to the designated evaluated slot value.

2. An expert system for processing knowledge of an object system that has a plurality of elements, comprising:

a) a knowledge base containing the knowledge of the object system, said knowledge base having a plurality of frames, each of said plurality of frames having a plurality of slots, each of said plurality of slots containing a slot value, wherein said plurality of slots is comprised of different types of slots; and

b) an indexer including;

(i) a plurality of frame managing tables each of which corresponds to one of the types of slot and manages said plurality of frames according to an ordering by the evaluated slot values resulting from an evaluation function executed on the slot values;

(ii) a slot type managing table for managing said plurality of frame managing tables according to an ordering by the types of slots corresponding to the frame managing table;

(iii) a reorganizer reorganizing said frame managing table in response to updating of said slot values, wherein said reorganizing is performed by executing said evaluation function on the updated slot values and changing the content of the frame managing table according to the evaluated slot values resulting from the evaluation function executed on the updated slot values; and

(iv) a retriever retrieving the frame managing table corresponding to the designated slot type by referring to said slot type table according to the designated slot type, retrieving a set of flames according to the designated evaluated slot value by referring to the retrieved frame managing table according to the designated evaluated slot value.

3. The expert system according to claim 2, further comprising:

c) a plurality of rules of inference being located within said knowledge base and each including;

(i) an instruction section having a plurality of instructions defined therein; and

(ii) a condition section having a plurality of conditional clauses and defining a second condition which selects an instruction to be executed; and

d) an inference engine performing a judgment of said second condition and providing said indexer with a designation of the slot type and the evaluated slot value to retrieve a set of frames needed for said judgment, selecting an instruction according to the result of said judgment, and executing the selected instruction.

4. The expert system according to claim 2, wherein if one of said slot values is updated by selected instruction which is executed by said inference engine, then said reorganizer reorganizes said frame managing table.

6. An expert system according to claim 2, wherein said indexer comprises a second retriever, said second retriever retrieving the frame managing table corresponding to each designated type by referring to said slot type table according to the designated slot type, retrieving a plurality of subsets of frames respectively according to a plurality of designated evaluated slot values by referring to the retrieved frame managing table according to the designated evaluated slot values, and obtaining a result set of frames by performing an operation on the retrieved subsets, wherein said operation is determined according to a designated first condition.

7. The expert system according to claim 6, further comprising:

a) a plurality of rules of inference being located within said knowledge base including;

(i) an instruction section having a plurality of instructions defined therein; and

(ii) a condition section having a plurality of conditional clauses and defining a second condition which selects an instruction to be executed; and

b) an inference engine performing a judgment of said second condition and providing said indexer with a designation of the slot type, evaluated slot values and said first condition to retrieve a set of frames needed for said judgment, selecting an instruction according to the result of said judgment, and executing the selected instruction.

5. An expert system according to claim 22, wherein said object system is an electric power system.

8. An expert system for processing knowledge of an object system that has a plurality of elements, comprising:

a) a knowledge base containing the knowledge of the object system, said knowledge base having a plurality of frames, each of said plurality of frames having a plurality of slots, each of said plurality of slots containing a slot value, wherein said plurality of slots is comprised of different types of slots; and

b) an indexer including;

(i) a plurality of frame managing tables each of which corresponds to each of the types of slots and manages said plurality of frames according to an ordering by the slot values;

(ii) a slot type managing table for managing said plurality of frame managing tables according to an ordering by the types of slots corresponding to the frame managing table;

(iii) a reorganizer reorganizing said frame managing table in response to updating of said slot values, wherein said reorganizing is performed by changing the content of the frame managing table according to the updated slot values; and

(iv) a retriever retrieving the frame managing table corresponding to the designated slot type by referring to said slot type table according to the designated slot type, retrieving a plurality of subsets of frames respectively according to a plurality of designated slot values by referring to the retrieved frame managing table according to the designated slot values, and obtaining a result set of frames by performing an operation on the retrieved subsets, wherein said operation is determined according to a designated first condition.

9. The expert system according to claim 8, further comprising:

c) a plurality of rules of inference being located within said knowledge base and each including;

(i) an instruction section having a plurality of instructions defined therein; and

(ii) a condition section having a plurality of conditional clauses and defining a second condition which selects an instruction to be executed; and

d) an inference engine performing a judgment of said second condition and providing said indexer with a designation of the slot type, evaluated slot values or slot values, and said first condition, to retrieve a set of frames needed for said judgment, selecting an instruction according to the result of said judgment, and executing the selected instruction.

10. An expert system for processing knowledge of an object system that has a plurality of elements, comprising:

a) a knowledge base containing the knowledge of the object system, said knowledge base having a plurality of frames, each of said plurality of flames having a plurality of slots, each of said plurality of slots containing a slot value, wherein said plurality of slots is comprised of different types of slots; and

b) an indexer including;

(i) a plurality of frame managing tables each of which corresponds to each of a plurality of predetermined evaluation functions for slot values and manages said plurality of flames according to an ordering by the evaluated slot values resulting from an corresponding evaluation function executed on the slot values;

(ii) a plurality of evaluation function managing tables each of which corresponds to each of the types of slot and manages said plurality of frame managing tables according to an ordering by the evaluation functions corresponding to the frame managing table;

(iii) a slot type managing table for managing said plurality of evaluation function managing tables according to an ordering by the types of slots corresponding to the evaluation function managing tables;

(iv) a reorganizer reorganizing said frame managing tables in response to updating of said slot values, wherein said reorganizing is performed by executing said evaluation functions on the updated slot values and changing the content of each frame managing table according to the evaluated slot values resulting from the evaluation function according to the frame managing table; and

(v) a retriever retrieving the evaluation function managing table corresponding to the designated slot type by referring to said slot type table according to the designated slot type, retrieving the frame managing table corresponding to the designated evaluation function by referring to the retrieved evaluation function managing table according to the designated evaluation function, and retrieving a set of frames according to the designated evaluated slot value by referring to the retrieved frame managing table according to the designated evaluated slot value.

11. An expert system according to claim 10, wherein said indexer comprises a second retriever, said second retriever retrieving the evaluation function managing table corresponding to the designated type by referring to said slot type table according to the designated slot type, retrieving the frame managing tables corresponding to the designated evaluation functions by referring to the retrieved evaluation function managing table according to the designated evaluation functions, retrieving subsets of frames according to each of the designated evaluation functions by referring to each of the retrieved frame management tables according to the one of the designated evaluated slot values which corresponds to the evaluation function corresponding to the frame managing table, and obtaining a result set of frames by performing an operation on the retrieved subsets, wherein said operation is determined according to a designated first condition.

12. The expert system according to claim 10, further comprising:

c) a plurality of rules of inference being located within said knowledge base and including;

(i) an instruction section having a plurality of instructions defined therein; and

(ii) a condition section having a plurality of conditional clauses and defining a second condition which selects an instruction to be executed; and

d) an inference engine performing a judgment of said second condition and providing said indexer with a designation of the slot type, evaluation functions, evaluated slot values of the evaluation functions, and said first condition to retrieve a set of frames needed for said judgment, selecting an instruction according to the result of said judgment, and executing the selected instruction.