Knowledge information generating system

• US 5,414,798 A
• Filed: 01/29/1990
• Issued: 05/09/1995
• Est. Priority Date: 06/17/1988
• Status: Expired due to Term

• Current Assignee: Hitachi, Ltd.
• Original Assignee: Hitachi, Ltd.
• Inventors: Fukai, Masayuki, Namba, Shigeaki, Nigawara, Seiitsu
• Primary Examiner(s): Downs, Robert W.
• Application Number:
  US07/460,093
• Time in Patent Office: 1,926 Days
• Field of Search: 364/513, 395/68, 395/75, 395/906, 395/51
• US Class Current: 706/59
• CPC Class Codes: G05B 13/028 using expert systems only
• CPC Class Codes: Y10S 706/906 Process plant

Abstract

A knowledge information generating system including: a fundamental rule storage for making and storing, for each process data to be controlled, another process data relating to the former and the relations between these process data in advance as fundamental rules; a process data input processor for inputting and storing the process data to be controlled; a knowledge information collection necessity decider for detecting that the process data obtained from the process data input processor have changed or deviated from a predetermined relation; a knowledge information editor for fetching the fundamental rules relating to the process data, which are decided to be collected by the knowledge information collection necessity decider, from the fundamental rule storage, to edit the process data on the basis of the relations between the data group stored in the process data input processor and the process data stored in the fundamental rules; and a knowledge data base for storing the knowledge information obtained from the knowledge information editor.

Claims

1. A knowledge information generating system comprising:

means for storing, for each of first process data indicating a current state of a subject of control to be monitored, second process data related to said first process data and fundamental rules indicating relations between said first and second process data, said second process data representing previously collected information concerning previous states of said subject of control;

means for fetching said fundamental rules based on said first process data indicating that a current state of the subject of control is abnormal; and

means for generating knowledge information to be used to operate said subject of control based on said fetched fundamental rules, said first process data input from said subject and said second process data.

2. A knowledge information generating system comprising:

a fundamental rule storage device for storing, for each of first process data indicating a current state of a subject of control to be monitored, second process data related to said first process data and fundamental rules indicating relations between said first and second process data, said second process data representing previously collected information concerning previous states of said subject of control;

a process data input processor for inputting said first process data from said subject of control;

a knowledge information collection necessity decider for deciding whether process data for knowledge information generation is to be collected by detecting whether said first process data input by said process data input processor has either changed from a normal state or deviated from a predetermined relation between said first process data and said second process data beyond a preset threshold value and outputting a necessary signal when it is detected that said first process data has either changed or deviated;

a knowledge information editor for fetching said fundamental rules and said process data from said fundamental rule storage device in response to said necessary signal, editing said process data based on relations between said first process data input by said process data input processor and said second process data indicated by said fundamental rules, and generating knowledge information for use in operating said subject of control based on said edited second process data; and

a knowledge data base for storing said knowledge information generated by said knowledge information editor.

3. A knowledge information generating system comprising:

a fundamental rule storage device for storing, for each of first process data indicating a current state of a subject of control, second process data related to said first process data and fundamental rules indicating relations between said first and second process data, said second process data representing previously collected information concerning previous states of said subject of control;

a process data input processor for inputting first process data from said subject of control;

a knowledge information collection necessity decider for deciding whether process data for knowledge information generation is to be collected by detecting whether said first process data input by said process data input processor has either changed from a normal state or deviated from a predetermined relation between said first process data and said second process data beyond a preset threshold value and outputting a necessary signal when it is detected that said first process data has either changed or deviated;

a knowledge information editor for fetching said fundamental rules and said process data from said fundamental rule storage device in response to said necessary signal, editing said process data based on relations between said first process data input by said process data input processor and said second process data indicated by said fundamental rules, and generating knowledge information for use in operating said subject of control based on said edited second process data;

an information adder for adding additional information input by an operator to said knowledge information generated by said knowledge information editor;

a knowledge data base for storing generated knowledge information and said additional information input by said operator; and

a fundamental rule information processor for correcting said fundamental rules, and storing said corrected fundamental rules in said fundamental rule storage device.

4. A fundamental rule correcting method comprising the steps of:

storing, for each of first process data indicating a current state of a subject of control to be monitored, second process data related to said first process data and fundamental rules indicating relations between said first and second process data, said second process data representing previously collected information concerning previous states of said subject of control;

extracting from first process data input from said subject of control, pairs of first process data having state values indicating a relation between said pair of first process data representing a physical quantity flow of said subject of control;

comparing said extracted pairs of first process data to related pairs of second process data indicated by said fundamental rules; and

correcting fundamental rules based on said extracted pairs of first process data and said related paris of second process data.

5. A fundamental rule correcting method comprising the steps of:

storing, for each of first process data indicating a current state of a subject of control to be monitored, second process data related to said first process data and fundamental rules indicating relations between said first and second process data, said second process data representing previously collected information concerning previous states of said subject;

extracting from first process data input from a subject of control, pairs of first process data having state values indicating a functional relation between said pairs of first process data;

comparing said extracted pairs of first process data to related pairs of second process data indicated by said fundamental rules; and

correcting said fundamental rules based on said extracted pairs of first process data and said related pairs of second process data.

6. A fundamental rule correcting method comprising the steps of:

storing, for each of first process data indicating a current state of a subject of control to be monitored, second process data related to said first process data and fundamental rules indicating relations between said first and second process data, said second process data representing previously collected information concerning previous states of said subject of control;

extracting, from first process data input from said subject of control, pairs of first process data having state values indicating a source and target relation between said pairs of first process data representing an influence propagation route of a physical quantity of said subject of control;

comparing said extracted pairs of first process data and said second process data based on said fundamental rules; and

correcting said fundamental rules based on said extracted pairs of first process data and related pairs of second process data.

7. A fundamental rule correcting method comprising the steps of:

storing, for each of first process data indicating a current state of a subject of control to be monitored, second process data related to said first process data and fundamental rules indicating relations between said first and second process data, said second process data representing previously collected information concerning previous states of said subject of control;

extracting, from first process data input from said subject of control, pairs of first process data having state values indicating operational relations between a running state of said subject of control and an operating state of auxiliary equipment of said subject of control;

comparing said extracted pairs of first process data to related pairs of second process data indicated by said fundamental rules; and

correcting said fundamental rules based on said extracted pairs of first process data and said related pairs of second process data.

8. An expert system for controlling a subject of control by using knowledge information effective for running said subject of control, comprising:

means for storing, for each of first process data indicating a current state of a subject of control to be monitored, second process data related to said first process data of said subject of control and fundamental rules indicating relations between said first process data and said second process data, said second process data representing previously collected information concerning previous states of said subject of control;

means for inputting first process data from said subject of control when said subject of control is running;

means for generating knowledge information based on said inputted first process data, said second process data and said fundamental rules;

means for displaying said generated knowledge information for viewing by an operator of said subject of control; and

means for applying said generated knowledge information to a controller for controlling the running of said subject of control.

9. A logger system for outputting knowledge information for use in controlling a subject of control, comprising:

means for storing, for each of first process data indicating a current state of a subject of control to be monitored, second process data related to said first process data and information indicating relations between said first and second process data as fundamental rules, said second process data representing previously collected information concerning previous states of said subject of control;

means for fetching said fundamental rules based on said first process data indicating that a current state of the subject of control is abnormal;

means for obtaining knowledge information, which includes time data indicating time periods between change of said first process data, based on said fetched fundamental rules and said second process data; and

means for outputting said obtained knowledge information in a predetermined interval and on demand.