Computer-based diagnostic expert system organized according to Bayesian theory

• US 5,133,046 A
• Filed: 01/03/1991
• Issued: 07/21/1992
• Est. Priority Date: 01/03/1991
• Status: Expired due to Fees

• Current Assignee: Pickard LOWE and Carrick
• Original Assignee: Pickard LOWE and Carrick
• Inventors: Kaplan, Stanley
• Primary Examiner(s): MacDonald, Allen R.
• Application Number:
  US07/637,190
• Time in Patent Office: 565 Days
• Field of Search: 364/513, 364/550, 364/551.02, 395/54, 395/61, 395/912, 395/914
• US Class Current: 706/52
• CPC Class Codes: G05B 23/0254 based on a quantitative model, e.g. mathematical relationships between inputs and outputs; functions: observer, Kalman filter, residual calculation, Neural Networks
• CPC Class Codes: G06F 11/2257 using expert systems
• CPC Class Codes: G06F 17/18 for evaluating statistical data , e.g. average values, frequency distributions, probability functions, regression analysis forecasting specially adapted for a specific administrative, business or logistic context G06Q10/04
• CPC Class Codes: Y10S 706/912 Manufacturing or machine, e.g. agricultural machinery, machine tool
• CPC Class Codes: Y10S 706/914 Process plant

Abstract

An expert system is organized according to Bayes'"'"' theorem. The system includes a diagnostic module that generates a diagnosis in the form of probability distributions. The diagnostic module also is responsive to evidence in the form of discretized time trajectories in the value space of the observable variables. The diagnostic module is also responsive to data from a knowledge base that represents the trajectories of the possible states of the system and the associated values of the likelihood that the various trajectories of observed evidence would be present if the system were in that state.

Claims

1. An apparatus for determining the status of a system together with the probability of that status from time-dependent observables of the system, said apparatus comprising a computer, said computer including first memory means for storing a knowledge base, said knowledge base comprising data representative of a set of possible states of a system in association with probabilities that time-variable patterns of said observables will be present given the occurrence of such states, said apparatus including means for obtaining observables indicative of the operation of the system at a succession of different times and means for representing said observables in a time-succession of digitized codes, said apparatus also including diagnostic means responsive to data selected from said knowledge base and to said succession of digitized codes for generating a second set of digitized codes representing probability distributions over the set of said states and for updating said probability distributions as a function of time.

2. An apparatus as set forth in claim 1 including display means for presenting said probability distributions visually.

3. An apparatus as set forth in claim 1 wherein said knowledge base in said first memory means is organized into explicit, residual and impossible subsets and said memory means includes a first output means for applying to said diagnostic means digitized codes representative of said explicit, residual and impossible subsets.

4. An apparatus as set forth in claim 1 wherein said means for representing said time-succession of digitized codes comprises an A/D converter and a clock means, said clock means being operative to enable said A/D converter to generate time-dependent sequences of codes, said apparatus including second memory means for storing said time-dependent sequences of codes.

5. An apparatus as set forth in claim 4 wherein said second memory means includes a second output means for applying to said diagnostic means digitized codes representative of said time-dependent sequences of codes and said diagnostic module is responsive to outputs from said first and second memory means for generating options and associated probability distributions.

6. A method of diagnosing the instant state of a system having a state which varies with time using a computer having a memory, comprising the steps of storing in said memory a knowledge base comprising sets of digitized data representative of the possible states of said system in time-dependent form in association with the probabilities that certain observables will be present given the occurrence of the associated state, inputting into said computer evidence in a time-dependent form, and providing a probability distribution of the possible states of said system responsive to said data from said knowledge base and to said evidence in time-dependent form.

7. A method as set forth in claim 6 wherein said step of storing in said memory includes the step of discreting the data in said knowledge base in digital strings representing a range of values and a point in time in said string.

8. A method as set forth in claim 7 wherein said step of inputting into said computer includes the step of organizing said evidence into scenerio strings of digitized data.

9. A method as set forth in claim 8 wherein said step of providing a probability distribution includes the step of matching each of said scenerio strings with the digital strings in said knowledge base.

10. A method as set forth in claim 9 also including the step organizing the strings of digitized data in said knowledge base into categories, each one of said categories corresponding to a family of problem initiators.