Expert System Tools OKPS

The knowledge base system of expert system tool OKPS adopts object-oriented knowledge representation method. Object-oriented knowledge representation method is a knowledge representation method based on the framework knowledge representation, combined with semantic network representation, and applying object-oriented concept definition. This representation method regards knowledge as an object class, encapsulates the attributes of objective things and laws and their behavior characteristics, and expresses their structure and connection through the inheritance and constraint relationship between objects.


In the design and development of OKPS, we completely adopt the object-oriented design method and programming technology. Each part of the knowledge processing system is relatively independent, and they are encapsulated in the corresponding classes. The message control mechanism of the object is used to interface between each part, including the interface between the inference engine and the ICL interpreter, the interface between the inference engine and the tool library, etc.

OKPS has the following functions:

  • Manage the knowledge base, acquire knowledge through VKMT, and use ODBC to access the knowledge base;


  • Reasoning and search;


  • Interpretation and implementation of methods and monitoring;


  • Participate in problem solving as an intelligent model;


  • In the problem processing system, intelligent scheduling;


  • Human computer interaction, external database access and blackboard control.

OKPS mainly consists of four parts:

  • Object Oriented Inference Engine

  • ICL Command Interpreter

  • Tool Library

  • Visual Knowledge Management Tool VKMTOKPS

Application Examples:

1. Expert System for Fishery Resources Evaluation in the East China Sea

System Function: the key of fishery resources evaluation and dynamic prediction is to select the best evaluation and prediction model according to specific objectives and specific sea areas. Practice has proved that the existing empirical models or classical models must be selected and improved for specific application conditions in order to make the evaluation and prediction reliable and practical. The working principle of the expert system for fishery resources assessment is to use the model base and expert knowledge in the expert system to select, fit and calculate the model according to the existing statistical data (yield, catch, etc.), biological parameters of fish groups, needs and objectives of assessment, and finally use the expert knowledge of the system to verify the fitted model, and judge and provide necessary explanations for the results of model calculation.


The expert system evaluates and predicts the main fish species (hairtail, mackerel, puffer fish), main fishing seasons or annual resources, maximum sustainable production and allowable catch in different sea areas, and the results are provided in the form of statements or charts.


The project achievements won the second prize of the scientific and technological progress award of the Chinese Academy of Sciences in 2001 and the second prize of the National Scientific and Technological Progress award in 2002.

2. Object-Oriented Intelligent Fault Diagnosis Expert System

This system is used for fault diagnosis during automatic test of equipment. The diagnosis software mainly obtains the data required for diagnosis by reading the database, and diagnoses the faults occurred in the test process, such as the diagnosis of communication fault between the automatic test system and the main control computer, the diagnosis of the fault of the integrated control device of the power system, and the diagnosis of the fault of independent small components in the electrical system on the equipment, Finally, analyze the data collected by the test system, including drawing data curves, making qualitative analysis of the curves, and displaying the analysis results.


The system has been highly evaluated by a department of the Ministry of space.


3. ICBC Fixed Asset Loan Risk Decision System

This system is an interactive system, that is, in the process of decision-making, it puts forward some questions that need to be answered by numbers to users, and the interface is simple and friendly. In the process of man-machine dialogue, the system requires users to input some parameters for decision-making in numerical form, such as enterprise operator quality evaluation, economic strength, capital structure, economic benefits, development prospects, credit rating coefficient, loan amount, etc. At the same time, some options are given for users to choose, such as mortgage loan, guaranteed loan, credit loan, and loan form. The system calculates the risk weight of all loan assets and the risk degree of all fixed assets loan assets according to the information provided by the user, and gives an evaluation chart combined with the credit rating of the enterprise, and finally gives suggestions on whether to loan or not.



Download File: