| 摘要 |
This invention relates to the design and implementation of real-time knowledge-based fuzzy controller system for general purpose industrial applications. The controller is implemented in three technological bases (software, VLSI and neural network). In another aspect the invention relates to the design of an Intelligent system which implements decision-making procedure based on approximation, association and reasoning with fuzzy patterns and their clearness assessments rather than the use of Max-Min computation over fuzzy relational matrices usually applied in approximate reasoning procedures in similar systems. According to this design the fuzzy controller is a device which operates at the level of fuzzy pattern processing where each control task is expressed through the attributes of fuzzy patterns (syntax and content, domain and clearness measure), and the elementary cognitive activities which the human performs with these patterns such as: recognition, generation, assessment, association, pattern matching, approximation, etc. The fuzzy controller utilizes a new scheme for approximate reasoning with fuzzy patterns called the Clearness Transformation Rule of Inference (CTRI). This mechanism offers a spectrum of advantages broadening the functional Intelligence of the controller to handle complex human tasks, improving the performance and accuracy of the controller and reducing the computational requirements. The fuzzy controller presented in the invention can be applied in general engineering practice, financial, medical, process control, pattern recognition and other areas requiring knowledge-based bahaviour in decision making. |
