Fuzzy models of intelligent industrial controllers and control systems. III: Design procedure (Q1901558)

The paper deals with problems of designing industrial controllers and control systems with various degrees of intelligence. Fuzzy models of controllers, particularly control systems that are `intelligent in the large', are of special interest in the paper. Simulation, learning, and adaptation by neural networks that make it possible to formulate a special knowledge base on the intelligence level used are the basis of the procedure for designing control systems with different intelligence levels. Examples of designed systems are such as a learning and adaptable neural network, an approximate reasoning unit, a knowledge base generator, and a fuzzy generator and interpreter.