The Intelligent Learning Algorithm Of Fuzzy Cognitive Maps And Its Application Research

Fuzzy Cognitive Maps (FCM) is introduced by Kosko. It is soft computing tools and combines Fuzzy Sets which is bring forward by Zadeh and Neural Network. Its information express and reasoning capacity is better than before. The concept of FCM can express action, reason, result, motive, sensation and tendency of system; it can image the attribute, capability and character of system.Firstly, the development of Fuzzy Cognitive Maps, the existing learning methods, Genetic Algorithm and Particle Swarm Optimization Algorithm are introduced. Secondly, after analyzing the shortage of existing learning methods, two novel methods of learning algorithm are produced. The two methods are: Genetic Algorithm and Quantum-behaved Particle Swarm Optimization. The most tasks of the two novel learning procedures are to find a set of the FCM's weights, which leads the FCM to a desired steady state. With introduction of two novel algorithms, the minimization of a properly defined objective function are achieved, the workings of the approach are applied to an industrial control problem. The results support the claim that the proposed technique is a promising methodology for Fuzzy Cognitive Maps learning, and the methodology is effective and efficient.