Automatic Knowledege Acquisition for Multivariable Fuzzy Control Using Neural Network Approach

This paper introduce a simple and systematic scheme capable of self-organizing and self-learning the required control knowledge for use with multivariable fuzzy controllers. The starting point of the approach is to structurally map a simplified fuzzy control algorithm (SFCA) into a counterpropagation network (CPN) in such a way that the control knowledge is explicitly represented in the form of connection weights of the nets, the control rule-base is gradually self-constructed with the fulfillment of the prespecified performance requirements, and finally the approximate reasoning is carried out by replacing a winner-take-all competitive scheme with a soft matching cooperative strategy. Two problems of multivariable control of blood pressure and anaesthesia have been studied as demonstration examples.