Training type-2 fuzzy system by particle swarm optimization

Many Intelligent techniques were established during last decades to handle nonlinear, multimode, noisy, nondifferentiable problems and to obtain optimum solution(s). This paper presents improving and implementations for two recently intelligent techniques; Type-2 Fuzzy System (T2 FS) and Particle Swarm Optimization (PSO) and presents a new method to optimize parameters of the primary membership functions of T2 FS by PSO to improve the performance and increase the accuracy of T2 FS model. The implementation of the suggested method on mean blood pressure estimation has very successful rate. © 2007 IEEE