The evolutionary-programming learning of linguistic fuzzy model for nonlinear system and designing of configuration for neural network

With the difficulty in building model of nonlinear system, the improving evolutionary programming in this paper is used to obtain a linguistic fuzzy model of the system by input-output data. A hierarchical evolutionary programming for RBF neural networks design is also proposed to train network configuration and parameters, overcoming shortcoming of grade algorithm The effectiveness of the evolutionary programming and the hierarchical evolutionary programming is proved through simulating SISO and MISO system.