A fuzzy brain emotional learning model based on particle swarm algorithm was proposed to improve the accuracy of neu-ral network model in solving nonlinear system recognition problems.The model contained a brain emotional learning network,and based on the training of the model using the system historical data,the adaptation function made the dynamic adjustment of the weight factors in the network structure to improve the network learning efficiency and recognition accuracy.In the identification test of continuous stirred tank reactor and the approximation test of sinE strong nonlinear object,compared with the conventional fuzzy brain emotional learning model,BP neural network and RBF neural network,this model had higher approximation ability and faster convergence speed,and solved the problems of long adjustment time of model parameters and model instability caused by trial-and-error based method,which provided a feasible model for practical application of recognition.
关键词
粒子群算法/类脑神经网络/大脑模糊情感学习模型/神经网络系统辨识/非线性系统
Key words
particle swarm algorithm/brain-like neural network/fuzzy brainemotional learning model/neural network system iden-tification/nonlinear system
维普
万方数据