具有混沌机制的模糊神经网络非线性系统建模研究

针对常规的BP算法收敛速度慢及容易陷入局部极小的缺点,在该算法中引入具有混沌机制的非线性自反馈项,给出了混沌BP算法,并利用其训练和学习模糊神经网络中的权值,从而构成一种引入型模糊混沌神经网络;最后,用提出的引入型模糊混沌神经网络对非线性系统进行仿真研究,仿真结果表明,所设计的引入型模糊混沌神经具有与混沌动力学特性同样复杂的动态特性,其对非线性系统具有很好的建模能力.

The Modeling Study of Nonlinear System of Fuzzy Neural Network with Chaotic Mechanism

For the shortcomings of conventional BP algorithm which has slown convergence and falls into local minimum easily,the nonlinear self-feedback term with chaotic mechanism was introduced into this algorithm.Thus chaotic BP algorithm（CBPA） was given.The weight of fuzzy neural network（FNN） was trained and learned.An introduction-type fuzzy chaotic neural network（IFCNN） was constituted.The simulation of nonlinear system given based on IFCNN was proposed.Simulation results show that the designed IFCNN has the same and complex dynamic characteristics with chaotic system,which has good modeling capabilities for nonlinear system.