论文部分内容阅读
小波神经网络(WNN)具有较强的逼近和容错能力,并具有良好的收敛性和鲁棒性。然而其网络收敛速度慢、搜索成功率低和易陷入局部极小值等缺点使得传统的小波神经网络难以得到广泛应用。本文介绍一种基于粒子群(PSO)算法的小波神经网络,其通过利用种群间信息共享进行寻优,以获得结构化的神经网络,克服了传统小波网络的诸多缺点,结合工程实例,检验了其具有较好的适用性和可靠性。
Wavelet neural network (WNN) has strong approximation and fault tolerance, and has good convergence and robustness. However, its slow convergence rate, low search success rate, and its vulnerability to local minima make the traditional wavelet neural network difficult to be widely used. This paper introduces a wavelet neural network based on Particle Swarm Optimization (PSO) algorithm, which uses the information sharing among the populations to optimize, to obtain a structured neural network and overcomes many shortcomings of the traditional wavelet network. Combined with engineering examples, It has good applicability and reliability.