【摘要】 提出了一种将小波变换同神经网络相结合的方法,旨在克服由于广为采用的神经网络在对不定信息处理方面存在的不足和容易陷入局部极小值的问题,将小波变换与神经网络相结合,利用混沌轨道的游动性有利于系统跳出局域极值的束缚而寻求全局最优。仿真结果表明,此种算法快速、有效,能很好地解决某些复杂的辨识问题。更多还原

【Abstract】 Due to the shortcoming of neural network in processing uncertain data and tending to local minimum,a chaotic learning algorithm of wavelet is put forward,which is a combination of wavelet transformation and neural network.It is advantageous to the system identification to escape from the limit of local extremum and to find the global optimum.The simulation results show that the algorithm is fast and efficient in complicated identification.更多还原