论文部分内容阅读
ＢＰ网络广泛应用于函数逼近、模式识别和系统辨识，但ＢＰ算法收敛速度很慢。为此提出了ＢＰ算法的一种新的改进方式，即在误差反向传播时，不仅改变网络的联接权值，也改变神经元模型参数。详细推导了改进ＢＰ算法的迭代公式。仿真研究表明，与传统ＢＰ算法相比，该算法具有收敛速度快，函数逼近精度高的优点。
BP network is widely used in function approximation, pattern recognition and system identification, but the BP algorithm converges very slowly. To this end, a new improved BP algorithm is proposed. That is, when the error is propagated backwards, not only the weight of the network but also the parameters of the neuron model are changed. The iterative formula of improving BP algorithm is deduced in detail. Simulation results show that compared with the traditional BP algorithm, the proposed algorithm has the advantages of fast convergence and high approximation of the function.