基于BP神经网络的弱信号提取方法研究

A Method of Extracting Weak Signal Based on BP Algorithm

提出一种基于BP神经网络结构与算法的弱信号提取方法，该方法在网络节点连接权向量域，从宽带背景噪声中提取弱信号，且对信号的先验知识没有要求。文中具体给出了BPWV方法的原理及推导过程。仿真结果表明，该方法能显著地提高信噪比。

A new method is presented based on BP algorithm of artificial neural network (ANN). The method can be used to extract weak signal from width band noise in net node linking power vector donlain.No a prior knDWledge about the signal is required. The principle and the deriving process about the BPWV method have been given in the paper. Simulation results show that it can improve signal-noise ratio rema rkably.