基于PSO-BP神经网络算法的绞线串扰预测研究

Research on Stranding Crosstalk Prediction Based on PSO-BP Neural Network Algorithm

绞线串扰预测研究对定量分析系统的电磁兼容问题起着重要作用,本文提出了基于粒子群算法(particle swarm optimization,PSO)优化误差逆向传播(back propagation,BP)神经网络算法的绞线串扰预测方法。首先,通过模量解耦法对建立的传输线等值电路进行求解,得到传输线的串扰;然后,利用PSO-BP神经网络算法实现绞线电磁参数矩阵提取,并表示其与绞线位置的映射关系;最后将该映射关系代入传输线方程,结合模量解耦法预测一实例三芯绞线的串扰。通过与传输线矩阵求得的数值解对比可知,本文所提方法与电磁场数值方法所求的串扰结果具有良好的一致性,验证了基于PSO-BP神经网络算法预测绞线串扰的有效性。

Stranding crosstalk prediction plays an important role in the electromagnetic compatibility of quantitative analysis system.In this paper,a stranding crosstalk prediction method based on Particle Swarm Optimization(PSO)algorithm optimized Back Propagation(BP)neural network algorithm was proposed.Firstly,the modulus decoupling method was used to solve the established transmission line equivalent circuit to get the crosstalk of the transmission line.Then PSO-BP neural network algorithm was used to extract the stranded electromagnetic parameter matrix and express its mapping relationship with the stranded position.Finally,the mapping relation was substituted into the transmission line equation,and the crosstalk of an example three-core stranded wire was predicted by combining the modulus decoupling method.Compared with the numerical solution obtained by the transmission line matrix,the crosstalk results obtained by the proposed method are in good agreement with those obtained by the electromagnetic field numerical method,which verifies the effectiveness of the PSO-BP neural network algorithm to predict stranded crosstalk.