多导体线束内串扰概率分布的预测

摘要

串扰是电气、电子系统内部多导体传输线间的相互电磁干扰,受其影响系统可靠性往往较差.电缆线束作为典型的多导体传输线,其串扰问题显得尤为突出.针对电缆线束内导线位置的不确定性,提出一种线束内串扰概率分布的预测方法,在假定线束横截面固定的条件下,将线束分为N段,计算单一子段的电磁参数分布律,并利用卷积算法求得整个线束的电磁参数概率分布;基于蒙特卡洛法,结合多导线传输线理论得到线束串扰的概率分布,进而帮助设计者发现系统内潜在的串扰问题、评估遭受串扰影响的系统可靠性.为了得到卷积算法中分段数N的合适取值,以线束的扭绞程度为依据,提出了相应的经验公式,并结合实际算例进行测量.实验结果表明,线束的分段数N应与线束的扭绞程度相适应;线束的扭绞程度越高,线束内串扰的概率分布越集中,有趋于"平均化"的规律.%Crosstalk is a kind of mutual electromagnetic interference(EMI) among transmission lines in electric or electronic system,which often leads to a low reliability of the system.As typical multiconductor transmission lines(MTLs),cable bundles have a prominent problem of crosstalk.Considering the uncertainty of the position of the wires in cable bundles,a method for predicting the probability distribution of crosstalk in cable bundles is proposed in this paper.Assuming that the cross-section is fixed,the cable bundle is divided into Nsegments.Then the distribution of electromagnetic parameters of a single segment is calculated,and the probability distribution of electromagnetic parameters of the whole cable bundle is obtained by convolution algorithm.Based on Monte Carlo method,the probability distribution of crosstalk is obtained according to the MTLs theory,which can help the designer to find potential crosstalk problems and to evaluate the reliability of system with influence.In order to acquire appropriate segement number N,an emprical formula according to the degree of twist is proposed.For a specific example,its simulation and experiment are both carried out.The experimental results show that the number of N should be commensurate with the degree of twist.What′s more,a higher twisting degree will bring the concentration of probability distribution,which means there is an averaging tendency.