Research On Time Domain Calculation Method For High-speed Railway Catenary Lightning Overvoltage

As the locomotive power transmission line,high speed railway catenary plays an important role in the operation of high-speed railway.The lightning protection level of the high speed railway in our country is low,and there are many lightning accidents,so it is of great significance to calculate and research the lightning overvoltage of high speed railway catenary.The high speed railway catenary is composed of AF line,T line and PW line.In this paper,a new type of overhead line model(A-Line model)is used to calculate the catenary lightning overvoltage.However,A-Line model is influenced by various factors,and it has certain limitations.The applicability of this method is not analyzed in the paper which puts forward the A-Line model.In this paper,the main factors affecting the method are analyzed and verified,and the applicability of the method is obtained,which provides a theoretical basis for the application of the method.The integrated grounding system of high speed railway should be taken into account when calculating the catenary lightning overvoltage.The bridge pier is a part of integrated grounding system of high speed railway.It has a large number of nodes and conductors in the course of calculation.How to calculate it quickly is a problem to be solved.In this paper,an improved node tearing method-node tearing method based on Norton circuit is proved in order to calculate the grounding grid.The bridge pier lightning overvoltage is calculated by the improved node tearing method and overlapping partition waveform relaxation method.The grounding grid capacitance network and inductance network are separated.The capacitance network is calculated by the improved node tearing method,and the inductance network is calculated by the overlapping network waveform relaxation method.So the lightning overvoltage of grounding grid can be calculated quickly and accurately.In order to speed up the calculation,this paper uses the GPU parallel acceleration technology.The parallel computation of matrix vector multiplication and LU decomposition for solving linear equations is implemented and optimized.The superiority of GPU parallel computing with respect to CPU serial computing is verified by the calculation of specific examples.