Study On AT Network Models

The development of high-speed railway determins the widely application of auto-transformer power supply. It is suitable for large current high-speed and heavy load railway because of its good ability of power supply and anti interference. While AT power supply has its unique superiority, its network structure and the electric calculation is very complex. To effectively simplify the calculation by choosing reasonable traction network model, it is a great significance of studying AT traction network model.This paper analysed the formulae of AT traction network rail potential, short-circuit impedance and its voltage loss. It also analysed the calculation methods of impedance matrix and admittance matrix of AT traction network wires, calculated and compared the results of the traction network wires parameter by using the Carsom formula and plural depth formula to write M files. Builded the multiconductor model and the equivalent circuit model without mutual impedance according to the calculation of traction network parameters.The equivalent circuit model with guard line was builed by merging the rail and the guard line.Three kinds of AT traction network simulation models were builded by Matlab/Simulink in this paper, and the rail potential, short-circuit impedance and voltage loss were simulated. Considering the impaction of CPW, earth connection, rail leakage conductance, rail cross connection in traction network CPW, simulation models were simulated by changing the settings of the above factors. Compared and analysed the the above factors’influence to the calculation results of the multiconductoer model and the equivalent circuit model. Three different models were simulated by setting the CPW interval as 1km. At last, verified the accuracy of the equivalent circuit model with guard line by comparing the different simulation results.Through the analysis of the simulation results, the variation trend of AT traction network rail potential, short-circuit impedance, voltage loss were concluded. The influence of simulation error to different models were analysed when the above factors were changing.