Abstract: On the basis of analyzing feature of middle and low speed linear induction motor for maglev train, the simulated model of regenerate breaking dynamic network for middle and low speed maglev train was studied and dynamic net simulation model was established and the massive nonlinear equations were calculated thus g knowing the variation of grid voltage transformation during regenerate breaking. In order to maintain matrix positive for object function Hesse, improved Broyden rank 1 algorithm was adopted and Beijing maglev S1 line project was used as calculating object and the grid voltage variation rate on occasion of stable motion was calculated on. The results show that Broyden rank 1 algorithm requires low initial value of iterative computation with comparison of traditional Newton-Raphson method. The convergence rate is improved effectively and the divergence of simulated calculation results is restrained.

Key words: maglev train, traction power supply, regenerative braking, Broyden rank 1 algorithm