Model Of Frequency-dependent Transmission Line Based On Vector Fitting And Genetic Algorithm

Most of the accidents in the power system are caused by operating over-voltage.In order to analyze the over-voltage process carefully, avoid the harm of over-voltage phenomenon, it is extremely important to establish a precise frequency-dependent transmission line electromagnetic transient model.This paper presents the modeling process of frequency-dependent transmission line in detail and indicates that the key is the accuracy of rational approximation of characteristic admittance and transmission function. Currently the result obtained by Vector Fitting is more accurate than those proposed method before. But its disadvantage is the great affection of the initial pole. Improper initial poles may cause inaccurate fitting result and reduce the rate of convergence seriously.In order to avoid that the frequency-dependent transmission line transient model established by vector fitting may be affected by initial iterative points easily,this paper presents an equivalent circuit modeling method of frequency-dependent transmission line based on the combination of vector fitting and genetic algorithm.Learning from the theory of biological evolution, we search the best initial poles with the global convergence ability of genetic algorithm. Genetic operations will be implemented in the group of the iterative points appeared during the vector fitting by constructing the fitness function,striving to obtain the global optimal approximate rational expression of characteristic admittance and delay function quickly. This method has been carried out on the simulation platform, it indicates that the frequency-dependent transmission line transient model established by this method has higher simulation accuracy. The search procedure based on the combination of vector fitting and genetic algorithm has a faster convergence speed than that by genetic algorithm merely. The practicability of the method has been confirmed.