Fault Location Method Based On SFOA-BP Neural Network

UHVDC transmission has the advantages of large transmission capacity,long transmission distance,and low power consumption.Therefore,it has irreplaceable advantages in the long-distance transmission of electrical energy.Because UHVDC transmission lines are long and the terrain is complicated,when the system fails,whether it can quickly and accurately find and eliminate faults is related to the reliability of the power system,and researches the fast and accurate UHVDC transmission line fault location technology is very important.According to the mapping relationship between traveling wave spectrum energy and fault distance of transmission line,this paper proposes a fault location method based on SFOA-BP neural network.In view of the good ability of the BP neural network to fit the nonlinear function,the BP neural network is used to fit the non-linear relationship between the traveling wave spectrum energy and the fault distance.However,the BP neural network has the slow convergence speed and easily falls into the local minimum value..In view of the powerful global search capability of the Drosophila algorithm with adaptive connection structure,the SFOA algorithm is used to optimize the weights and thresholds of the BP neural network and then perform fault location.Based on the Yun-Guang 800 kV UHVDC transmission system model and related parameters,a simulation model of the Yunguang UHVDC transmission system was built based on PSCAD/EMTDC software.Assume that the short circuit fault occurs in the positive line,and in the case of different fault resistances and fault distances,the transient voltage signal after the failure of the transmission line is collected,and the voltage signal is decoupled by the phase mode to extract its line mode component,and the three-layer wavelet packet is used.Decomposition of the fault transient wave line mode component line,the wavelet energy spectrum in the extracted 3-7 bands is used as the fault feature quantity to train the SFOA-BP neural network,and the SFOA-BP neural network fault location model is obtained.Simulation results show that the fault location algorithm based on SFOA-BP neural network proposed in this paper has high accuracy,ranging error is kept within 0.1%,and the resistance to transient resistance is strong.