Research On Synchronous Control Technology Of Rail Transit Circuit Breakers

The rail transit locomotive main circuit breaker(hereinafter referred to as the rail transit circuit breaker)is a device that connects the main circuit of the locomotive and the locomotive traction network.Its intelligent level affects the safe and reliable operation of rail transit vehicles.Synchronous control technology as one of the directions of intelligent development of circuit breakers,its research is of great significance.This thesis uses rail transit circuit breakers as the carrier to study its synchronization control technology.First,on the basis of analyzing the basic working principle of the electromagnetic operating mechanism of rail transit circuit breaker,combining its electromagnetic operating mechanism control circuit and special working environment,analyze the factors affecting the opening time of the electromagnetic mechanism of rail transit circuit breaker,and research the principle and realization method of rail transit circuit breaker synchronization control technology.Secondly,in view of the large dispersion of rail transit circuit breaker opening time,machine learning methods are used to establish a neural network-based circuit breaker opening time prediction model,and the model is optimized by using Bayesian regularization training algorithm,to improve the convergence speed and prediction accuracy of the model;in view of the prediction error of the opening time caused by the aging of the spring mechanism and the contact wear,the law of the opening time change is analyzed,and an adaptive compensation scheme based on historical data weighted filtering is proposed.Finally,in the current zero prediction method,based on the mathematical model of the short-circuit current,the recursive least square method is used to analyze the transient shortcircuit current containing the DC component,the short-circuit current zero prediction model is established,and analyze the impact of sampling frequency,analog-to-digital conversion bits,noise,harmonics,frequency offset and other factors on the accuracy of short-circuit current zero prediction;aiming at the problem of excessive prediction error caused by frequency offset during asynchronous sampling,the reason for the error formation is analyzed,and a method for calculating the actual frequency of the power grid based on the principle of maximum correlation coefficient is proposed,which modifies the short-circuit current prediction model to eliminate the model error caused by frequency offset;design short-circuit current and grid frequency offset experiments to verify the correctness and effectiveness of the short-circuit current zero point prediction method and improved method proposed in this thesis.