Achieving rapid and accurate fault diagnosis of automation equipment is crucial to ensuring stable operation of smart substations featuring high autonomy and controllability.In view of regular types of faults that may be encountered by smart substation automation equipment and their electrical characteristics,the present work analyzed the feasibility of utilizing structural advantages of the attention mechanism-based gated recurrent unit(AM-GRU)in fault diagnosis of au-tomation equipment,and made a preliminary attemtp to establish an AM-GRU-based automatic fault diagnosis model and algorithm.The proposed algorithm was indicated by case verification,compared with three selected conventional algo-rithms,more superior in fault diagnosis accuracy and convergence speed.
维普
万方数据
