摘要
提出利用GA-BP神经网络的系统对变压器的故障诊断进行优化。利用GA遗传算法优化BP的初始权值,得到GA-BP神经网络。同时使用L-M算法训练GA-BP,使其可精确识别故障变压器内部的气体含量变化,并针对变压器故障诊断过程进行高效处理。GA-BP神经网络具备模糊算法,具有计算快速和判断准确等优点,可在很多的领域内保障电气安全,因而其具有良好的发展前景。
The idea will be proposed of the optimization means for the diagnosis of transformer fault with GA-BP neural network system. The GA-BP neural network will be gotten if GA heredity method optimizes BP＇s initial power parameter. Simultaneously, the GA-BP nerve net can be trained by L-M method so that it can accurately identify the change of air content situation and high effectively deal with the procedure of diagnosis of transformer fault. With the fuzzy method, GA-BP has advantages such as quick computation, accurate judgment etc. It may ensure the electric safe case in many areas,so it＇s provided with nicer prospect.
出处
《煤矿机械》
2015年第7期318-320,共3页
Coal Mine Machinery
