论文部分内容阅读
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 system using GA-BP neural network is proposed to optimize the fault diagnosis of transformers. GA genetic algorithm to optimize the initial weight of BP, get GA-BP neural network. At the same time, L-M algorithm is used to train GA-BP, which can accurately identify the gas content changes inside the fault transformer, and treat the transformer fault diagnosis process efficiently. GA-BP neural network with fuzzy algorithm, with the advantages of rapid calculation and accurate judgment, can be in many areas to ensure electrical safety, so it has good prospects for development.