摘要
设计了一种模糊神经网络推理系统诊断发动机故障的方法,该方法应用遗传算法对系统进行优化并对BP网络算法进行改进。这种方法能够优化模糊系统的参数和结构,并且能删除无用的模糊规则。最后以发动机点火系统的次级电压为例,说明这种方法的应用过程。结果证明这种推理系统具有收敛速度快、泛化能力好、推广性强的特点。
Fuzzy inference neural network method was designed in engine fault diagnosis, and GA （genetic algorithm）was also adopted to optimize the system and improve BP network. The system was the right combination of fuzzy system and neural network. Parameters and structures of the fuzzy system were optimized and redundancies of fuzzy inference were reduced. At last, the application course was illustrated by analyzing secondary voltage of ignition system. The rapid and overall convergence, advanced comprehensibility and good generalization of the system were proved by the diagnosis conclusion, and the accuracy and the efficiency of fault diagnosis system were improved by the method.
出处
《农业机械学报》
EI
CAS
CSCD
北大核心
2005年第8期122-124,共3页
Transactions of the Chinese Society for Agricultural Machinery
关键词
模糊推理
神经网络
遗传算法
故障诊断
Fuzzy inference, Neural network, Genetic algorithm, Fault diagnosis
