基于LMBP算法的柴油机喷油器故障诊断

Fault Diagnosis of Diesel Injector Using Levenberg-Marquardt Back Propagation Algorithm

为了实现柴油机燃油供给系统的故障诊断,采用了一种新的快速收敛学习算法——Levenberg-marquardt BP(简称LMBP算法),建立了面向LMBP神经网络柴油机喷油器故障诊断模型。通过对柴油机喷油器典型故障的试验研究,结果表明该方法是可行的,辨识效果好。

In order to diagnose faults of a diesel injector,this paper proposed a learning algorithm—Levenberg-Marquardt back propagation (LMBP) algorithm that converges very fast. A fault diagnosis model was built by using the LMBP neural network. Experiments of the typical faults of diesel injector were conducted. The results show that the diagnosis method is feasible and effective.