BP神经网络的汽车故障诊断系统

Auto's Fault Diagnosis System Based on BP Neural Network

汽车发动机故障呈现多部位、多现象、非线性等特点,因此诊断汽车发动机故障较为困难。经过对汽车发动机故障的诊断以及BP神经网络理论的研究。将BP神经网络的联想、推测、记忆、学习等优点和误差反向传播算法应用于汽车故障诊断。分析了系统网络结构和性能,利用实际测试的故障样本训练网络并进行测试。结论证明,BP神经网络应用于汽车故障诊断,效果良好,具有较高的诊断效率和准确度。

It is difficult to diagnose the faults of automobiles because the issue refers to multi-position, multi-phenomenon, and non-liner characteristic. Through researching the fault diagnosis of engines and BP NN theory, the BP network and its advantages of association, inference, memory, and learning, as well as BP arithmetic are used in fault diagnosis of automobiles. The structure and performance of network are analyzed, by using fault samples from practical tests the network is trained and tested. The conclusion is that the method offers excellent results of diagnosis and the efficiency and accuracy have been enhanced.