基于神经网络的汽车EPS传感器故障诊断

The Fault Diagnosis of the Sensor Device of Automotive EPS System Based on Neural Network

汽车EPS传感器的故障诊断对提高汽车行驶的安全性具有重要的意义。提出了将BP神经网络引入汽车EPS传感器中的故障诊断思想。结合故障树分析(FTA)原理,给出基于EPS系统的传感器故障树模型结构图,以BP神经网络为训练网络,从传感器故障模型中选取故障征兆,进行网络训练,最终实现对EPS传感器的故障诊断。训练结果表明该方法在EPS传感器故障诊断中具有较高的诊断精度以及很好的稳定性。

The fault diagnosis of the sensor device of automotive EPS system is important for the automotive safety on the course. BP neural network was introduced into the fault diagnosis of the sensor device of EPS system. Combined with the fault tree, the tree structure chart of the sensor device faults of EPS system was presented. By selecting the fault symptom from the model of sensor device faults and using the BP neural network as training network to carry out network practice, the fault diagnosis of the sensor device of EPS system was realized. The result shows that the presented method has high precision and good robustness in the fault diagnosis of the sensor device of EPS system.