基于BP神经网络的电动助力转向器在线故障诊断研究

应用人工神经网络和推理法则,提出了一套针对电动助力转向器(EPS)、转矩转角传感器(TPS)和电子控制单元(ECU)的在线故障诊断策略。以传感器采样值作为神经网络的输入,TPS及ECU故障代码作为输出,对EPS关键部件进行在线故障诊断。通过EPS试验台,获取相应的数据样本,利用Matlab神经网络工具箱,对该策略进行了仿真分析,结果验证了故障诊断策略的有效性。

On-line Fault Diagnosis of EPS System Using BP Neural Network

Based on artificial neural network(ANN) and inference rules,a set of strategies for on-line fault diagnosis for the torque/position sensor(TPS) and electronic control unit(ECU) of the electric power steering(EPS) system are put forward.The on-line fault diagnosis on key components of EPS is conducted with sensor data as input of ANN and the fault codes of TPS and ECU as output.Based on sampling data obtained from test rig and utilizing neural network toolbox in Matlab,a simulation on the strategies is performed and the results verify the effectiveness of the strategies.