基于神经网络的汽车发动机机械设备故障定位

Fault Location of Automobile Engine Mechanical Equipment Based on Neural Network

为提升发动机故障定位方法的定位性能,提出基于神经网络的汽车发动机机械设备故障定位。该方法首先依据汽车发动机具体结构,分析汽车发动机的各项故障类型,结合小波原理分析发动机振动信号,获取发动机故障特征向量值;再依据获取的特征向量,建立基于人工神经网络的汽车发动机故障定位模型;最后通过遗传算法更新模型参数,计算模型输出值,实现发动机的故障定位。结果表明,运用该方法对汽车发动机实施故障定位时,故障定位精度高、定位效果好以及定位性能优秀。

In order to improve the performance of engine fault location method,a neural network based fault location method for automobile engine mechanical equipment was proposed.The method firstly analyzes various fault types of the automobile engine according to the specific structure of the automobile engine,analyzes the engine vibration signal with the wavelet principle,and obtains the characteristic vector value of the engine fault;and then establishes the fault location of the automobile engine based on the artificial neural network according to the obtained characteristic vector.Finally,the model parameters are updated through the genetic algorithm,the model output value is calculated,and the fault location of the engine is realized.The experimental results show that when the method is used to locate the fault of the automobile engine,the fault location accuracy is high,the location effect is good,and the location performance is excellent.