基于神经网络的内燃机活塞-缸套磨损的故障诊断

Fault Diagnosis of Internal Combustion Engine Pistonliner Wear Based on Neural Network Theory

分析了BP算法。在MATLAB环境下以改进的BP网络为识别模型对内燃机活塞-缸套磨损的几种故障进行分类训练,并应用待识别的故障样本识别仿真。结果表明,该方法在活塞-缸套磨损诊断中是行之有效的。

This paper analyzes Back Propagation training algorithm. Based on the improved Back Propagation neural network, some fault samples of internal combustion engine piston-liner wear condition trained in MAT-AB environment, and also the neural network model were applied to identify the samples for identification. The result indicates that the method is effective in the fault diagnosis of piston-liner wear condition.