基于神经网络趋势预测的装甲车辆柴油机寿命预测

Life Prediction for Armored Vehicle Diesel Engine Based on Neural Network Trend Model

针对装甲车辆柴油机现有定期维修方式存在的"维修过剩"和"维修不足"的问题,提出了装甲车辆柴油机神经网络寿命预测步骤,运用主成分分析方法将多个柴油机状态参数简化为2个综合参数,利用插值法得到柴油机各个时刻的综合参数数据,并作为神经网络的训练数据和测试数据,建立了装甲车辆柴油机寿命预测模型。结果表明:该模型预测精度较高,具有一定的应用和推广价值,为实现装甲车辆柴油机状态维修提供了技术支撑。

The process of life prediction for armored vehicle diesel engine based on nerual network is pres-ented to solve the problems of “excessive maintenance”and “inadequate maintenance”in Time Based Maintenance (TBM).The Principal Component Analysis (PCA)method is applied to turn several diesel engine condition parameters into two integrated parameters.Then the interpolation method is used to get the integrated parameters data of various times as training and validation data of neural network.Finally, a reliable neural network trend model of life prediction is founded for armored vehicle diesel engine.The application shows that the model is of high accuracy,which has applying and promotion value.And it pro-vides technical support to realize Condition Based Maintenance(CBM)of armored vehicle engine.