一种基于层叠分离卷积模块的机械设备剩余寿命预测方法,先获取不同工况下机械设备的原始振动信号,进行预处理,建立基于层叠分离卷积模块的机械设备剩余寿命预测模型;利用层叠分离卷积模块提取原始振动信号中与机械设备健康状况最相关的高层次典型特征;然后将高层次典型特征输入到全连接网络,得到机械设备的剩余寿命预测值;构建剩余寿命预测模型的均方误差目标函数,通过Adam优化算法迭代更新预测模型的待训练参数,获得最优的剩余寿命预测模型;最后输入预处理后的机械设备振动信号,完成机械设备得剩余寿命预测;本发明具有预测精度更高、稳定性更好、鲁棒性更强的优点。
A residual life prediction method of mechanical equipment based on cascade separation convolution module is proposed. Firstly, the original vibration signals of mechanical equipment under different working conditions are obtained and preprocessed, and the residual life prediction model of mechanical equipment based on cascade separation convolution module is established; The cascade separation convolution module is used to extract the high-level typical features most related to the health status of mechanical equipment in the original vibration signal; Then the high-level typical features are input into the full connection network to obtain the residual life prediction value of mechanical equipment; The mean square error objective function of the remaining life prediction model is constructed, and the parameters to be trained of the prediction model are iteratively updated by Adam optimization algorithm to obtain the optimal remaining life prediction model; Finally, input the vibration signal of mechanical equipment after pretreatment to complete the remaining life prediction of mechanical equipment; The invention has the advantages of higher prediction accuracy, better stability and stronger robustness.