摘要
针对传统的滚动轴承剩余使用寿命预测精度低、计算效率低等问题,提出了一种基于改进Informer深度学习模型结构的滚动轴承剩余使用寿命预测方法。为解决现有Informer模型中的self-attention结构存在内存占用高、计算复杂度高等问题,提出CSPA结构对输入数据进行处理,大幅度减少内存占用,提升计算效率的同时提高计算精度。因此,将CSPA替换原Informer模型中的self-attention结构,提出了基于CSPA-Informer的滚动轴承剩余寿命预测方法。输入数据分为两个通道进行特征提取和线性投影,并通过解码器快速生成预测序列。将CSPA-Informer与其他预测模型在公开数据集上的预测结果进行对比,其MAE、MSE和RMSE分别提升了21%、32%和17%以上,验证了该方法在滚动轴承剩余寿命预测方面的有效性。
Aiming at the problems of low accuracy and computational efficiency of traditional methods towards rolling bearing,a rolling bearing remaining useful life prediction method based on improved Informer deep learning model structure was proposed.In order to solve the problems of high memory occupation and high computational complexity of the self-attention structure in the existing Informer model,the CSPA structure was proposed to process the input data,which greatly reduces the memory occupation,improves the computational efficiency and accuracy.Therefore,CSPA was replaced by the self-attention structure in the original Informer model,and a method for predicting the remaining life of rolling bearings based on CSPA-Informer was proposed.The input data was divided into two channels for feature extraction and linear projection,and the prediction sequence was quickly generated by the decoder.Compared with the prediction results of other prediction models on the open data set,the MAE,MSE and RMSE of CSPA-Informer have increased by 21%,32%and 17%respectively,which verifies the effectiveness of this method in the prediction of the remaining life of rolling bearings.
作者
颜家威
易灿灿
黄涛
肖涵
YAN Jiawei;YI Cancan;HUANG Tao;XIAO Han(Key Laboratory of Metallurgical Equipment and Control,Ministry of Education,Wuhan University of Science and Technology,Wuhan 430080,China;Hubei Provincial Key Laboratory of Mechanical Transmission and Manufacturing Engineering,Wuhan University of Science and Technology,Wuhan 430080,China;Institute of Precision Manufacturing Research,Wuhan University of Science and Technology,Wuhan 430080,China)
出处
《组合机床与自动化加工技术》
北大核心
2023年第10期85-90,共6页
Modular Machine Tool & Automatic Manufacturing Technique
基金
国家自然科学基金项目(51805382)
湖北省重点研发计划项目(2021BAA194)。
