摘要
针对液压信号复杂且难以诊断的难点,提出一种多尺度一维卷积神经网络与多传感器信息融合的深度神经网络模型(MS1D-CNN-MSIF)对液压泵与蓄能器进行故障诊断。在提出方法中,采用不同大小的卷积核对故障信号进行多尺度特征提取;然后使用多传感器信息融合策略将多个传感器的特征信号进行融合,最后使用Softmax进行分类识别。诊断蓄能器压力状态与液压泵泄漏状态的实验结果表明,与支持向量机、堆栈自编码、深度置信网络比较,提出模型具有更好的故障诊断性能,蓄能器识别精度可达99.50%,液压泵识别精度可达99.73%。
Because it is difficult to diagnosecomplex hydraulic signals,the multi-scale one-dimensional convolutional neural network and multi-sensor information fusion(MS1D-CNN-MSIF)deep neural network model is proposed to diagnose the fault of hydraulic pump and accumulator.Under the proposed method,the convolution check signals of different sizes are used to extract multi-scale features.The multi-sensor information fusion strategy is used to fuse the characteristic signals of multiple sensors,which are thenclassified and recognized with the Softmax software.The experimental results on diagnosing the accumulator pressure state and hydraulic pump leakage state show that,compared with the support vector machine,stack self coding and deep confidence network,the proposed method has a better fault diagnosis performance and that the method’s recognition accuracy of the accumulator and hydraulic pump reaches 99.50%and 99.73%respectively.
作者
陈书辉
章猛
刘辉
张超勇
CHEN Shuhui;ZHANG Meng;LIU Hui;ZHANG Chaoyong(State Key Laboratory of Digital Manufacturing Equipment and Technology,Huazhong University of Science and Technology,Wuhan 430074,China)
出处
《机械科学与技术》
CSCD
北大核心
2023年第5期715-723,共9页
Mechanical Science and Technology for Aerospace Engineering
基金
广东省重点领域研发计划项目(2019B090921001)。
