数字式气动桥式回路故障诊断  

Digital Pneumatic Bridge Loop Fault Diagnosis

作　　者：孙健 度红望 熊伟 SUN Jian;DU Hong-wang;XIONG Wei(School of Naval Architecture and Marine Engineering,Dalian Maritime University,Dalian,Liaoning 116026)

机构地区：[1]大连海事大学船舶与海洋工程学院,辽宁大连116026

出　　处：《液压与气动》2023年第6期156-162,共7页Chinese Hydraulics & Pneumatics

摘　　要：针对数字式气动桥式回路可靠性不高的问题,提出了基于支持向量机的气动系统故障诊断方法。通过分析气动桥式回路常见故障,得出9种故障类型和1种正常状态,使用AMESim建立了回路的仿真模型,基于多传感融合思想获取系统正常动作、工作压力低、气缸泄漏、开关阀不工作或者产生泄漏、减压阀不减压等正常/故障状态数据。针对系统整体工作数据集,在MATLAB中建立支持向量机故障模型进行桥式回路故障诊断实验与分析,结果表明该方法可准确分辨出系统的1种正常状态和9种故障状态。Aiming at the problem that the reliability of digital pneumatic bridge circuit is not high,this paper proposes a fault diagnosis method for pneumatic system based on support vector machine.By analyzing the common faults of pneumatic bridge circuit,9 fault types and 1 normal state are obtained,The simulation model of the circuit is established by using AMESim,and the normal/fault status data such as normal operation of the system,low working pressure,cylinder leakage,switching valve is not working or leaking,and pressure reducing valve is not decompressed is obtained based on the multi-sensor fusion idea.According to the overall working data set of the system,the fault model of the support vector machine is established in MATLAB for bridge loop fault diagnosis experiment and analysis,and the results show that the proposed method can accurately distinguish 1 normal state and 9 fault states of the system.

关 键 词：气动回路 故障诊断 支持向量机 多传感器融合 

分 类 号：TH137[机械工程—机械制造及自动化]