基于ANFIS的空气悬架执行器故障诊断研究  

Fault diagnosis of air suspension actuators based on ANFIS

作　　者：孙涛 张丽萍[1] 朱永博 祁升升 SUN Tao;ZHANG Liping;ZHU Yongbo;QI Shengsheng(College of Automotive and Traffic Engineering,Liaoning University of Technology,Jinzhou 121001,Liaoning,China)

机构地区：[1]辽宁工业大学汽车与交通工程学院,辽宁锦州121001

出　　处：《农业装备与车辆工程》2024年第11期37-43,共7页Agricultural Equipment & Vehicle Engineering

基　　金：辽宁省教育厅2021年度科学研究经费项目(面上项目)(LJKZ0620)。

摘　　要：悬架作为汽车底盘不可或缺的一部分,一旦发生故障会对汽车甚至驾驶员产生恶劣影响。为了找到悬架发生故障的位置,短时间内减轻对车辆结构的进一步损伤,并保障驾驶员及乘客的安全,针对电控空气悬架执行器故障,提出了一种自适应模糊神经网络系统(ANFIS)的故障诊断与隔离方法。设计ANFIS观测器组,通过传感器的输出值为真实数据,ANFIS观测器判断发生故障的位置进行诊断。利用AMESim软件搭建电控空气悬架系统物理模型与Simulink搭建PID/PWM车身高度控制器模型、故障模型、随机路面模型与故障诊断模型,最后通过AMESim-Simulink联合仿真进行验证。仿真结果得出不同故障下输出值与其对应的诊断结果的预期输出值差异较小,整个悬架系统的平均诊断误差为0.048,准确率为95%,验证了此方法的有效性与准确性。As an integral part of the automobile chassis,the suspension will have a bad impact on the vehicle and even the driver once it fails.In order to find the location where the suspension failure occurs,to mitigate further damage to the vehicle structure in a short time,and to protect the safety of the driver and passengers,an adaptive fuzzy neural network system(ANFIS)fault diagnosis and isolation method is proposed for the electronically controlled air suspension actuator failure.The ANFIS observer set was designed to diagnose by the output value of the sensor as real data and the ANFIS observer determines the location where the fault occurs.AMESim software was used to build the physical model of electronically controlled air suspension system and Simulink to build the PID∕PWM body height controller model,fault model,stochastic road model and fault diagnosis model,and finally verified by AMESim-Simulink joint simulation.The simulation results yielded different faults under the output value and its corresponding diagnostic results had a small difference in the expected output value,the average diagnostic error of the entire suspension system was 0.048,the accuracy rate was 95%,which verifies the effectiveness and accuracy of this method.

关 键 词：电控空气悬架 执行器故障 自适应模糊神经网络系统 故障诊断 

分 类 号：U463.33[机械工程—车辆工程] TP273[交通运输工程—载运工具运用工程]