>> 自然科学版期刊 >> 2019年06期 >> 正文
电液锥阀电导率自适应控制研究
供稿: 陈军章;胡万强;卫世乾 时间: 2019-10-23 次数:

作者:陈军章胡万强卫世乾

作者单位:许昌市耕新信息科学研究院许昌职业技术学院信息工程学院许昌学院电气(机电)工程学院许昌学院学报编辑部

摘要:针对电液锥阀电导率易受外界因素影响而难于控制的问题,提出一种基于节点感知器网络的自适应智能控制器。该控制器借助节点感知器网络自适应学习电液锥阀逆输入状态映射,逼近被控对象目标函数,迫使其追踪流量系数特性和期望系统压力。另外,给该控制器加入一故障检测算法,该算法实时检测被控目标状态参数,与预先设定的性能界限偏差值相比较,进而迫使被控目标转移到对应状态模块,并提取故障特征,给出诊断结果。仿真结果表明,将节点感知器网络和故障检测算法相结合,能够实现对电液锥阀期望电导率和系统工作压力的良好追踪,平均追踪误差分别为-0.2×10-4m3/(s·Pa)和-0.004 MPa,优于其他控制算法。该控制器的提出为液压阀模型识别及快速准确故障诊断提供了参考依据。

基金:河南省科技攻关项目(182102210506);河南省高等学校重点科研项目(19B460011);

关键词:电液锥阀;节点感知器网络;自适应学习控制;故障检测算法;

DOI:10.16186/j.cnki.1673-9787.2019.6.17

分类号:TH137;TP273.2

Research on adaptive control for electro-hydraulic poppet valve conductivity

CHEN JunzhangHU WanqiangWEI Shiqian

Xuchang City Farming Institute of Information ScienceSchool of Information Engineering, Xuchang Vocational Technical CollegeSchool of Electrical Engineering & Mechano and Electronic Engineering, Xuchang UniversityEditorial Department of Journal, Xuchang University

Abstract:In order to solve the problem that the conductivity of electro-hydraulic poppet valve (EHPV) is easily affected by external factors and difficult to be controlled, an adaptive intelligent controller was proposed based on the nodal link perception link (NLPL) .The controller adaptively learned the inverse input state mapping of the EHPV by means of the NLPN, and approximated to the object function and forced it to track the flow coefficient characteristics and the expected system pressure.In addition, a fault detection algorithm was added to the controller to detect the state parameters of the controlled target, and the parameters of the algorithm were compared with the pre-established performance bound deviation.The controlled object was forced to transfer to the corresponding state module to extract its fault features and obtain diagnostic results.The simulation results showed that the adaptive intelligent controller could achieve a good tracking for the expected NLPN conductivity and the system operating pressure, and the average tracking error was-0.2×10-4m3/ (s·Pa) and-0.004 MPa, respectively.The results were superior to other control algorithms.The proposed controller could establish foundations for the model identification and the rapid and accurate fault diagnosis of hydraulic valves.

最近更新