基于BP神经网络的独立节流口阀流量-压力复合控制方法  

Compound Control of Flow and Pressure Method of Independent Metering Valve Based on BP Neural Network

作　　者：魏列江[1] 赵保才 孙辉 刘亚斌 安贞嬛 强彦[1] WEI Lie-jiang;ZHAO Bao-cai;SUN Hui;LIU Ya-bin;AN Zhen-huan;QIANG Yan(Energy and Power Engineering College,Lanzhou University of Technology,Lanzhou,Gansu 730050;Jiangsu Advanced Construction Machinery Innovation Center Ltd.,Xuzhou,Jiangsu 221004)

机构地区：[1]兰州理工大学能源与动力工程学院,甘肃兰州730050 [2]江苏汇智高端工程机械创新中心有限公司,江苏徐州221004

出　　处：《液压与气动》2023年第8期58-65,共8页Chinese Hydraulics & Pneumatics

基　　金：国家重点研发计划(2020YFB2009800)。

摘　　要：相较传统比例阀,独立节流口阀因阀口独立控制,更易实现流量-压力复合控制功能。但由于流量、压力控制过程为多级闭环控制,仅采用传统PID控制算法难以满足流量-压力复合控制的要求。针对某型双阀芯独立节流口阀,提出了基于BP神经网络PID的流量-压力复合控制策略,利用BP神经网络算法可以逼近任意非线性函数的特性,实时调整PID控制器的参数值,并进行了理论分析和实验研究。结果表明:采用BP神经网络PID控制算法在流量-压力复合控制时,流量、压力控制精度分别为1.8%和2.1%,可以满足大多数流量-压力复合控制场合的需求。Compared with traditional proportional valves,independent metering valves are easier to achieve compound control of flow and pressure because of independent control of valve ports.However,since the flow and pressure control processes are multi-stage closed-loop control,it is difficult to meet the requirements of compound control of flow and pressure when only the traditional PID control algorithm is adopted.For a dual-spool independent metering valve,the strategy of compound control of flow and pressure based on BP neural network PID are proposed,utilizing the characteristics that BP-NN algorithm can approximate any nonlinear function to adjust the parameters of PID controller in real time.Theoretical analysis and experimental studies were carried out.The results show that when BP-NN PID control algorithm is used for flow-pressure compound control,the precision of flow and pressure control is 1.8% and 2.1% respectively,which can meet the requirements of most flow-pressure compound control occasions.

关 键 词：双阀芯独立节流口阀 BP神经网络PID 流量-压力复合控制 

分 类 号：TH137[机械工程—机械制造及自动化]