足式机器人腿部关节改进单神经网络PID控制算法研究  

Research on Improved Single Neural Network PID Control Algorithm for Leg Joints of Footed Robot

作　　者：马程 蒋刚 郝兴安[2] 蒲虹云 陈清平 黄建军 徐文刚 黄璜 MA Cheng;JIANG Gang;HAO Xingan;PU Hongyun;CHEN Qingping;HUANG Jianjun;XU Wengang;HUANG Huang(School of Nuclear Technology and Automation Engineering,Chengdu University of Technology,Chengdu Sichuan 610059,China;School of Mechanical and Electrical Engineering,Chengdu University of Technology,Chengdu Sichuan 610059,China;Chengdu Lingchuan Special Industry Co.,Ltd.,Chengdu Sichuan 610110,China)

机构地区：[1]成都理工大学核技术与自动化工程学院,四川成都610059 [2]成都理工大学机电工程学院,四川成都610059 [3]成都陵川特种工业有限责任公司,四川成都610110

出　　处：《机床与液压》2024年第3期60-66,共7页Machine Tool & Hydraulics

基　　金：四川省重大科技专项(2020ZDZX0019);四川省科技计划重点研发项目(2021YFG0076,2021YFG0075)。

摘　　要：为了满足液压足式机器人在复杂环境中实现精确、快速的腿部关节控制需求,把单神经网络PID能够实时调节参数的优点运用到足式机器人液压机械腿关节的控制中,在单神经网络PID的基础上增加机械腿关节的位置和速度控制算法,形成改进单神经网络PID,实现了对神经元比例参数自调整、PID参数的自整定,能够较好地适应内、外参数的变化,增强了腿部关节的快速性、精确性。在Simulink中进行建模仿真以及在设计的以STM32为中央处理芯片的控制平台上进行实验测试,结果表明:改进单神经网络PID在足式液压机器人的腿部关节控制中具有响应速度快、超调量小、控制精度高、鲁棒性强等优点。In order to meet the requirements of accurate and fast leg joint control of hydraulic footed robots in complex environments,the advantage of single neural network PID of being able to adjust parameters in real time was applied to the control of hydraulic legs of footed robots.On the basis of single neural network PID,the position and speed control algorithms of mechanical leg joints were added,and an improved single neural network PID was formed,then the self-adjustment of neuron proportional parameters and the self-tuning of PID parameters were realized,which could better adapt to the changes of internal and external parameters.So the speed and precision of the leg joints were enhanced.The experimental results show that the improved single neural network PID has a fast response speed in the leg joint control of the foot hydraulic robot.It has the advantages of small overshoot,high control precision and strong robustness.

关 键 词：电液伺服控制 足式机器人 改进单神经网络PID 参数自整定 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置] TH137[自动化与计算机技术—控制科学与工程]