摘要
针对园艺拖拉机线控液压大转向角系统的高度非线性和抗干扰性较差等特点,设计了一种基于神经网络的拖拉机线控液压大转向角系统复合滑膜控制策略。此策略使用径向基(RBF)神经网络自适应逼近系统不确定项并加以补偿,且采用饱和函数法抑制系统抖动,同时使用AMESim和MATLAB/Simulink对线控液压大转向角系统进行联合仿真实验。结果表明:设计的控制器相对于常规PID控制器能够较好地适应转向系统的非线性,瞬态响应时间由0.94s缩短为0.39s,跟随响应最大误差由14.8mm减小至5.5mm,抗干扰能力也更强。
The Hydraulic Steer-by-wire system has relatively serious nonlinearity,Poor anti-interference and so on,based on these issues,the strategy of neural network compound sliding mode control for hydraulic steer-by-wire system of tractor was proposed.Based on the advantages of the sliding mode control,the RBF neural network was used to adaptively approximate the system uncertainty,and the saturation function was used to suppress the system jitter.Based on the design of RBF neural network compound sliding mode controller,AMESim and MATLAB/Simulink co-simulation research was used.Compared with conventional PID controller,the strategy of neural network compound synovial control for hydraulic steer-by-wire system of tractor can adapt to the nonlinearity of steering system,the transient response is also better and it has good tracking accuracy and anti-jamming ability.
作者
王琪
孙宇峰
卞翔
张金铮
闫东东
陆凤祥
Wang Qi;Sun Yufeng;Bian Xiang;Zhang Jinzheng;Yan Dongdong;Lu Fengxiang(College of Mechanical Engineering,Jiangsu University of Science and Technology,Zhenjiang 212003,China;Jiangsu Yueda Huanghai Tractor Manufacturing Co.Ltd.,Yancheng 224000,China)
出处
《农机化研究》
北大核心
2022年第8期252-257,共6页
Journal of Agricultural Mechanization Research
基金
国家重点研发计划项目(2016YFD0700900)
江苏省自然科学基金项目(SBK2020042632)
教育部高等教育司产学合作协同育人项目(201802187002)。
关键词
园艺拖拉机
线控液压转向系统
神经网络滑膜
常规PID控制
联合仿真
horticultural tractor
hydraulic steer-by-wire system
neural network sliding mode
convention PID control
co-simulation
