基于滑模变结构的轮式机器人运动误差控制器设计

Design of Motion Error Controller for Wheeled Robot Based on Sliding Mode Variable Structure

面对当前使用的基于黎卡提微分方程、综合位姿误差控制优化仿真方法受到不确定性因素干扰,导致控制误差较大的问题,提出了基于滑模变结构的轮式机器人运动误差控制器设计;使用双框架陀螺仪,为控制器提供电力;采用FB900C/E角位变送器,将交流信号转换为角位移输出;使用TMS320F2812的DSP控制器,负责控制整个控制器的数据传输以及电平转换;充分考虑相变量输入的n阶线形数据,设计滑模变结构控制律,计算滑模变结构控制滑动面;在滑模变结构模态控制阶段,构建滑模变结构模态控制阶段的运动方程矩阵;对滑模切换面强制状态点运动,通过控制目标实现稳定控制;对DSP程序控制流程进行设计,将基于滑模变结构的误差控制结果传递到电位器上,并将运行数据发送到主机,由此完成轮式机器人运动误差控制;由实验结果可知,该控制器通过自适应调整参数后,滑模抖振得到明显消除,且最大控制误差为0.01 rad,具有精准控制效果。

The motion error controller of wheeled robot based on sliding mode variable structure was proposed to solve the problem of large control errors caused by the disturbance of uncertain factors in the current optimization simulation method of position and pose error control based on Riccati differential equation.A dual-frame gyroscope is used to provide power to the controller.FB900C/E Angle transmitter is used to convert AC signal into angular displacement output.The DSP controller of TMS320F2812 is used to control the data transmission and level conversion of the whole controller.Considering the n-order linear data of phase variable input,the control law of variable structure control of synovial film was designed and the sliding surface of variable structure control was calculated.In the mode control stage of the synovial variable structure,the motion equation matrix of the mode control stage was constructed.The state point is forced to move on the synovial switching surface,and the stability control is achieved by controlling the target.The DSP program control flow is designed,and the error control result based on sliding mode variable structure is transmitted to the potentiometer,and the operation data is sent to the host computer,so as to complete the motion error control of the wheeled robot.The experimental results show that the sliding mode chattering is obviously eliminated after the controller adaptively adjusts the parameters,and the maximum control error is 0.01 rad,which has an accurate control effect.