基于弹性动力学模型与遗传算法的Delta机器人模糊PID控制

Fuzzy PID control of delta robot based on elastic dynamic model and genetic algorithm

为了提高Delta机器人控制的实时性与准确性,同时使控制过程简单方便、运行效率高,以Delta机器人为研究对象,基于空间有限元理论与拉格朗日方程,充分考虑主、从动臂间的运动及动力特性,建立了其弹性动力学模型。通过分析系统的结构特点,采用模糊比例—积分—微分(PID)控制器与传统PID控制器并联的形式对Delta机器人主动臂输入力矩进行控制。以动平台轨迹误差的平均值与系统前三阶固有频率平均值的倒数为目标函数,利用遗传算法优化得到模糊PID控制参数的取值范围,从而提高控制的实时性与准确性。通过ADAMS与MATLAB/simulink联合仿真和物理样机的现场实验,验证了控制策略的有效性与正确性。

To improve the real-time and accuracy of Delta robot＇s control,and to improve the control process and operating efficiency,Delta robot was taken as the study object,the elastic dynamic model for fully considering the characteristics of motion and dynamic of active arms and driven arms was built based on space finite element theory and Lagrange equation.Through analyzing the structural features of system,the active arms input torque of Delta robot was controlled by the fuzzy Proportional-Integral-Derivative （PID） controller and the traditional PID controller in parallel form.By taking the average values of moving platform＇s trajectory errors and the average frequency of first three order natural frequency as target functions,the parameters range of fuzzy PID control was obtained by genetic algorithm,so as to improve the real-time and accuracy of control.Combined simulation of ADAMS and MATLAB/simulink and the field test of physical prototype were used to verify the validity and correctness of the control strategy.