基于单神经元PID调节的弹跳机器人落地姿态控制

Based on Single-Neuron-PID Governor for Landing Pose Control of Jump-robot

为解决弹跳机器人在落地前的姿态控制问题,将单神经元PID调节技术运用到机器人飞行姿态控制中,使机器人底面在落地前能够与地面调节至平行关系,以利于机器人稳定落地。根据控制要求,在对机器人"弹跳—落地"过程以及飞轮调节系统数学模型分析的基础之上,讨论了机器人空间姿态角变化与飞轮旋转角之间的关系,建立了基于机器人姿态角反馈的飞轮旋转输出力矩的单神经元PID调节控制方案,最后给出了在这种控制方案下基于Adams与MATLAB的联合仿真结果。该仿真结果表明,在落地前,通过飞轮输出力矩的变化,机器人底面能够与地面逐渐调节至平行关系,从而稳定落地。同时,仿真结果也为弹跳机器人在空间任意姿态角时的精确控制的进一步研究提供了理论基础和研究依据。

In order to solve the landing pose control problem of jumping robot before landing, and in favor of the robot＇s landing, the single-neuron-propotion intergration differential （PID） control technology is applied to the flight pose control of the robot, to make the robot floor bottom regulated to the parallel relationship with the ground, for its better stable landing. As the control task required, on the basis of analysis about the process of robot bouncing-landing and the mathematical model of flywheel adjusting control system, the relationship between the angle change in space pose of robot and the flywheel rotation angle was discussed, and a control scheme was also established which used the single-neuron-PID control strategy adjusts the performance of the torque of flywheel based on the feedback of the posture-angle of the robot. At last, a co-simulation used this control scheme based on Adams and Matlab was carried out. The simulation result indicates that the robot floor bottom has been regulated gradually to be parallel with the ground by changing the performance of the torque of flywheel before landing, and a smooth land has been gotten. At the same time, it also provides the the- oretical foundations and basis for the further study of the jump-robot＇s precision-control in space at any posture angle.