基于自适应动态规划的半被动双足机器人控制  

Adaptive dynamic programming for semi-passive biped robot control

作　　者：李诚[1] 张奇志[1] 周亚丽[1] 

机构地区：[1]北京信息科技大学自动化学院,北京100192

出　　处：《北京信息科技大学学报（自然科学版）》2016年第4期17-21,26,共6页Journal of Beijing Information Science and Technology University

基　　金：国家自然科学基金资助项目(11172047)

摘　　要：针对半被动双足机器人系统中存在的强耦合和非线性导致控制困难的问题,采用基于自适应动态规划方法对机器人步态进行最优控制。首先建立机器人系统模型,依据控制目标,确定系统的效用函数,然后,在极限环存在的前提下,通过执行依赖启发式动态规划方法计算得到最优控制力矩。在半被动双足机器人支撑腿脚跟与地面发生碰撞之前,把脉冲力矩沿着支撑脚的方向施加给机器人。仿真结果表明:执行依赖启发式动态规划方法可以控制半被动双足机器人精确地跟踪双足机器人行走的期望步长,并且在系统的稳态误差、超调、响应速度和鲁棒性方面与迭代学习及模糊PID相比有更好的控制效果。To solve the problem of difficult control caused by strong coupling and nonlinearity in semi-passive biped robotic system,an adaptive dynamic programing approach is adopted to control the gait of semi-passive biped robot optimally. Firstly,according to the model of the robot system and control object,the utility function is determined. Then,under the condition of the existence of limit cycle herein,the optimal control torque is calculated by the action-dependent heuristic dynamic programming method. Finally,the biped robot is powered by the extra impulsive torque along the stance leg whose heel is going to strike the ground. The results of simulation show that the action-dependent heuristic dynamic programming approach for semi-passive biped robot can track the desired walking step length effectively and accurately,and compared with iterative learning approach and fuzzy PID,achieve better control in steady state deviation,overshoot,speed of response and robustness.

关 键 词：半被动 双足机器人 执行启发式动态规划 迭代学习 脉冲力矩 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]