带滑移铰空间机器人运动规划的混合优化策略

HYBRID OPTIMIZATION STRATEGY FOR MOTION PLANNING OF SPACE ROBOT SYSTEM WITH PRISMATIC JOINT

  • 摘要: 研究了自由漂浮带滑移铰空间机器人非完整运动规划的最优控制问题,提出一种由高斯伪谱法求解可行解与直接打靶法求解最优解相结合的混合优化策略.首先,根据多体系统动力学理论建立空间机器人的动力学模型,给定系统的初始和目标位形,将空间机器人运动规划问题描述成博尔察(Bolza)型最优控制问题;然后,利用高斯伪谱法将最优控制问题离散为非线性规划问题,求解在较少勒让德-高斯(Legendre-Gauss,LG)点时状态变量和控制变量对应的可行解;最后,在LG点处离散控制变量,作为直接打靶法的初值,利用序列二次规划算法求解空间机器人系统的优化运动轨迹和最优控制输入.通过数值仿真,系统优化运动轨迹光滑平稳,最优控制输入也能很好地满足各种约束条件,仿真结果验证了该混合优化策略的鲁棒性和有效性.

     

    Abstract: This paper studies the optimal control of the nonholonomic motion planning for the free-floating space robot with prismatic joint. A hybrid optimization strategy based on the Gauss pseudospectral method (GPM) and the direct shooting method (DSM) is proposed. First, a dynamics model for the space robot is established based on the multibody dynamics theory. For a given initial and flnal conflgurations, the motion planning problem is described as an optimal control problem in the Bolza form. Then, the continuous optimal control problem is transformed into a nonlinear programming problem based on the GPM. With fewer Legendre-Gauss (LG) points, the feasible solution of the state variables and the control variables can be obtained. Finally, the control variables are discretized at the LG points and used as the initial values of the DSM, the optimal trajectory and the optimal control inputs are obtained by the sequential quadratic programming (SQP) algorithm. As shown by the numerical simulation, the optimal trajectory of the system is stable with no detours, and the control inputs match the various constraints under actual conditions. The results indicate that the hybrid optimization strategy is efiective with a good robustness.

     

/

返回文章
返回