基于动态规划的机器人运动规划最优控制

发布时间：2018-10-12 21:06

【摘要】：针对轮式移动机器人运动规划的非完整性问题,采用自适应动态规划(Adaptive Dynamic Programming,ADP)方法求解其最优控制。通过建立轮式机器人的动力学模型,选取状态变量和控制变量,将其转化为对动态规划问题的求解。根据自适应动态规划网络结构,利用评价网络近似性能指标函数,执行网络逼近控制变量,并给出了适合该类问题的一种效用函数的具体表达式,实现了对轮式机器人非完整运动规划的最优控制。通过数值仿真实验,验证了ADP求解轮式机器人非完整运动规划最优控制问题的有效性。
[Abstract]:In order to solve the non-integrality problem of wheeled mobile robot motion planning, adaptive dynamic programming (Adaptive Dynamic Programming,ADP) method is used to solve the optimal control problem. The dynamic model of wheeled robot is established, and the state variable and control variable are selected to solve the dynamic programming problem. According to the network structure of adaptive dynamic programming, the approximate performance index function of the evaluation network is used to perform the network approximation control variable, and the concrete expression of a utility function suitable for this kind of problem is given. The optimal control of nonholonomic motion planning of wheeled robot is realized. The effectiveness of ADP in solving the optimal control problem of nonholonomic motion planning of wheeled robot is verified by numerical simulation.
【作者单位】：北京信息科技大学自动化学院;北京信息科技大学理学院;
【基金】：国家自然科学基金(11472058)
【分类号】：TP242

【相似文献】