基于高斯过程回归预测的舞台机器人避撞路径滚动时域规划  

Moving horizon planning for obstacle avoidance of mobile stage robots based on Gaussian process regression prediction

作　　者：陈伟轩 徐晨辉 何德峰[1] 余世明[1] CHEN Weixuan;XU Chenhui;HE Defeng;YU Shiming(College of Information Engineering,Zhejiang University of Technology,Hangzhou 310023)

机构地区：[1]浙江工业大学信息工程学院,杭州310023

出　　处：《高技术通讯》2023年第5期528-536,共9页Chinese High Technology Letters

基　　金：国家重点研发计划(2018YFB1403704)资助项目。

摘　　要：针对移动舞台机器人的避撞运动规划问题,提出基于高斯过程回归和模型预测控制(MPC)的方法。在获取舞台环境中运动体位置信息后,使用高斯过程回归方法对运动体位置进行预测,再利用运动体的未来位置信息,将控制时域内存在的碰撞风险作为硬性约束,在线优化求解移动机器人未来一段时刻的局部最优速度轨迹。进一步驱动移动舞台机器人跟踪规划速度,实现对移动舞台机器人运动轨迹的实时滚动规划与跟踪。最后在两类舞台场景下,仿真验证了本文方法的有效性。Considering the problem of collision avoidance motion planning for mobile stage robots,a moving horizon planning algorithm for collision avoidance based on Gaussian process regression and model predictive control is proposed.After obtaining the position of moving bodies in the stage environment,Gaussian process regression method is used to predict the position of other moving bodies.Then,using the predicted location information of other moving bodies and taking the collision risk in control time as rigid constraint,the local optimal speed trajectory of a mobile robot in a fixed horizon of future is solved by online optimization.Further,the mobile stage robot is driven to track the planning speed,the mobile stage robot and realizes real-time moving trajectory planning and tracking.Finally,the effectiveness of the proposed algorithm is verified by simulation in two scenes.

关 键 词：移动舞台机器人 高斯过程回归 模型预测控制(MPC) 运动规划 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置] O212.1[自动化与计算机技术—控制科学与工程]