摘要
在变电站巡检机器人作业中,有效的避障策略不仅能防止机器人与已知障碍物碰撞,还能确保实际行进路线紧密贴合预设巡检轨迹。基于此,提出基于数字孪生SLAM感知的变电站巡检机器人避障控制算法。定义数字孪生模型,并在其中实施SLAM感知转换,实现机器人巡检路径的建图。在此基础上,预测可达巡检区域,再结合障碍物视场模拟条件,求解避障算法奖励函数,实现数字孪生SLAM感知下变电站巡检机器人避障控制算法的设计。实验结果表明,所提方法在实施避障控制时,机器人实时行进距离始终小于障碍物安全阈值且最大差距为3 dm,避障成功率更高;同时轨迹节点数量与预设值相差较小,偏离数量小于2,展现出更优的避障稳定性和控制精度。
In the operation of substation inspection robots,effective obstacle avoidance strategies can not only prevent collisions with known obstacles,but also ensure that the actual travel route closely follows the preset inspection trajectory.Based on this background,a substation inspection robot obstacle avoidance control algorithm based on digital twin SLAM perception is proposed.Define a digital twin model and implement SLAM perception transformation to complete the mapping of robot inspection paths.On this basis,the reachable inspection area is predicted,and combined with the obstacle field of view simulation conditions,the obstacle avoidance algorithm reward function is solved to achieve the design of obstacle avoidance control algorithm for substation inspection robot under digital twin SLAM perception.The experimental results show that when implementing obstacle avoidance control,the real-time walking distance of the robot is always less than the obstacle safety threshold,and the maximum difference is 3dm,resulting in a higher success rate of obstacle avoidance.At the same time,the number of trajectory nodes is relatively small compared to the preset value,with a deviation of less than 2,demonstrating better obstacle avoidance stability and control accuracy.
作者
许宏洋
丁再贤
周敏
段一平
乔海春
XU Hongyang;DING Zaixian;ZHOU Min;DUAN Yiping;QIAO Haichun(Haidong Power Supply Company of State Grid Qinghai Provincial Electric Power Company,Haidong 810600,China)
出处
《电子设计工程》
2025年第16期107-110,119,共5页
Electronic Design Engineering
基金
国网青海海东供电公司开发项目(742802220001)。
