基于大数据聚类分析的爬壁机器人位姿定位控制系统设计

Design of pose positioning control system for wall-climbing robot based on big data cluster analysis

摘要:

为提高爬壁机器人运动位姿精准度与稳定性,解决现有位姿定位控制系统存在的控制效果不佳的问题,利用大数据聚类分析技术,通过软、硬件结构的设计,实现爬壁机器人位姿定位控制系统的优化。改装爬壁机器人位姿传感器、爬壁机器人驱动元件、爬壁机器人位姿定位控制器的内部连接结构及工作方式,扩大系统存储器的存储空间,通过系统电路的连接完成硬件系统的设计。根据爬壁机器人的组成结构和工作机理,建立机械结构与运动模型。在构建模型下,采集爬壁机器人实时运行数据,利用大数据聚类分析技术处理初始采集数据,判定爬壁机器人的当前位姿。规划爬壁机器人位姿及关节轨迹,结合当前爬壁机器人的运行数据,计算爬壁机器人位姿定位控制量,在控制器的约束下,实现爬壁机器人位姿定位控制功能。通过系统测试实验得出结论：通过设计位姿定位控制系统的应用,爬壁机器人样机的足端轨迹控制误差、关节角度控制误差和占空比控制误差均低于预设值,即设计系统具有良好的位姿定位控制效果。

Abstract:

In order to improve the accuracy and stability of the motion pose of the wall-climbing robot and solve the problem of poor control effect of the existing pose positioning control system, the big data clustering analysis technology is used to realize the wall-climbing through the design of software and hardware structure. Optimization of robot pose positioning control system. Modify the internal connection structure and working method of the wall-climbing robot pose sensor, the wall-climbing robot driving element, and the wall-climbing robot pose positioning controller, expand the storage space of the system memory, and complete the hardware system design through the connection of the system circuit. According to the composition structure and working mechanism of the wall-climbing robot, the mechanical structure and motion model are established. Under the construction of the model, the real-time operation data of the wall-climbing robot is collected, and the big data cluster analysis technology is used to process the initial collected data to determine the current pose of the wall-climbing robot. Plan the pose and joint trajectories of the wall-climbing robot, and combine the current running data of the wall-climbing robot to calculate the position and orientation control amount of the wall-climbing robot. Through the system test experiment, it is concluded that by designing the application of the pose positioning control system, the foot trajectory control error, joint angle control error and duty cycle control error of the wall-climbing robot prototype are all lower than the preset values, that is, the designed system has Good pose positioning control effect.