基于惯性传感器的球形机器人位姿控制系统及实验研究  

Experimental research of spherical robot pose control system based on inertial sensor

作　　者：王琳玮 邵星灵 杨卫[1] WANG Linwei;SHAO Xingling;YANG Wei(State Key Laboratory of Electronic Testing Technology,North University of China,Taiyuan 030051,China)

机构地区：[1]中北大学电子测试技术国家重点实验室,山西太原030051

出　　处：《中国测试》2020年第3期123-127,共5页China Measurement & Test

基　　金：国家自然科学基金(61803347);山西省应用基础研究项目(201701D221123)。

摘　　要：针对复杂环境下对机器人多自由度灵活稳定运动的需求,设计一种基于神经网络自整定PID控制算法的球形机器人;该系统采用重心偏离的方式为机器人提供动力,利用对质量块的智能调速改变机器人的方向,结合陀螺仪加速度计等惯性传感器实现对机器人方向和位移的精确控制;机器人内部安装包括图像采集系统在内的多种传感器,利用设计的遥控程序,可以通过电脑或手机APP功能通过WiFi实现机器人远程控制。对机器人性能进行测试,实验结果表明该机器人运动性能良好,运动轨迹和姿态角误差在控制器接受范围内,在管道、密闭空间探查等方面具有良好应用前景。Aiming at complex environment more degrees of freedom robot flexible stable movement of the demand,design a self-tuning PID control algorithm based on neural network of the spherical robot.The system use a center of gravity deviation to provide power for robot,using the change of mass of intelligent control for the direction of the robot,in combination with gyroscope inertial sensors such as accelerometer to realize accurate control of the direction and displacement of the robot.Robot installed inside,including image acquisition system,a variety of sensors,Remote control procedures,can be remotely controlled by WiFi through the function of computer or mobile APP.The performance of the robot was tested,and the experimental results showed that the robot had good motion performance,the motion trajectory and attitude angle errors are within the range accepted by the controller and have a good application prospect in pipeline confined space exploration.

关 键 词：球形机器人 重心偏离 神经网络 PID算法 惯性传感器 

分 类 号：TH39[机械工程—机械制造及自动化]