Human-following And Obstacle Avoidance Control Method For Wheeled Mobile Robot

Due to the aging of the world population and the rise of labor cost,human-robot integration system has been paid an increasing number of attention and has a good development prospect in many fields.In the human-robot shared space,effective human-following and obstacle-avoiding is an important prerequisite for human-robot integration.Therefore,the human-following control of wheeled mobile robot is deeply explored by taking the following,obstacle avoidance and cooperation between human and robot as the research direction.Firstly,the kinematics and dynamics model of the wheeled mobile robot is established to accurately describe the motion characteristics of the robot.The human-following is abstracted into the pilot-follow form,a range-angle human-following kinematics model is established.The motion coupling relationship between human and robot is expounded,which is the basis of human-following and obstacle avoidance control.In addition,the design requirements of human-following system are given,which makes a reasonable explanation for applying human behavior law to robot control.Secondly,according to the movement characteristics of people following each other,the impedance control method is applied to the human-following control to ensure the comfort of the human and the reliability of the robot.Combined with the impedance control of virtual interaction force,the robot can simulate the dynamic behavior of human beings and improve the comfort of human in the human-following system.In order to solve the problem of nonholonomic constraint of wheeled mobile robot and ensure the convergence of posture errors of the robot,a double closed-loop control structure is used to control kinematics and dynamics of the robot.On this basis,an obstacle avoidance component based on behavioral dynamics of human being is integrated to extend the impedance controller for collision avoidance in multi-obstacle environment.Thirdly,the system model based on distance and angle is applied to human-following control,the following robot can simulate the dynamic behavior of human when avoiding pedestrians by combining the interference degree and risk degree.The double closed-loop control structure is still used for the human-following dynamic control of the robot,in which LQR method is used in outer loop to control the following distance and angle.Heuristic rules is developed according to the behavior characteristics of people avoiding pedestrians to design the intervention degree and risk degree for real-time planning of the following angle of robot,so as to effectively complete pedestrian avoidance and improve the social acceptability of human-following.Finally,considering that the human-following system is faced with the execution of some practical tasks,in which the human-robot needs to maintain a precise position relationship.The coupling error can be used to combine the posture error of the robot with the position error of the human,and then the posture controller can be designed based on the coupling error to realize the cooperative control of the human-robot following system under specific working conditions.In this paper,human behavior and control methods are combined to improve the social acceptability of human-following system,and achieve the robot's cooperative control of following and avoiding obstacles.The effectiveness of the control strategy is verified by Matlab/Simulink platform.