Tracking Control For Nonholonomic Mobile Robots

Since the birth of mobile robots, they have no image which can be substituteddisplaying in sight of people.Mobile robot is a new discipline which is developing inrecent years. It is a comprehensive discipline which involves mechanical manufacture,electronic technology, automation, computer and intelligent science and so on. Mobilerobots reflect the peak of the electromechanical integration achievement. In order toget rid of natural forces and realize the human dream, people extremely hope thatmobile robot can complete all the tasks in daily life for human beings one day, andobey the orders given by human beings. Nowaday, mobile robots have become themain ones that explore unknown environment. From the exploration of Mars to theoperation of deep sea, it is inseparable from the help of mobile robot. Therefore, thetechnology of mobile robot has important theoretical significance and practical value.The primary technology of mobile robot is to solve the problem of mobile robotmotion control.Tracking control and path planning are the most fundamental and mostcore part of mobile robot motion control which deeply draw researchers’ attention.Path planning refers to how to find a safe motion path with no collision from astarting point to an end point based on the premise of surrounding information fromdifferent kinds of sensors of mobile robot. According to path planning, tracking controlmake the mobile robot move along the desired trajectory through adjusting themovement direction and velocity of mobile robot. This paper mainly studies dynamictracking and trajectory tracking of nonholonomic wheeled mobile robot based on thekinematic model. Specifically, the main research work involves the following severalaspects.Firstly, two kinds of different sensors are fused in view of different characteristicsand deficiencies of sonar sensors and infrared sensors. Based on the integration ofRSR type wheeled mobile robot system, the dynamic tracking of moving objectmovement is proposed in this paper.Secondly, the trajectory tracking control problem for mobile robots is analyzed.The discrete kinematic model is taken as the research object.This paper proposes amethod which uses Kalman filtering and iterative learning control. This controlmethod is combined with Kalman filtering algorithm based on the improved iterativelearning control law.It speeds up convergence and reduces tracking error. Also, this control approach makes it robust and more resistant to interference. Computersimulation verifies the correctness and effectiveness of this method.Thirdly, according to the kinematic model of mobile robot, the control law formobile robot is constructed by two parts including feedforward and feedback. Thefeedforward consists of sliding mode controller with the use of backstepping. Thefeedback adopts the method of Lyapunov function to design a feedback controller.The simulation results show that this control law obviously improves the trackingcontrol accuracy.Fourthly, the tracking controller is completed by two steps in order to meet globalasymptotic stability. In the initial stage, the control law of angular velocity is designedby the fast terminal sliding mode so as to stabilize asymptotically forward Angle error;Next, The control law of linear velocity is constructed by Lyapunov method in orderto stabilize asymptotically position error. The computer results show that thiscontroller is effectiveness.Fifthly, according to the characteristics of trajectory tracking control of thenonholonomic mobile robot, this paper combines finite time control with sliding modecontrol. The ability of mobile robot tracking control is improved effectively.Sixthly, this paper proposes a fuzzy open-closed-loop PID-P type nonlineardiscrete iterative learning control law in view of the kinematic model ofnonholonomic mobile robot. The convergence of the control law is proved. Theexperimental results show that this control law improves the convergence rate of theiterative learning control algorithm and reduces trajectory tracking error for themobile robot.Finally, this paper summarizes all the work, and the follow-up work isprospected.