Study Of Trajectory Tracking Control Of Methods On Nonholonomic Mobile Robot

With the constant development of robotics, the control science of mobile robot has gradually become an important branch of robotics. The research of control algorithm of mobile robot, which has great theoretical and practical value will also promote the cutting-edge discipline research, such as cognitive science, pattern recognition, nonlinear control method and so on.The thesis aims at the trajectory tracking problem in mobile robot motion control, based on the non-homonymic mobile robot, analyzes several related trajectory tracking control algorithms, their merits as well as demerits. Based on the in-depth analysis of the track with the characteristics and mathematical model of mobile robot, the thesis also aims at the uncertainty in the process of mobile robot operation and the anti-jamming of control algorithm, which bring up two control algorithms. These control algorithms meet the requirements of control system, and can be able to adapt to the complex trajectory tracking.Firstly, we design two fuzzy controllers for trajectory tracking of robot control based on the in-depth analysis of the control volume on impact of tracking error. The speed fuzzy controller determine the tracking speed, according to different curvature, while the velocity fuzzy controller control the angular, based on the e and ec. The designed controller improves the tracking speed, reduces the tracking error, overcomes the non-linear and uncertain factors, solves the dependence of mathematical model, and has strong anti-interference for Gaussian noise and can adapt to the complex environment, and the simulation results also validate the feasibility and superiority of algorithm.Secondly, we give a PD control algorithm BP neural network based on the in-depth analysis of structural characteristics of the path tracking. The algorithm can identify the path of the track based on fluctuations in the path and the slope of the situation, which can optimize two parameters of PD in operation at real time. The PD control algorithm in BP neural network has simple structure, can be conveniently achieved and can be adapted to the complexity of tracking the path, which significantly increases tracking speed and ensures the tracking error. Finally, we design the hardware and software of control system for the intelligent vehicle, using MC9SDG128chips as a master control, and CMOS camera for detection of environmental information. We has prepared the control software by the FOR HCS12V4.6. After debugging, the intelligent smart car can track the path with high speed and accuracy.