Research On Path Planning And Control Methods For Wheeled Mobile Robot

Mobile robot technology,which has already involved the military industry,product manufacturing,aerospace and other aspects,is causingmore and more experts and scholars' attention from all fields,because of its broad application prospects and development space at home and abroad.The research on mobile robot involves many fields and areas,and the problem of path planning is one of the mostly basic topic,which copes with how the mobile robot to walk in a real environment.Through reading a lot of literature and analyzing the short comings and deficiencies existing in local path planning methods,some improved methods are proposed.Finally,the final task is to complete simulation test with software.To solve the problem of local minima in traditional artificial potential field method,a modified form of repulsion function is proposed by adding a virtual force and judging the local minima occasions with the robot's behavior.Then the robot is able to pull out of the plight timely and quickly.Considering the complex robotic environment,control accuracy and the local oscillation problem of fuzzy control itself,a hierarchical fuzzy control structure has been put forward.The fuzzy-PID control is used for the non-obstacle mode,and the improved fuzzy control is for the obstacle mode.With the help of neural network,the fuzzy control rules can be easily constructed.In order to solve the problems of slow convergence rateof BP algorithm,a modified method is proposed to optimize the convergence speed and adjust fuzzy membership functions.Finally,the proposed solutions mentioned above will be verified by Matlab software,and thesimulation results show that the improved method has solved the problem of local minima and reached the target point,the control accuracy has become better by combining with PID control and adjusting domain,and the convergence speed of neural network has accelerated,and finish the task of path planning with T-S neural network in the end.From the whole,by comparing the simulation result before and after,the improved methods are much more reasonable and effective.