Trajectory Trackirig Control Of Nonholonomic Mobile Robots Based On Neural Network And Sliding Mode Control

In recent decades,the field of mobile robots has received increasing attention,and related research has made great progress.Now robots have been applied to various fields,such as military,industrial and agricultural,daily life,etc.,to complete some heavy and difficult tasks,which also put forward higher requirements for the design of control schemes.Nonholonomic mobile robots are an important part of the robotics field.In this paper,a type of typical tractor-trailer robots are taken as research objects.In the first two chapters,the background,significance,research status and basic knowledge of stability theory and nonholonomic system are introduced respectively.In the third and fourth chapters,the trajectory tracking control of the tractor-trailer robot under the condition of ideation and time delay is studied respectively.The summary and prospect are given.The main research work of this article is as follows:1.Aiming at the trajectory tracking control problem of tractor-trailer robot kinematic model with n trailers under ideal conditions,a class of adaptive neural network controller is designed based on the backstepping method.First,its kinematic chain model is introduced,and then a tracking error model is derived.Secondly,a neural network controller is designed using backstepping method and adaptive method,in which the neural network is used to approximate the unknown terms in the error system.Using Lyapunov's stability theory,an adaptive law for network weights is designed.The robot can track the reference trajectory.2.For the kinematic trajectory tracking control problem of tractor-trailer robot model with n trailers with input time-delay,first convert its time-delayed kinematic model into a chain form,and then obtain the tracking error model.After that,a class of neural network sliding mode controller was designed based on the backstepping method and self-adaptation.The time delay problem was solved by using the Lyapunov-Krasovskii functional,so that the robot can still track the reference trajectory with time delay.3.Simulate the two cases separately to verify the effectiveness of the proposed control scheme.The main innovations of this paper include: using the non-linear mapping ability of neural network to approach the unknown term in the controlled system;Considering the traction trailer robot with time delay and transforming its kinematic model into a chain form,The error system designs a new type of sliding mode surface.