Research On Probabilistic Uncertainty Trajectory Planning Of Wheeled Mobile Robot

Wheeled mobile robots are widely used in modern industry,aerospace,military and civilian fields because of their excellent characteristics such as strong carrying capacity and high flexibility.Trajectory planning,as an important part of realizing autonomous robot motion,can not only plan feasible paths for mobile robots,but also contains information such as speed and acceleration that are essential for the subsequent trajectory tracking control of mobile robots.However,due to the interference of uncertain factors in itself or in the environment,it is difficult for wheeled mobile robots to operate normally according to the pre-planned path.In the past,the research on trajectory planning of mobile robots rarely predicts the effect of uncertainty on the global trajectory.influences.Therefore,this paper considers the influence of probability uncertainty on the trajectory planning of wheeled mobile robots,so as to provide a robust trajectory for wheeled mobile robots.This dissertation takes the mobile robot with two-wheel differential drive as the research object,establishes the kinematics and dynamics model of the wheeled mobile robot under the pure rolling condition,and selects the dynamic model as the research model of the trajectory planning problem.Considering the existence of uncertain factors from two perspectives,the influence of probability uncertainty on the global trajectory is studied separately.One is to consider the uncertainty of the wheeled mobile robot's own model parameters and the initial state of the system,and propose a robust trajectory planning problem to predict the influence range of uncertainty on the trajectory.Firstly,the problem of deterministic trajectory planning for wheeled mobile robots is established.On this basis,the problem of uncertain trajectory planning is formed by considering the model parameters and the uncertainty of the initial state.Then,in order to better reflect the influence range of uncertainty on the trajectory,a robust trajectory planning formula is proposed to describe the uncertainty trajectory planning problem.Since the proposed robust trajectory planning formula contains the mean and variance functions of the random function terms,it cannot be directly solved numerically.Therefore,this paper seeks to use the generalized polynomial chaos to expand the random function terms involved in the quantified robust trajectory planning sequence to obtain the deterministic function expression of the corresponding statistical characteristics.Finally,the robust trajectory planning formula is transformed into a high-dimensional deterministic trajectory planning problem that can be solved directly,and the statistical characteristics of the state quantity can be obtained by solving the problem.The other is to consider the parameter uncertainty or random interference in the constraint function that the wheeled mobile robot needs to meet,and propose a probabilistic constrained trajectory planning problem for the wheeled mobile robot to solve the reliable trajectory.Probabilistic constraints are different from previous hard constraints in that they allow the existence of constraint violations in the sense of probability.By setting the constraint violation threshold,the influence of uncertainty on the trajectory is limited,thereby improving the control performance of the system.The probabilistic constraint trajectory planning problem is solved by using the convex function approximation method to convert the probability constraints contained in the problem into corresponding deterministic constraints,so that the original problem can be directly solved numerically.The constraint violation rate of the mobile robot trajectory obtained under the framework of probability constraints in the entire time domain is lower than the given threshold,and the reliable trajectory of the wheeled mobile robot under the specified threshold can be obtained.