Intelligent trajectory tracking control method for mobile robot

Abstract

The invention discloses an intelligent trajectory tracking control method for a mobile robot, and the method is characterized by estimating a weight value of a given mobile robot model through the convergence of a least square method; updating the weight of a controller by introducing a neural network and a strategy iteration method in reinforcement learning, solving an optimal value function. The method comprises the following steps: 1) establishing a mobile robot nonlinear system; 2) initializing a state variable of the system, and giving an initial strategy; 3) updating the weight by using a reinforcement learning strategy iteration method, and solving an optimal strategy; 4) introducing a neural network algorithm and calculating a weight by using the convergence of least squares; and 5) stopping strategy updating. According to the method, a relatively good trajectory tracking effect is ensured under the condition that internal dynamic knowledge of a power system is not needed.

Description

technical field [0001] The invention belongs to the technical field of control, and specifically provides an intelligent trajectory tracking control method for a mobile robot, which ensures a better trajectory tracking effect without the need for internal dynamic knowledge of a power system. Background technique [0002] The value of mobile robots in actual production and life is increasing day by day, and the research on mobile robots has also received more attention. With the emergence of various mobile robots in various forms, the research on their control problems has also attracted the attention of relevant researchers. [0003] The dynamic equations of mobile robots are nonlinear, and most of the existing control methods are based on models. Although many feedback linear methods have been widely used in mobile robot control, including exact feedback linearization, approximate feedback linearization and pseudo-linearization, etc., but The model obtained after linearizat...

AI Technical Summary

Problems solved by technology

[0003] The dynamic equations of mobile robots are nonlinear, and most of the existing control methods are based on models. Although many feedback linear methods have been widely used in mobile robot control, including exact feedback linearization, approximate feedback linearization and pseudo-linearization, etc., but The model obtained after linearization still cannot completely replace the nonlinear robot model, and it is very difficult and not feasible to use linearization to model mobile robots, so a data-driven method is proposed

Images