Route planning of mobile robot based on improved RRT star and TEB algorithm.

Area of Science:

• Robotics
• Artificial Intelligence
• Computer Science

Background:

• Mobile robot path planning is crucial for autonomous navigation.
• Traditional algorithms like RRT* can suffer from suboptimal paths and kinematic limitations.
• Ensuring smooth, efficient, and kinematically feasible trajectories is a key challenge.

Purpose of the Study:

• To develop and evaluate a fusion algorithm combining an enhanced RRT* for global path generation and an optimized TEB for local trajectory planning.
• To improve path optimality, reduce length and turning points, and enhance path smoothness.
• To ensure the generated trajectories adhere to mobile robot kinematic constraints.

Main Methods:

• An enhanced Rapidly-exploring Random Tree star (RRT*) algorithm featuring adaptive sampling and node bias for accelerated global path generation and reduced local optimality.
• Path redundancy elimination and turning angle constraints to minimize path length and improve smoothness.
• Kinematic modeling of the mobile robot and optimization of the Trajectory Encoding Bundle (TEB) algorithm for kinematically feasible local trajectories.

Main Results:

• The enhanced RRT* algorithm reduced path length by 5.8% and the number of turning points by 62.5% compared to the traditional RRT*.
• The fusion algorithm successfully generated globally optimal, smooth paths that effectively circumvent obstacles.
• Local trajectories generated by the fusion algorithm precisely matched the mobile robot's kinematic constraints.

Conclusions:

• The proposed fusion algorithm offers a superior approach to mobile robot path planning.
• The enhanced RRT* and optimized TEB integration results in more efficient, smoother, and kinematically compliant robot navigation.
• This method demonstrates significant improvements over conventional RRT* for complex path planning tasks.