Research on path planning for four-wheel independent drive mobile robots based on minimum turning radius hybrid mode

Abstract

Robotic systems are increasingly integrated across various sectors, from manufacturing to logistics, yet robot navigation in construction environments remains challenged by issues such as safety concerns, high energy consumption, and suboptimal operational efficiency. To address these challenges, this study introduces a global path-planning framework for mobile robots based on a four-wheel four-steering (4W4S) kinematic model. The proposed framework combines Informed-RRT ^* with RRT-Connect to accelerate exploration and employs an Euclidean Signed Distance Field map for efficient obstacle representation and collision checking. After path generation, redundant waypoints are removed using an enhanced Douglas–Peucker algorithm, incorporating obstacle detection to boost efficiency and eliminate manual threshold tuning. The key contribution of this work is to propose a 4W4S-specific path optimization strategy that transforms sampling-based paths into straight–arc connections satisfying the minimum turning-radius constraint, while allowing zero-turn motion when no Ackermann-feasible connection is available. Finally, polynomial trajectory optimization fine-tunes the velocity profile to guarantee dynamic feasibility and path smoothness in construction environments. Algorithm comparisons show that the proposed planner generates compact and smooth paths while maintaining sufficient clearance from obstacles. The resulted trajectories are more concise and consistent with human navigation behavior. In real-world tests, the 4W4S robot executes these optimized paths with reduced velocity fluctuations and lower energy consumption, demonstrating enhanced efficiency and reliability.