IACO-DQ Path Planning for AGV in Complex and Dynamic Environments

This study addresses the problem of autonomous navigation and obstacle avoidance for mobile robots in complex and dynamically changing environments. A path planning strategy is proposed, namely the IACO-DQ (I-ACO and Dyna-Q) algorithm, which utilizes I-ACO (Improved Ant Colony Optimization) to construct an initial global optimal path framework. This is combined with the model learning and planning capabilities of the Dyna-Q algorithm, allowing the AGV to predict future states through the environment model during actual exploration, thus achieving online real-time dynamic path planning. The goal is to provide AGV with an efficient and flexible path planning solution to handle dynamic obstacles, ensuring safe and rapid task completion in ever-changing environments. Simulation results demonstrate that the IACO-DQ algorithm significantly improves the AGV's obstacle avoidance capabilities and path planning efficiency in dynamic environments, showcasing enhanced robustness and adaptability.

I. Introduction

Path planning for mobile robots refers to the process of determining the optimal route for a robot to travel from a starting point to a target point [1]–[2]. This process is one of the core challenges in the field of mobile robotics. Common path planning algorithms include the A * algorithm, RRT (Rapidly-exploring Random Tree) algorithm, and ACO (Ant Colony Optimization) algorithm [3]–[4]. These algorithms utilize heuristic information and cost functions to search for the optimal path within the environment. Furthermore, due to the potential variability of environments, path planning necessitates real-time and dynamic processing capabilities to promptly adjust routes in response to changes. Commonly employed algorithms for handling these dynamic challenges include the DWA (Dynamic Window Approach) algorithm and reinforcement learning algorithms, which further enhance the flexibility and adaptability of path planning [5]–[6].