Development of metaheuristic algorithms for efficient path planning of autonomous mobile robots in indoor environments

摘要

Application of efficient path planning algorithms for Autonomous Mobile Robots (AMRs) in environments with obstacles is a significant challenge in robotics research. Existing methods, such as A-star (A*) algorithm, can provide optimal paths but suffer from high computational complexity and may not be suitable for dynamic environments. This study explores the potential of three metaheuristic algorithms - Improved Particle Swarm Optimization (IPSO), Improved Grey Wolf Optimizer (IGWO), and Artificial Bee Colony (ABC) algorithm – in planning high-speed and smooth paths. These algorithms are selected due to their ability to find near-optimal solutions efficiently, avoid local optima, and adapt to changing environments. In this study, the researchers designed and built an AMR using a Raspberry Pi 4 microcontroller as the main processing unit, working in conjunction with an Arduino Mega for controlling the DC motor drive through an MDD10A motor driver circuit. The robot is equipped with an RPLiDAR A1 sensor to read 360-degree distance values for mapping and obstacle avoidance. The experimental results clearly indicate that the metaheuristic algorithms, especially ABC, can calculate paths up to 19% shorter than A* while requiring only one-tenth of the time. Moreover, ABC demonstrates superior motion smoothness when applied to the actual robot, enabling it to better adapt to rapidly changing work environments. This work represents a significant step in developing algorithms for robots that are ready to support real-world operations in industries, logistics, healthcare, or various service sectors, helping to increase efficiency and reduce operating costs in the future.