Intelligent Optimisation for Autonomous Mobile Robot Path Planning: Enhanced GA and PSO Approaches

Abstract

Autonomous mobile robots (AMRs) require efficient, reliable, and adaptive path planning to operate safely in complex environments. This paper reviews path planning algorithms for AMRs and discusses intelligent optimisation methods for improving navigation performance.It first introduces the classification of path planning methods, including global planning, local planning, and hybrid planning, and explains key evaluation metrics. Conventional algorithms, including Dijkstra, A*, and Bezier-curve-based smoothing, are analysed in terms of their principles, advantages, and limitations.The paper then focuses on intelligent optimisation algorithms, especially Genetic Algorithm (GA) and Particle Swarm Optimisation (PSO), and proposes improvement strategies such as hybrid initialisation and mutation probabilities. Simulation results show that optimised GA and PSO have better paremeters including planning time and stability compared with their standard versions. Finally, the paper discusses practical applications in logistics, manufacturing, and hospital service robots, while identifying challenges such as parameter sensitivity, simulation-to-reality transfer, multi-robot coordination, and human-robot trust.