AI-Based Optimization Framework for Scheduling Autonomous Rail-Guided Vehicles in Warehouse Systems

Abstract

Scheduling tasks for autonomous Rail-Guided Vehicle (RGV) systems presents a complex optimization challenge that critically influences warehouse automation performance. This research develops an AI-based RGV scheduling framework that allows configuration of robot parameters such as maximum velocity, acceleration, deceleration, and track dimensions, accounting for velocity constraints imposed by curved tracks. The study includes five computational intelligence algorithms: Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Artificial Bee Colony (ABC), hybrid GAPSO, and hybrid SUPER-SAPSO. The framework integrates path planning, layered collision penalty models, and multi-RGV task assignment under a physics-based travel time model while minimizing RGV idle time and addressing workload imbalance. Experimental results demonstrate comparative analyses of the efficiency and convergence speed of the various computational intelligence algorithms in optimizing overall warehouse efficiency.