AN IMPROVED ANT COLONY OPTIMIZATION ALGORITHM FOR AGV PATH PLANNING IN SEMICONDUCTOR MANUFACTURING SYSTEMS

Abstract

Automated Guided Vehicles (AGVs) play a critical role in ensuring efficient and collision-free logistics within modern semiconductor manufacturing systems. However, navigating the highly complex layouts of fabrication plants presents significant challenges for traditional path-planning algorithms. While the standard Ant Colony Optimization (ACO) algorithm is robust, it often suffers from slow early-stage convergence and a tendency to stagnate in local optima due to initial blind searching. To address these limitations, this paper proposes an Improved Ant Colony Optimization (IACO) algorithm tailored for grid-based AGV routing. The proposed IACO enhances search efficiency by introducing a distance-guided heuristic function that exerts a strong directional pull towards the target, thereby minimizing blind exploration. Furthermore, a non-linear adaptive pheromone updating strategy is designed to dynamically balance global search capabilities with local exploitation speed. Simulation experiments conducted in an abstract semiconductor factory environment demonstrate that the IACO algorithm significantly outperforms traditional ACO. The proposed method not only generates shorter and smoother trajectories with minimized redundant turning maneuvers but also achieves an approximately 18% reduction in total path length and a 45% decrease in computational iterations. These improvements confirm that the IACO algorithm is highly effective for the real-time, dynamic scheduling demands of complex industrial logistics.