Improved Ant Colony Algorithm-based Automated Guided Vehicle Path Planning Research for Sensor-aware Obstacle Avoidance

Abstract

Automated guided vehicles (AGVs) are the main delivery vehicle for the horizontal transport of containers between the quayside and yard of automated container terminals (ACTs). The coordination of AGVs with the quayside bridge and yard bridge is necessary for loading and unloading operations at the wharf and to improve logistics management efficiency. Toward solving the problem of AGV path planning and sensor-aware obstacle avoidance in a dynamic complex environment for the Internet of Things (IoT), we proposed an improved ant colony algorithm based on an adaptive dynamic parameter adjustment strategy (IACA-ADPA) in this paper. The grid method is first used to construct a motion space model because it is easy to implement, analyze, store, and express, and make the AGV reach its target node safely and smoothly. Then the proposed IACA-ADPA is used for global path planning and for efficient AGV path design and adjustment. Finally, the improved time window adjusts the waiting time of the AGV to avoid local collisions. The simulation results of different scale paradigms show that the IACA-ADPA can effectively avoid road section obstacles and node obstacles, and improve the safety and efficiency of a multi-AGV system.

Adaptive path planning for unknown environment monitoring

“…Researchers have shown the context in action via handling complex indoor 3D quad copter situations of navigation that need planning ahead of time. Waiting for other people and quad copters, as well as taking diversions around them [30]. Qing and team explores path searching, it records all equidistant shortest paths and chooses the best path among numerous equidistant paths based on the turn time, resulting in the path with the shortest distance and time [47].…”

“…Another extension is to use principles from fast MPC to increase real-time performance [6]. Third, to improve robustness, obstacle motion uncertainty must explicitly be included [30]. Fourth, the path planning process can be combined with Deep Learning or SLAM to assist UAVs to create a real-time map of the environment [18].…”