Real time cooperative path planning for multi-autonomous vehicles

Abstract

The paper discusses an effective approach on cooperative path planning for two or more autonomous ground vehicles (AGVs). These multi-AGVs coordinate their path effectively to accomplish complex and critical tasks for various applications. The proposed architecture accommodates two or more AGVs in the same workspace and each vehicle in the workspace shares their kinematic and sensor information with the others. Each individual has to replan its own path accounting for other AGVs and obstacles. Effective D* lite algorithm is used to calculate individual paths. The AGVs utilize both proprioceptive (internal sensor data) and exteroceptive (external sensor data) sensors to perceive their surroundings for avoiding interference with each other and with the obstacles. The proposed approach is tested on MobileSim, simulation software, and on pioneer 3DX mobile robots in real-time scenarios. Later, the simulated and experimental results are compared and discussed.