Multi Robotic Path Planning in a Static Environment

摘要

Robotic systems are becoming inevitable part of our lives. The demand for robotic system, which can perform multiple tasks with less complexity and energy consumption, is very high in the market. Multi robot system is the most recommended system for such requirements. In the case of multi-robot system reaching their intended destination and performing the assigned task determines its usefulness and relevance. In any mobile robot system path planning is a crucial step which should be designed carefully. Over the years, many path-planning algorithms have been realized yet there is scope of improvement in this area. Usually robotic path planning will happen either in static or dynamic environment. This work focuses on multi robotic path planning in a known environment (static in nature). The system consist of two robot, which will traverse through a planned path depending on the obstacles. The selection of the robot will be based on a score which is generated by considering multiple parameters. The cooperative behavior of these robots enables them to perform the task assigned to the other robot while the other robot is facing a malfunction or failure. The path planning process is realized using reinforcement learning for namely the Temporal Difference (TD) algorithm. The work also emphasizes on catering the existing TD algorithm to work for multi robotic system thereby increasing the efficiency of the system. The simulation of multi robotic system is carried out with the helpof Python 3.8.5 and SpyderIDE.