Y6 tricopter autonomous evacuation in an Indoor Environment using Q-learning algorithm

Emergency situations in large public and residential buildings, earthquake, fire, flood, terrorist attacks, cause extreme physical and emotional behaviours, inter alia, anxiety, hyperactivity, anger, etc.; in these situations, people are often unable to take the right action or even unable to make a decision. This paper addresses the problem of generating a building evacuation plan with the help of a Y6 coaxial tricopter UAV in an emergency situation where GPS signal is not available. The proposed algorithm, stochastic Q-Learning, learns the shortest path to leave the building. The traditional 2D space navigation is extended to the challengeable 3D space, which makes our approach more applicable in the real world. The emergency evacuation system proposed in herein can navigate people to evacuate a building safely in the wake of an emergency situation.