Multi-Agent Based Simulation of an Unmanned Aerial Vehicles System
Abstract
The rapid growth of using Unmanned Aerial Vehicles (UAV) for civilian and military applications has promoted the development of research in many areas. Most of the unmanned aerial vehicles in use are manually controlled. Often, UAVs require highly trained pilot operators. Hence, the main challenge faced by researchers has been to make UAVs autonomous or semiautonomous. The goal of this research project is to develop and implement a simulation for a user-defined environment allowing UAVs to maneuver in free environments and obstacle-laden environments using Boid's algorithm of flocking with obstacle avoidance. The users are permitted to analyze the maneuvering area and coverage efficiency of the UAVs and to dynamically change environments. This project makes use of Boid's flocking algorithm to generate different kinds of movements for the flying agents, enabling the user to analyze the effectiveness of patrolling in that particular scenario. The number of UAVs and the type of environment are set by the user. The set number of UAVs moves as a flock or swarm inside the set environment by using Boid's rules of flocking: cohesion, alignment, and separation. The coverage efficiency of the UAVs in that particular environment is reported based on the ratio between the area covered and the time when the search time reaches a threshold. The advantages and feasibilities of the approach are discussed with the simulation results.