Decentralized Task Assignment for Multiple UAVs using Genetic Algorithm with Negotiation scheme approach

This paper deals with a task assignment problem of cooperative multiple Unmanned Aerial Vehicles (UAVs). The problem about assigning the tasks to each UAV can be interpreted as a combinatorial optimization problem such as Travelling Salesman Problem (TSP), Vehicle Routing Problem (VRP), and Generalized Assignment Problem (GAP). These problems have NP-complete computational complexity which has features such that the computation time cannot be determined in polynomial scale and the problem cannot be solved correctly except for examining all possible solution cases. To solve this combinatorial optimization problem, Genetic Algorithm (GA) which is one of the meta-heuristic algorithms is adopted. By using GA, multiple UAVs-multiple targets-multiple tasks scenario example is simulated, and the results of GA are compared with those of Mixed Integer Linear Programming (MILP) method to verify the optimality. Then the decentralized task assignment method based on chromosomes negotiation scheme approach is employed, and the simulation for a decentralized task assignment scenario is performed to evaluate the validity of the proposed method.