MURMOEA: A pareto optimality based multiobjective evolutionary algorithm for multi-UAV reconnaissance problem

Abstract

The objective of multiple Unmanned Aerial Vehicles(UAVs) reconnaissance is to employ different kinds of UAVs conducting reconnaissance on a set of targets within predefined time windows at minimum cost, without violating the real-world constraints. This paper presents a mathematical formulation for the problem, which is a multi-objective optimization problem. A Pareto optimality based multi-objective evolutionary algorithm, MURMOEA, is put forward to solve the problem. In MURMOEA, an integer string is used to represent the chromosome. Pareto dominance based tournament selection with elitism strategy is introduced, which ensures that MURMOEA converges toward the Pareto set and prevents bias to any object. A novel sequence crossover operator is designed to ensure the feasibilities of the children, and a problem specific forward insert mutation operator is designed to ensure the validity of the mutated individuals. Finally the simulation results show the efficiency of our algorithm. © Springer-Verlag Berlin Heidelberg 2006.