Bacterial evolutionary route planning for unmanned aerial vehicle

Abstract

Based on bacterial evolutionary algorithm and multi-attribute decision making theory, a new route planning method is presented. The vehicle overload is considered during angle encoding; and then during the creation of initial population, the start angle is decentralized selected to avoid premature convergence. To evaluate each flight candidate route synthetically, a multi-attribute decision making algorithm is described including: (1) the Euclidean distance of a route from the origin to its destination, (2) the survival probability of a route, and (3) the turning angle of a route. The experimental result reveals that the planned route can avoid threaten effectively, and the optimization efficiency is improved significantly.