Three-dimensional offline path planning for UAVs using multiobjective evolutionary algorithms

In this paper, we present 3D offline path planner for unmanned aerial vehicles (UAVs) using multiobjective evolutionary algorithms for finding solutions corresponding to conflicting goals of minimizing length of path and maximizing margin of safety. In particular, we have chosen the commonly-used NSGA-II algorithm for this purpose. The algorithm generates a curved path which is represented using B-spline curves. The control points of the B-spline curve are the decision variables in the genetic algorithm. In particular, we solve two problems, assuming the normal flight envelope restriction: i. path planning for UAV when no other constraint is assumed to be present and ii. path planning for UAV if the vehicle has to necessarily pass through a particular point in the space. The use of a multiobjective evolutionary algorithm helps in generating a number of feasible paths with different trade-offs between the objective functions. The availability of a number of trade-off solutions allows the user to choose a path according to his/her needs easily, thereby making the approach more pragmatic. Although an automated decision-making aid is the next immediate need of research, we defer it for another study.