Path planning for UAVs

In this paper, a two step path-planning algorithm for UAVs is proposed. The algorithm generates a stealthy path through a set of enemy radar sites of known location, and provides an intuitive way to trade-off stealth versus path length. In the first step, a suboptimal rough-cut path is generated through the radar sites by constructing and searching a graph based on Voronoi polygons. In the second step, a set of nonlinear ordinary differential equations are simulated, using the graph solution as an initial condition. The ODEs describe the dynamics of a set of virtual masses located in a virtual force field. The virtual forces push the masses away from the radar and toward one another. The ODEs are simulated to find a locally, exponentially stable equilibrium solution, which is interpreted as the optimal path. A simulation is provided to illustrate the idea.