Coverage path planning for UAVs based on enhanced exact cellular decomposition method

In this paper, an enhanced exact cellular decomposition method to plan the coverage path of UAVs in a polygon area is proposed. To be more specific, the contributions of the paper are: firstly, the turning motion of UAVs is shown to be less efficient from the viewpoints of route length, duration and energy. Secondly, the problem of coverage Path Planning (CPP) in a convex polygon area is transformed to width calculation of the convex polygon, and a novel algorithm to calculate the widths of convex polygons with time complexity of O(n) is developed. The path of the least number of turns for an UAV based on the widths of convex polygons is devised. Thirdly, a convex decomposition algorithm for minimum width sum based on the greedy recursive method which revolves around decomposing the concave area into convex subregions is developed. It is proved that the algorithm is a polynomial time algorithm. To avoid unnecessary back and forth motion, some entirely adjacent subregions are combined. Finally, comparing different weights of two joint-points, a subregion connection algorithm based on minimum traversal of weighted undirected graph is proposed to connect the coverage paths of the subregions. Simulation results show that the proposed method is feasible and effective.

Research highlights
► Coverage Path Planning in a convex polygon area is transformed to width calculation of the convex polygon.
► A novel algorithm to calculate the widths of convex polygons is developed.
► The path of the least number of turns for an UAV is devised.
► A convex decomposition algorithm for minimum width sum is developed.
► A subregion connection algorithm is proposed.