Abstract: Our target is finding the optimal coverage path planning (CPP) in given area including obstacles for multi small unmanned air vehicles (UAV). First, the given area is discretized into cells. And to degrade the computation complexity, the divide areas based on robots initial positions (DARP) algorithm is utilized to divide the area into separated equal subareas, transforming the multi-CPP problem into several CPP problems. Then, an end nodes exchanged method based on minimum spanning trees is proposed to adjust the shape of subareas and minimum spanning trees. Finally, the coverage paths of each subareas are constructed basedon the optimized minimum spanning trees. And the simulation confirms that the proposed method can guarantee the completeness, nonoverlaps, minimum lengths and the least number of turns.

Key words: unmanned air vehicle (UAV), coverage path planning (CPP), area subdivision, spanning trees （STs）, minimum turns