Energy-aware Multi-UAV Coverage Mission Planning with Optimal Speed of Flight

TL;DR

A novel mCPP method is proposed that uses the optimal flight speed for minimizing energy consumption per traveled distance and a simple yet precise energy consumption estimation algorithm that is utilized during the mCPP planning phase to outperform state-of-the-art methods in terms of computational time and energy efficiency.

Abstract

This paper tackles the problem of planning minimum-energy coverage paths for multiple UAVs. The addressed Multi-UAV Coverage Path Planning (mCPP) is a crucial problem for many UAV applications such as inspection and aerial survey. However, the typical path-length objective of existing approaches does not directly minimize the energy consumption, nor allows for constraining energy of individual paths by the battery capacity. To this end, we propose a novel mCPP method that uses the optimal flight speed for minimizing energy consumption per traveled distance and a simple yet precise energy consumption estimation algorithm that is utilized during the mCPP planning phase. The method decomposes a given area with boustrophedon decomposition and represents the mCPP as an instance of Multiple Set Traveling Salesman Problem with a minimum energy objective and energy consumption constraint. The proposed method is shown to outperform state-of-the-art methods in terms of computational time and energy efficiency of produced paths. The experimental results show that the accuracy of the energy consumption estimation is on average 97% compared to real flight consumption. The feasibility of the proposed method was verified in a real-world coverage experiment with two UAVs.

Figures (4)

• Figure 1: Stitched images taken by two UAVs in a real-world experiment of the proposed energy-aware coverage path planning method together with the photos of used UAVs. The white polygon represents the area of interest, the red triangle is a no-fly zone and the green lines are the flown trajectories of the UAVs.
• Figure 2: Problem transformation into an MS-TSP instance. Each value $w_{s\mathbf{x}}$ corresponds to the energy needed to perform sweeping pattern $\mathbf{x}$. Each $w_{x,y}$ denotes the energy needed to get from the last point of sweeping pattern $x$ to the first point of sweeping pattern $y$. Init and End nodes represent UAV's initial and end positions.
• Figure 3: (a) velocity vectors during turn, (b) assumed turn trajectory, (c) speed profile projected on path segment.
• Figure 4: Scenarios used for experimental evaluation of the proposed method. Shade-of-green polygons represent fly-zones, red polygons are no-fly-zones, and red dot markers represent UAVs' initial positions. All values in the scales are in meters.