Mechanical and Computational Energy Estimation of a Fixed-Wing Drone

Abstract

In this paper, we present a case study on the energy estimation of drones and derive a general modeling approach that estimates computational and mechanical energy separately. The mechanical energy model can easily be extended to other drones and is built using a Fourier series from a number of training flights. The computational energy model is more advanced as it handles heterogeneous hardware and incorporates a specification that defines the quality-of-service ranges for software components of the robotic system. The computational model is suitable for any mobile robot and is implemented in a modeling tool. The tool automatically generates an energy model from the specification by performing a set of empirical trials for selected configurations while approximating others. Information about the battery State of Charge is also included in the tool, hence allowing the evaluation of how different software configurations impact the battery. This approach can be used for dynamic mission assessment regarding different planning decisions. We here have demonstrated its ability to model the energy of a specific mission performed at varying levels of quality-of-service using a specific drone.