LAUSR

Abstract In this study, Artificial Neural Networks, Extreme Machine Learning and Support Vector Machine (SVM) are used to estimate power consumption of Brushless DC motor in Unmanned Aerial Vehicle (UAV). Durafly 3648 Brushless DC motor of UAV is modelled with the Finite Element Analysis software. Then it is simulated with Ansys-Rmxprt according to pulse degree, speed and battery voltage of the UAV. The training times and Mean Absolute Percentage Errors (MAPE) of the Artificial Intelligence Techniques (AITs) are calculated for motor input power estimation. The best result among the implemented AITs is achieved with the MAPE of 0.269% by the SVM model. Then the graphical user interface (GUI) is developed to easily obtain information about how efficient the battery can be used, and the flight time of UAVs. Thanks to the proposed GUI software, the input power of the motor can be estimated via input parameters without the long-time consuming simulations.