Publication

Abstract : Considering the recent trends and the rapid growth of Unmanned Aerial Vehicle (UAV) and their applications, enhancing the reliability and robustness of the Unmanned Aerial Vehicle involves various factors to be taken into consideration such as Flight Time, Navigation, Geo-positioning, Payload, Safety, and Security of the Unmanned Aerial Vehicle. One of the most pressing issues of the UAV deployment is the Flight Time. The Flight Time is directly proportional to the energy that is required by the Unmanned Aerial Vehicle, the most common energy source for a UAV is lithium-based batteries. The Flight time of a UAV can be increased by increasing the capacity of the battery as well as the efficiency of the battery. Increasing the battery capacity in turn affects the payload capacity of the Unmanned Aerial Vehicle. Hence, increasing the efficiency of the battery ensures the improvement in the flight time of the Unmanned Aerial Vehicle. The proposed framework is based on Machine Learning techniques, that involve SoC estimation and comparative analysis of the battery's performance. The performance data of both the battery's Lithium Ion and Lithium Polymer are collected and then subjected to training and testing of the data through various ML regression techniques. The MAE, MSE, RMSE, and R-squared values are then analysed to provide the necessary conclusion on the model's accuracy in estimating the SOC of the batteries and analysing the performance of the batteries. The proposed framework is deemed effective in estimating the SoC of the batteries and the performance analysis of the battery, compared to the conventional methods of estimation and observation.