Real-time battery SOC estimation under hybrid power conditions using fast-OCV curve with unscented Kalman filters

Abstract

Unmanned aerial vehicles (UAVs) are drawing more and more attentions recent years and widely applied. Nevertheless, it is generally challenged by short time of duration because of the lack of energy density for battery. A robust power supply is indispensable for advanced UAVs thus hybrid-power might be a promising solution. State of charge (SOC) estimation is essential for power system of UAVs. While, accurate SOC estimation always challenged partly ascribed to accurate open circuit voltage identification. Considering the actual working condition of battery under hybrid condition, this paper proposed a novel method “fast-OCV” for obtaining open circuit voltage of battery. It is proved that fast-OCV showed great advantages related to simplicity and time cost over traditional way for obtaining OCV. Moreover, fast-OCV also showed better accuracy over traditional OCV. As a supplementary, this paper also proved that limited memory recursive least square algorithm was a good way for parameter estimation. With such algorithm, distinct noise was encountered when using single-mode. In compare, batch mode for parameter estimation showed much better performance with distinctively weaker noise.