Search

The lifespan of a satellite is primarily dependent on its battery performance. Thus, proper management and monitoring of the battery is important. Most miniaturized satellites of cubeSat and nanosatellite primarily rely on battery voltage readings for monitoring and seldom provide battery health status in a satellite. As the voltage readings can be affected by satellite operating conditions such as temperature and battery lifespan, it can give unreliable readings that might jeopardize satellite operation. The availability of the battery health status can prevent unexpected battery failures and provide useful insights into the planning of satellite operations. In this article, the battery management system of a satellite with state-of-charge (SOC) and state-of-health (SOH) monitoring is presented. For SOC estimation, a scaled unscented Kalman filter (UKF) is proposed. When compared to the existing UKF approach, it requires fewer sigma points and the positive weights used in the scaled unscented transform ensure the positive semidefiniteness of the covariance matrix leading to the improved numerical stability of the filter. Conversely, SOH monitoring is achieved by taking advantage of the opportunity arising from the satellite operations. The battery parameters are extracted without artificially injecting charge and discharge pulses. To validate the performance of the BMS, the experimental prelaunch tests and the actual in-flight data of the VELOX-II satellite is used as an example. From the results, the SOC estimation error is approximately 1% and the SOH estimation is consistent with the manufacturer's datasheet. [ABSTRACT FROM AUTHOR]

Commercial photovoltaic (PV) modules are made of different PV cell technologies such as monocrystalline, multicrystalline, amorphous silicon, and copper indium diselenide. Usually, a single- or a double-diode PV model is used to model the output characteristic. However, it is unclear which PV model is optimal for each PV cell technology as the modeling accuracy is dependent on the PV model used. To fill up this research gap, a generalized multidimension $(n \times m)$ diode PV model is proposed in this paper to determine the optimal PV model. The proposed PV model allows the diode network to be configured to better fit the output characteristics of different PV cell technologies. Both simulation and experimental results are presented to illustrate the advantages of the proposed model. From the results, the optimal PV model that matches each of the cell technology is established. They can be used as a reference PV model for future works. [ABSTRACT FROM PUBLISHER]