FPGA based Online Battery SOC Estimator using Weighted Mix Estimation

The untimed power failure is one of major issues confronting the successful commercialization of battery powered electric cars. So, an intelligent model predictive controller which adaptively schedules the load according to the predicted remaining charge of the Electric Vehicles(EV) battery saves the energy content of the battery and ensures the safe state of the Electric Vehicle. The most important part of the model predictive controller is the accurate online estimation of the battery state of charge using an area and time efficient methodology. In this paper the authors proposes and implements an efficient yet less computationally intensive weighted mix estimation method, that combines the enhanced coulomb counting, adaptive battery model and the Open Circuit Voltage(OCV) method. The battery model parameters are tuned adaptively as the battery discharges and the tuned battery model is used by the weighted mix estimator (WME) to estimate the SOC of the battery. Due to adaptive nature of battery model used, WME serves as a robust methodology for online SOC estimation of batteries powering dynamic systems like Battery Electric Vehicles. A comparative analysis of WME has been facilitated with battery model and coulomb counting method for static and dynamic load profiles. It was found that the SOC estimated by WME was accurate and reliable, as it tracks SOC in all discharge phases of the battery under arduous conditions. Further the controller proactively optimizes the constrained battery energy by varying the power delivered to the noncritical loads and supports critical loads as per demand.