New paradigm for battery life management in data centers

In this paper new paradigm for battery life management in data centers is presented. A method for remaining useful life (RUL) estimation of batteries based on the voltage drop under full load. In the following study real time data of batteries deployed in a datacenter is used. Dataset has historic values of 3100 batteries over one year, having features like voltage, resistance and temperature. There is very little research been done on this topic the main reason is the unavailability of battery data used in the datacenter. The drop of voltage under full load of a battery have been considered as the performance parameter as most of the battery management systems use these parameters as a remaining useful life of a battery. We propose a new paradigm for battery life management to avoid complete ups failure for data center. We used real dataset obtained from datacenter and verified our results with success in the datacenter.