<![CDATA[The battery is an important component in the context of implementing renewable energy. The type of battery that has a density in energy storage is lithium polymer. The parameter in the battery that must be considered is the State of Charge (SOC) estimation. In general, the SOC battery estimation uses the coloumb counting method because the difficulty level is low. However, there are weaknesses in the dependence on the utility of the current sensor which is used as an accumulation of the integral of the incoming and outgoing currents over time. In this study presents Back Propagation Neural Network (BPNN) as an algorithm for estimating SOC based on OCV-SOC characteristic curves. The OCV-SOC characteristic curve of the battery is obtained from the battery pulse test. Battery voltage, current and discharging time are used as the first BPNN input layer for the estimation of Open Circuit Voltage (OCV). OCV will be learned as the second BPNN input layer for estimating battery SOC. The results of SOC estimation simulations obtained an average error of 0.479% against the real SOC based on the characteristic curve of OCV - SOC.]]>
Copyrights © 2020
