Artificial Neural Network-Based Hybrid Model for Efficient Battery Management

The utilization of renewable energy is increasing due to environmental pollution. Power transmission to a remote location is also challenging. In this research work, a hybrid standalone power generation framework has been developed, which effectively utilizes the concept of renewable-based power generation and energy storage system for a real-time scenario. A detailed modeling approach for the power generation sources like solar PV and fuel cells along with the battery-based energy storage system has been carried out to develop an efficient artificial neural network (ANN)-based hybrid energy management system. The performance of the ANN was evaluated using the following performance metrics mean-squared error = 0.0461, root-mean-squared error = 0.2148, normalized root-mean-squared error = 0.043, and R-value > 0.99. The system is designed to operate the battery within safe operating limits. For the proposed energy management system, six modes of operation are considered. The system has been developed in MATLAB/Simulink environment and tested for a real-world scenario. With the proposed control architecture, the system performance is found better in terms of reducing the electricity cost by utilizing renewable sources efficiently.