Demand-side management in Grid-Connected Energy Storage System using Deep Neural Network
A novel energy management system to improve the efficiency of grid-connected energy storage systems using a deep neural network is developed. The high penetration of renewable energy and decentralization of the grid has led to an increase in the instability of the grid. To reduce this instability, a balance between the consumption demand and production rate needs to be maintained. For this objective, electric vehicle batteries can be integrated with demand-side management techniques using a deep neural network. The controller can be programmed with the timing of the peak and the off-peak hours obtained from the demand curve data and state of charge of the battery. The controller will take two inputs: The time of the day and the State of Charge of the battery. The NN controller will detect the arrival of the peak and will send a message to the EV battery to supply a programmed percentage of power to the household appliances. The direct communication between the grid and the battery can be eliminated to reduce the infrastructure requirements and data processing. The grid can operate successfully during normal working hours and can supply the total power consumption by the loads at any time of the day. The peak to average power ratio can be reduced by operating the EV battery during peak hours for providing that programmed percentage to the appliances for better grid operation. This drained battery will be further fully charging during low loading of the grid and keep ready for the following days’ operation. According to the results of simulation studies, it is demonstrated that our proposed model not only enhances users’ utility but also reduces energy consumption costs.