Please use this identifier to cite or link to this item:
|Title:||Distributed Renewable Energy Sources for Load Balancing in Smart Grid|
|Keywords:||Artificial neural network;Load forecasting;Machine learning;Renewable energy sources;Smart grid|
|Abstract:||Existing electric grid is facing one of the major concerns, to decarbonize the electricity generation and consumption at various levels in power sector. However, with an inclusion of information and communication technology (ICT)-based infrastructure in the existing electric grid, it can act as smart grid (SG) by making a balance between demand and supply which in turn decarbonize the environment. Various strategies for efficient energy consumption with reduced dependency on fossil fuels to control carbon emissions are under development across the globe. However, in extreme load conditions, these strategies may not work well due to inefficient usage of renewable energy sources (RES). To achieve the aforementioned goals, an adaptive approach towards distributed generation by incorporating RES in the existing electric grid is required. Moreover, there is a requirement to shift the conventional users from passive consumers to active ”Prosumers” to meet the ever-increasing growth of energy demand during peak hours. Prosumers can feed locally generated energy back to the grid to make a balance between demand and supply in the peak hours. In this paper, a novel scheme to address the aforementioned issues is proposed in which many prosumers are combined into a single unit to incorporate RES in SG. In the proposed scheme, an intelligent Artificial Neural Network (ANN)-based controller is designed for day-ahead load prediction to manage the mismatch between load demand and renewable generation supply in real-time. Also, to ensure energy availability at all times to the end users, a greedy heuristic scheduling algorithm is designed. The proposed scheduling algorithm allows the controller to select between various power options to meet the energy demands. The proposed scheme is evaluated with respect to various evaluation metrics such as-Mean Square Error (MSE), Mean Absolute Error (MAE), and Root Mean Square Error (RMSE). Also, simulation of the proposed scheme for a set of twenty-five prosumers illustrates that the amount of energy drawn from grid is reduced by 46.90% in comparison to the case when RES are not used. The results obtained clearly show the efficacy of the proposed scheme in real-time scenario.|
|Description:||Master of Engineering-CSE|
|Appears in Collections:||Masters Theses@CSED|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.