Agent Based Economic Framework For Resource Distribution in Household Smart Grid

Electrical grids are extremely large, highly complex and dynamic systems. The depletion of natural resources used for power generation has led to the facelift of the traditional grid system to a more efficient and smarter grid system. Smart grids aim at better energy utilization, right from generation to consumption. The electrical grid is similar to the computational grids in many ways where the issue of resource utilization has been researched upon. In a computational grid the resources are usually CPU time or storage space and the producers and consumers of the resource are the computers themselves. In an electrical grid, the resource is the electricity, a house (a neighborhood, a city, a country) can be considered as the network, the producers are generation stations & distributed sources and the consumers are the various appliances at home. In the present scenario all the appliances connected to the grid have unrestricted access to electricity, meaning that all devices can consume electricity at will. This setup on a global scale is very expensive to maintain as a high reserve capacity is required to meet the dynamic and unanticipated demand which is inefficient. In this thesis, a market based approach is proposed for efficient allocation of limited resources. The introduction of a consumption limit in the system leads to a more optimal and efficient grid which helps in load forecasting, reduction in the amount of spinning reserves and dynamic pricing. By using a combination of producer pricing policy (perfect competition and monopoly) and agent bidding strategies (prioritized bidding and budgets limits) we are able to achieve a suitable level of resource utilization and device utilization pertaining to the needs of the end consumer. By achieving a satisfactory level of operation at a household level when a consumption limit is introduced, we can achieve better resource management and stability at the higher system levels.Embedded SystemsComputer EngineeringElectrical Engineering, Mathematics and Computer Scienc