An enhancement of agent-based power supply-demand matching by using ANN-based forecaster

Local supply-demand matching in power grids by means of advanced information and communication technology (ICT) is emerging due to the increasing integration of distributed energy resources (DER). Although advantages of the local matching mechanism have been proved by either research works or demonstrations, there are some difficulties on being proactive to handle uncertainty from renewable energy sources (RES) and new types of load consumption. This paper aims to enhance the matching mechanism using multi-agent systems (MAS) and artificial neural network (ANN) to investigate and determine DER´s flexibility to compensate that uncertainty. Under a more general platform for smart grid functions, this paper presents a model to achieve a match between the forecasted supply and demand. Short-term forecasting based on an ANN model is used to predict the stochastic behavior of weather data. The model considers various scenarios and the potential from household demand side management. The results from the performed simulations indicate feasible DER´s flexibility for power matching, which can be further adapted for different scenarios to serve local or more grid-related optimization objectives.