A Dynamic Residential Load Model Based on a Non-homogeneous Poisson Process

Smart grids treat energy in a much more efficient manner than it is done currently by conventional power systems. One important aspect in the design and analysis of smart grids is to take into account some real characteristics of the power system and its loads. In this context, we propose a dynamic load model for residential applications based on a non-homogeneous Poisson process, whose parameters depend on the electrical characteristics of the loads and their time-varying power profiles. Some of the advantages of this model are its flexibility to represent any specific energetic scenario found in different regions of the globe and the possibility to independently control individual low-voltage loads of the evaluated system by modifying the activation function of the corresponding load model. This last characteristic is fundamental to represent adequately and analyze the behavior of smart grids. To demonstrate the accuracy and effectiveness of the proposed strategy, several load curves were generated with the aid of the proposed model and compared against real measurements.