Simulación Causal para el Consumo Eléctrico Residencial

A dynamic model based on causal diagrams aims to identify all the factors involved in the behavior of a specific phenomenon. This paper implements a dynamic model based on causal diagrams that seeks to identify the behavior and respective residential electricity consumption. The applied methodology involves random variables that seek to replicate the stochastic behavior inside a house during the day, the intention of applying a causal model lies in the interaction and dependence between variables, i. e. the conditioning that must exist between the execution of an activity and its consequent response, such criteria are reflected in the use of binary variables, which simulate on/off states as well as true/false Boolean states. For the simulation, we consider typical consumption values in household appliances for an average residence in the city of Quito, temporality of use and its probability of turning on under certain conditions. The simulation is executed in VENSIM, as it is a software designed to work with dynamic models. The results obtained establish that the proposed methodology presents 24.95% error with respect to real measurements.