Probabilistic modeling of smart residential energy systems

Negative effects of massive industrialization, high rates of fossil-fuel consumption, fast economic growing and technological development have positioned renewable energies as a promising manner to reach an environmentally sustainable society. Detailed knowledge about renewable resources is an important factor, but it is difficult to obtain in most places; in the case of solar and wind resources, energetic potential could vary significantly according to the local conditions. Implementation of Measure Correlate-Predict (MCP) methodology offers a partial solution to this problem; however, the associated error related to the extrapolation process could be in some cases significant. Hence, this paper presents an analytical method to incorporate MCP extrapolation error on the simulation of smart residential energy systems. Beta probability distribution function (PDF) is used to model the extrapolation error and it is combined with a simulation model to estimate PDF of renewable power generation, battery state of charge, and power imported from the distribution system, which allows obtaining a complete perspective of energy system performance.