Probabilistic analysis of masked loads with aggregated photovoltaic production

• Published in: Electric power systems research Vol. 189; p. 106670
• Main Authors: Liu, Shaohui, Maldonado, Daniel Adrian, Constantinescu, Emil M.
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 01.12.2020; Elsevier Science Ltd; Elsevier
• Subjects: Demand; Differential equations; Distributed PV system; Irradiance; Photovoltaic cells; Probabilistic analysis; Probability distribution; Probability theory; Real-time prediction; SDE; SOLAR ENERGY; spatio-temporal model; Spatiotemporal model; Stochastic models; Stochastic processes; Distributed PV system; SDE; Spatiotemporal model; Real-time prediction
• ISSN: 0378-7796; 1873-2046
• DOI: 10.1016/j.epsr.2020.106670

•System operators cannot observe behind-the-meter solar generation.•Load demand can be characterized as an stochastic Ornstein-Uhlenbeck process.•We model irradiance and solar generation with Gaussian processes.•With PMUs we estimate statistics of these processes.•Bayesian statistics are used to estimate and predict unobservable solar generation. In this paper we present a probabilistic analysis framework to estimate behind-the-meter photovoltaic generation in real time. We develop a forward model consisting of a spatiotemporal stochastic process that represents the photovoltaic generation and a stochastic differential equation with jumps that represents the demand. We employ this model to disaggregate the behind-the-meter photovoltaic generation using net load and irradiance measurements.