Hung Y.-CMichailidis G.YING-CHAO HUNG2022-11-112022-11-11201919493053https://www.scopus.com/inward/record.uri?eid=2-s2.0-85049063073&doi=10.1109%2fTSG.2018.2850326&partnerID=40&md5=cc9ca0f01534434329c17743a6e81b9bhttps://scholars.lib.ntu.edu.tw/handle/123456789/625019Time-of-use (TOU) pricing is an important strategy for electricity providers to manage supply and make the grid more efficient; as well as for consumers seeking to manage their costs. In this paper, we discuss a general stochastic modeling framework for consumer's power demand based on which TOU contract characteristics can be selected to minimize the mean electricity price paid by the customer. We exploit the characteristics of power demand observed in real grids to model it during homogeneous peak periods as a constant level with fluctuations described by a scaled fractional Brownian motion. We analyze the exceedance process over pre-specified thresholds and use this information to formulate an optimization problem to determine the key features of the TOU contract. Due to the analytical intractability of certain expressions with the exception of short-range dependence fluctuations, the solution of the posited optimization problem requires using techniques such as Monte Carlo simulation and numerical search. The methodology for two pricing schemes is illustrated using real data. © 2010-2012 IEEE.contract capacity; fractional Brownian motion; mixed-integer programming; Monte Carlo simulation; self-similarity; Time-of-use pricing[SDGs]SDG7Brownian movement; Costs; Electric power transmission networks; Electric power utilization; Intelligent systems; Monte Carlo methods; Power markets; Stochastic systems; Electricity pricing; Fractional brownian motion; Mixed integer programming; Modeling and optimization; Optimization problems; Self-similarities; Short range dependences; Time-of-use pricing; Integer programmingjournal article10.1109/TSG.2018.28503262-s2.0-85049063073