2017/10/11 by Joseph L. Durante, Juliana Nascimento, Durante, Joseph L. +3
Engineering · #Electric Power System Optimization #FOS: Mathematics #Microgrid Control and Optimization #Optimization and Control (math.OC) #Smart Grid Energy Management
paper · pdf · doi:10.48550/arxiv.1710.03914
openalex publication_date 2017/10/11 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
We consider an energy storage problem involving a wind farm with a forecasted\npower output, a stochastic load, an energy storage device, and a connection to\nthe larger power grid with stochastic prices. Electricity prices and wind power\nforecast errors are modeled using a novel hidden semi-Markov model that\naccurately replicates not just the distribution of the errors, but also\ncrossing times, capturing the amount of time each process stays above or below\nsome benchmark such as the forecast. This is an important property of\nstochastic processes involved in storage problems. We show that we achieve more\nrobust solutions using this model than when more common stochastic models are\nconsidered. The new model introduces some additional complexity to the problem\nas its information states are partially hidden, forming a partially observable\nMarkov decision process. We derive a near-optimal time-dependent policy using\nbackward approximate dynamic programming, which overcomes the computational\nhurdles of classical (exact) backward dynamic programming, with higher quality\nsolutions than the more familiar forward approximate dynamic programming\nmethods.\n