2019/08/21 by Nathaniel Tucker, Tucker, Nathaniel, Ahmadreza Moradipari +3
Decision Sciences · Engineering · #Advanced Bandit Algorithms Research #Energy Load and Power Forecasting #FOS: Electrical engineering #Smart Grid Energy Management #Systems and Control (eess.SY) #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.1908.07964
openalex publication_date 2019/08/21 · openalex created_date 2022/07/28 · openalex updated_date 2026/07/28
We consider the problem of an aggregator attempting to learn customers' load\nflexibility models while implementing a load shaping program by means of\nbroadcasting daily dispatch signals. We adopt a multi-armed bandit formulation\nto account for the stochastic and unknown nature of customers' responses to\ndispatch signals. We propose a constrained Thompson sampling heuristic,\nCon-TS-RTP, that accounts for various possible aggregator objectives (e.g., to\nreduce demand at peak hours, integrate more intermittent renewable generation,\ntrack a desired daily load profile, etc) and takes into account the operational\nconstraints of a distribution system to avoid potential grid failures as a\nresult of uncertainty in the customers' response. We provide a discussion on\nthe regret bounds for our algorithm as well as a discussion on the operational\nreliability of the distribution system's constraints being upheld throughout\nthe learning process.\n