TY - JOUR
T1 - Privacy-Aware Load Ensemble Control
T2 - A Linearly-Solvable MDP Approach
AU - Hassan, Ali
AU - Deka, Deepjyoti
AU - Dvorkin, Yury
N1 - Publisher Copyright:
© 2010-2012 IEEE.
PY - 2022/1/1
Y1 - 2022/1/1
N2 - Demand response (DR) programs engage distributed demand-side resources, e.g., controllable residential and commercial loads, in providing ancillary services for electric power systems. Ensembles of these resources can help reducing system load peaks and meeting operational limits by adjusting their electric power consumption. To equip utilities or load aggregators with adequate decision-support tools for ensemble dispatch, we develop a Markov Decision Process (MDP) approach to optimally control load ensembles in a privacy-preserving manner. To this end, the concept of differential privacy is internalized into the MDP routine to protect transition probabilities and, thus, privacy of DR participants. The proposed approach also provides a trade-off between solution optimality and privacy guarantees, and is analyzed using real-world data from DR events in the New York University microgrid in New York, NY.
AB - Demand response (DR) programs engage distributed demand-side resources, e.g., controllable residential and commercial loads, in providing ancillary services for electric power systems. Ensembles of these resources can help reducing system load peaks and meeting operational limits by adjusting their electric power consumption. To equip utilities or load aggregators with adequate decision-support tools for ensemble dispatch, we develop a Markov Decision Process (MDP) approach to optimally control load ensembles in a privacy-preserving manner. To this end, the concept of differential privacy is internalized into the MDP routine to protect transition probabilities and, thus, privacy of DR participants. The proposed approach also provides a trade-off between solution optimality and privacy guarantees, and is analyzed using real-world data from DR events in the New York University microgrid in New York, NY.
KW - Differential privacy
KW - Markov decision process~(MDP)
KW - cost of privacy
KW - dirichlet distribution
KW - linearly solvable MDP
UR - http://www.scopus.com/inward/record.url?scp=85115693362&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85115693362&partnerID=8YFLogxK
U2 - 10.1109/TSG.2021.3114370
DO - 10.1109/TSG.2021.3114370
M3 - Article
AN - SCOPUS:85115693362
SN - 1949-3053
VL - 13
SP - 255
EP - 267
JO - IEEE Transactions on Smart Grid
JF - IEEE Transactions on Smart Grid
IS - 1
ER -