@Article{cmc.2025.066097,
AUTHOR = {Yunhao Yu, Meiling Dizha, Boda Zhang, Ruibin Wen, Fuhua Luo, Xiang Guo, Junjie Song, Bingdong Wang, Zhenyong Zhang},
TITLE = {DH-LDA: A Deeply Hidden Load Data Attack on Electricity Market of Smart Grid},
JOURNAL = {Computers, Materials \& Continua},
VOLUME = {85},
YEAR = {2025},
NUMBER = {2},
PAGES = {3861--3877},
URL = {http://www.techscience.com/cmc/v85n2/63791},
ISSN = {1546-2226},
ABSTRACT = {The load profile is a key characteristic of the power grid and lies at the basis for the power flow control and generation scheduling. However, due to the wide adoption of internet-of-things (IoT)-based metering infrastructure, the cyber vulnerability of load meters has attracted the adversary’s great attention. In this paper, we investigate the vulnerability of manipulating the nodal prices by injecting false load data into the meter measurements. By taking advantage of the changing properties of real-world load profile, we propose a deeply hidden load data attack (i.e., DH-LDA) that can evade bad data detection, clustering-based detection, and price anomaly detection. The main contributions of this work are as follows: (i) We design a stealthy attack framework that exploits historical load patterns to generate load data with minimal statistical deviation from normal measurements, thereby maximizing concealment; (ii) We identify the optimal time window for data injection to ensure that the altered nodal prices follow natural fluctuations, enhancing the undetectability of the attack in real-time market operations; (iii) We develop a resilience evaluation metric and formulate an optimization-based approach to quantify the electricity market’s robustness against DH-LDAs. Our experiments show that the adversary can gain profits from the electricity market while remaining undetected.},
DOI = {10.32604/cmc.2025.066097}
}