XI Lei, PENG Dianming, CAO Wei, et al. FDIA Location Detection for Data-driven Algorithms in Cyber-physical Power Systems[J]. 2025, 45(18): 7110-7122.DOI: 10.13334/j.0258-8013.pcsee.240412.
False data injection attacks (FDIA) seriously threaten the security of cyber-physical power systems (CPPS). For the problem that the traditional detection methods are unable to identify attacks accurately and locate the attacked buses quickly
a FDIA localization method is proposed based on data-driven algorithms in CPPS. Firstly
the kernel extreme learning machine is combined with the autoencoder to form a multi-layer kernel extreme learning machine
which learns the power measurement data layer by layer. Then
the Harris hawk optimization
which incorporates the tent mapping and golden sine strategy
is utilized to optimize the parameters of the multi-layer kernel extreme learning machine
in order to improve the optimization speed and convergence accuracy. Finally
the proposed algorithm is verified by simulation in IEEE-14 and IEEE-118 bus test systems. The results show that the proposed algorithm has better detection speed
accuracy
precision
recall
and F1-Sorce compared to other algorithms
and can locate the attacked buses quickly and accurately.