Vulnerability assessment of distributed load shedding algorithm for active distribution power system under denial of service attack

Abstract

In order to deal with frequency deviation and supply-demand imbalance in active distribution power system, in this paper a distributed under frequency load shedding (UFLS) strategy is proposed. Different from conventional centralized UFLS schemes, no centralized master station gathering all the buses' information is required. Instead, each bus decides its own load shedding amount by only relying on limited peer-to-peer communication. However, such UFLS strategy may suffer from some unexpected cyber-attacks such as integrity attacks and denial of service (DoS) attack. The latter DoS attack aims to degrade the system performance by jamming or breaking the communication, which is of high probability to happen in practical power system. To assess the vulnerability of proposed distributed UFLS algorithm, the effect of DoS attack on distributed average consensus algorithm is theoretically derived, which indicates that the final consensus value can be estimated by a given attack probability. It is also investigated that such DoS attack does harm to the load shedding amount and finally affects the system frequency performance in the active distribution power system. Several case studies implemented on an IEEE 33-bus active distribution power system are conducted to verify the effectiveness of the theoretical findings and investigate the vulnerability of the considered power system.