Frequency stability-based controlled islanding scheme based on clustering algorithm and electrical distance using real-time dynamic criteria from WAMS data

Abstract This paper presents online criteria of frequency stability-based controlled islanding scheme based on the density-based clustering algorithm and the system electrical distance from wide-area measuring system (WAMS) data. For this issue, at each time window, based on evaluating the correlation coefficient criteria between all two pairs of synchronous generators (SGs), the possible coherent generator groups are identified. Next, by using clustering algorithm, the identified coherent SGs are clustered and sorted through online working mode. In this case, an online index is proposed which and developed index in the form of mixed-integer linear programming (MILP) which through real-time evaluations, the system frequency stability is evaluated. In the use of MILP processes, the network electrical distance is considered as the main objective function (OF) which by solving the proposed OF through minimum criteria, proper islanding locations are identified. Finally, at the post-islanding period, a nonlinear programming (NLP) procedure is performed trough each individual islands where corresponding proper operational feasibilities are satisfied. In this case of transient stability criteria, based on the coherent SGs located at each island, the stability criteria is preserved through coherent groups as the core of islands. Also, by allocating proper load buses based on the SGs governor capacities, the islands frequency stabilities are controlled through acceptable regions. The effectiveness of the proposed scheme is carried through two IEEE 39-bus test system and Iran practical power grid. Results present the effectiveness of the proposed approach through different islanding conditions.