Data-driven approach to fuzzy inference system tuning for static security assessment of multi-area power systems with renewable energy

Abstract

Energy market deregulation and increased integration of intermittent renewable energy resources (RE) have pushed modern power systems to operate close to their security limits, which if violated could be catastrophic. To prevent power system failure and damage, a fast and accurate security assessment tool is required. In this paper, a data-driven fuzzy inference system tuned by hybrid genetic-simulated annealing (HGSA-FIS) is proposed for quick static security assessment (SSA) of a power system under a single component failure. An existing composite security index is modified to make a reliable SSA by distinguishing between severe and non-severe operating scenarios and contingencies. The prediction performance of HGSA-FIS was evaluated on an RTS-GMLC test system with operating scenarios varying with RE generation, load demand, and tie-line flows. The findings demonstrate the robust performance of HGSA-FIS in making static security assessments under varying operating scenarios, exceeding ANFIS in most and ANN in a few scenarios.