Data Clustering-Driven Fuzzy Inference System-Based Optimal Power Flow Analysis in Electric Networks Integrating Wind Energy

Abstract

The development of smart grids has led to an increased focus by transmission and distribution network operators on the Optimal Power Flow (OPF) problem. The solutions identified for an OPF problem are vital to ensure the real-time optimal control and operation of electric networks and can help enhance their efficiency. In this context, this paper proposed an original solution to the OPF problem, represented by optimal voltage control in electric networks integrating wind farms. Based on a fuzzy inference system (FIS) built in the Fuzzy Logic Designer of the Matlab environment, where the fuzzification process was improved through fuzzy K-means clustering, two approaches were developed, representing novel tools for OPF analysis. The decision-maker can use these two approaches only successively. The FIS-based first approach considers the load requested at the PQ-type buses and the powers injected by the wind farms as the fuzzy input variables. Based on the fuzzy inference rules, the FIS determines the suitable tap positions for power transformers to minimise active power losses. The second approach (I-FIS), representing an improved variant of FIS, calculates the steady-state regime to determine power losses based on the suitable tap positions for power transformers, as determined with FIS. A real 10-bus network integrating two wind farms was used to test the two proposed approaches, considering comprehensive characteristic three-day tests to thoroughly highlight the performance under different injection active power profiles of the wind farms. The results obtained were compared with those of the best methods in constrained nonlinear mathematical programming used in OPF analysis, specifically sequential quadratic programming (SQP). The errors calculated throughout the analysis interval between the SQP-based approach, considered as the reference, and the FIS and I-FIS-based approaches were 5.72% and 2.41% for the first day, 1.07% and 1.19% for the second day, and 1.61% and 1.33% for the third day. The impact of the OPF, assessed by calculating the efficiency of the electric network, revealed average percentage errors between 0.04% and 0.06% for the FIS-based approach and 0.01% for the I-FIS-based approach.