Uma proposta metodológica não invasiva para o compartilhamento da responsabilidade sobre desequilíbrios de tensão baseada na minimização da correlação das fontes

Abstract

Electrical power systems have undergone substantial transformations due to the incorporation of new technologies, bringing significant impacts on power quality, notably the voltage imbalance arising from operational conditions imposed by modern power grids. In fact, this indicator and others associated with the loss of power quality have grown and, for many installations, reach proportions exceeding the limits established by regulatory bodies. Such violations represent sources of harmful impact for the agents involved, who may suffer various losses, such as a decrease in network operational efficiency, reduction in the useful life of installation components, and untimely supply interruptions, among others. In these circumstances, when such regulatory standard violations occur, mitigation alternatives emerge that represent high financial costs; recognizing that connection buses — the Points of Common Coupling — unite two or more agents, this can lead to conflicts between the parties. Focusing on the search for means of responsibility sharing regarding the final effects, the literature presents a vast bibliography that, however, still lacks non-invasive analysis methods with consolidated physical and application properties for practical field use. Considering this fact, this thesis proposes the development of the Correlation Minimization Method, a new non-invasive strategy for the sharing of responsibility over total voltage unbalances measured at the PCC that offers reliability, robustness, and practical applicability. The methodology is based on the statistical independence of the fast-varying components of the agents' current sources and stands out for not requiring physical interventions in the network for the analysis process. Besides establishing the theoretical foundations of this new strategy, the work evaluates its performance comparatively with the main methodologies in literature, such as Least Squares and Complex Independent Component Analysis Method, demonstrating, through computational analysis and experimental validations in a real facility, its superiority in terms of reliability, robustness, and operational practicality.