Nova metodologia para o controle de intercâmbio de potência reativa

Abstract

Reactive Power Interchange Control (RPIC) is a methodology whose main objective is to maintain the reactive power interchange of an area at a specified value. Such a scheme has the potential to support the study of the reactive power support provided or required by a given area and its impact on the operation of the system. In addition, such control strategy was elaborated aiming at the decoupling of tension control and reactive power of the area, in order to allow greater efficiency of coordinated voltage control strategies. The basic mathematical formulation, which will be presented, was inspired by the method of control of active power interchange between areas, where multiple generators contribute to control the active power exchange in specified values, as described in [1]. The main contribution of this work is associated with a new augmented formulation, where only one additional equation is used to represent the reactive power exchange control between two areas. The methodology provides great flexibility, allowing not only the application in the RPIC, but in other to generally model control strategies, in which the contribution of more than one device for a given control strategy is required. For the application of this work, the modelling represents the various generators participating in the RPIC, in which only one additional line and column in the Jacobian matrix is required, being this formulation the main contribution of this work. In addition, the methodology presents great flexibility in the incorporation or removal of a control device, which has reached its corresponding limit. The proposed mathematical model was tested in several test systems, obtaining good results. The simulations were validated using the ANAREDE program owned by the Electric Energy Research Center (CEPEL). The power flow models and the main core of the proposed methodology were developed in MATLAB, aided by several programs for formulating and verifying the results. The main ones being: Mathematica, ANAREDE, PSIM, R language, Rstudio.