No abstract

No abstract

No abstract

No abstract

No abstract

Estimating grid admittance is essential for assessing impedance stability and for designing adaptive controllers for distributed generation (DG) units. This paper proposes a new multivariable grid admittance identification algorithm that involves adaptive model order selection as an ancillary function within inverter-based DG controllers. Cross-coupling between d-and q-axis grid admittances necessitates multivariable estimation. To ensure persistence of excitation for grid admittance, sensitivity analysis is first employed in order to determine the injection of controlled voltage pulses by the DG. Grid admittance is then estimated from the processing of the extracted grid dynamics by the refined instrumental variable method for continuous-time system identification (RIVC) algorithm. The theoretical background underlying the RIVC algorithm is introduced, along with its integration within the proposed method for adaptive model order selection. Unlike nonparametric identification algorithms, the proposed RIVC algorithm provides a parametric multivariable model of grid admittance, which is essential for designing DG adaptive controllers. A hardware-in-the-loop application using OPAL-RT real-time simulators has been used to validate the proposed algorithm for both grid-connected and isolated active distribution networks.Index Terms-Distributed generation (DG), impedance stability, multivariable grid admittance identification, refined instrumental variable method for continuous-time system identification (RIVC).

This paper proposes multiagent supervisory control for precise power management in isolated dc microgrids. Two power management aspects are considered: 1) equal power sharing, which is realized via a proposed distributed equal power sharing algorithm; and 2) optimal power dispatch, which is achieved through a proposed distributed equal incremental cost (DEIC) algorithm. Both algorithms offer the additional advantage of the ability to restore the average system voltage to its nominal value. The proposed algorithms are based on the application of the average consensus theory along with voltage sensitivity analysis. Each distributed generation (DG) unit exchanges information with its neighbors, thus locally updating its no-load voltage setting to achieve the supervisory control objectives. The incorporation of DG droop-based control renders the proposed algorithms fully distributed with a reduced number of agents. The stability of the proposed algorithms is addressed, as well as the convergence of the proposed DEIC algorithm. Real-time OPAL-RT simulations demonstrate the effectiveness of the proposed algorithms in a hardware-in-the-loop application.

No abstract