Desempenho da Estimação de Estados em Redes de Distribuição de Energia: Modelo Linear e Mínimos Quadrados Ponderados

Abstract

In electrical energy distribution networks with radial topology, forward and backward sweeping methods stand out compared to traditional methods for solving power flow due to their simplicity and speed of convergence. The sweeping can thus be used in state estimation algorithms. First, a linear state estimation algorithm is proposed in which measurements are obtained by micro phasor measurement units present in modern distribution networks. The second algorithm is based on the formulation using the sum of powers in the lines. A biquadratic equation that preserves the direct and explicit relation between the magnitude of the nodal voltage and the active and reactive power flows. This solution is applied to weighted least squares state estimation due to its ease of correlation with electrical variables in real form without the need for phase angle. The proposed algorithms are developed and tested on a distributed network with 135 buses. In the comparative analysis, the error reduction in estimated states is verified in comparison with the generated pseudomeasures. The performance of the weighted least squares estimator is superior to the linear estimator.