Resumen:
The rollout of smart metering infrastructure brings new opportunities to better monitor and model low voltage systems. This paper presents a novel methodology to estimate the sensitivities of voltages to power injections using smart meter data. Unlike previous works, this paper focuses on how some highly correlated active and reactive power measurements affect the estimated parameters. Moreover, a clustering-based algorithm is proposed as a viable alternative when the number of unknown parameters needs to be reduced, without compromising the accuracy of the estimations. Power flow calculations of the IEEE European test system are performed to simulate the smart meter information. The tests demonstrate the robustness of the proposed algorithms with noisy and correlated measurements.
