Time-domain harmonic state estimation using filtered measurements based on Fourier transform

Power quality monitoring is required to provide an evaluation of the harmonic status of the power system. However, only a few buses can, with acceptable costs, be equipped with measuring devices. As a consequence, lack of information has to be covered by means of suitable harmonic state estimation techniques. A technique recently developed in the time-domain where Kalman filter has been used for this purpose; however, it requires the harmonic sources, the error covariance matrices associated with the measuring process, the plant model and an initial condition. Besides, it uses serial branch measurements, which in practical cases are not available. The proposed methodology relates the measurement vector with state variables in the time-domain where measurements at sending or receiving end of lines instead of serial branch measurements are used. In order to ensure the appropriate differentiation process needed by the time-domain solution, filtered noisy measurements based on Fourier transform are used. The proposed methodology is validated with a 5-bus test system including linear and nonlinear-balanced loads. A direct comparison of the state estimation response against the actual response obtained from the time-domain power system simulation is performed. Partial measurements have been used to evaluate the harmonic state of the power system. The obtained results indicate that the proposed methodology can adequately assess the harmonic state of the power network.