New dynamic filter based on least absolute value algorithm for on-line tracking of power system harmonics

The authors present the application of the technique of least absolute value (LAV) dynamic filtering to optimal tracking of power system harmonics. The proposed technique uses digitised samples of the voltage and current waveform at a power system bus, where a harmonics standard is contemplated. The proposed technique can easily handle time-varying harmonic parameters. Two models are developed and tested. In the first, the measurements matrix varies with time, with an identity transition matrix. In the second model, the state transition matrix is a function of the sampling rate and the number of harmonics chosen. The algorithm is tested using a simulated and actual recorded data set. A sample set of results is reported