Implementing a flexible method based on polynomial fit for harmonic analysis

A new and computationally convenient method is used to design the Taylor-Fourier harmonic estimator. This estimator get best approximations (in comparison with that of the FFT) of the harmonics components of a signal under transient conditions. The estimator can be seen as a FIR symmetric filter bank. Each filter of the bank meets maximally flat conditions around the harmonic frequencies. This conditions can be used to solve the least squares problem instead the classical pseudoinverse solution. The maximally flat conditions also ensure the quality of the estimate, as they make that the gains of the filters correspond to the ideal differentiator gains around the harmonic frequencies. An implementation of the proposed estimator using Labview and the cRIO platform is shown. Several experimental tests are used to compare the performance of the estimator with that of the Fourier analysis.