A high‐precision frequency estimation algorithm based on multi‐segment signal fusion

As the current multi-segment signal fusion algorithms cannot meet requirement of high-precision frequency estimation at different signal-to-noise ratios (SNRs), an adaptive algorithm is proposed in this paper. At high SNR, autocorrelation weighted utilised, which adopts real-time signals, resulting denoised sinusoidal signals. Then obtained by minimising error function based on coarse estimation. low improved spectrum adopted, uses three-parameter sine-fitting method to determine phase difference compensation factor, and obtain a long with continuous phase. result, can be estimated spectral peak search. Simulation results show that algorithm, compared existing algorithms, not only improves accuracy about 20%–50% but also perform for signals lengths, frequencies, abnormal situations.