Joint Estimation Of Frequency And Initial Phase Of Components In Short Signal

The estimation of frequency and initial phase of signal components is a basic task in signal processing.And it is widely used in many fields such as power systems,radars,biomedical equipment,and scientific instrument measurement.In reality,there are many scenarios with high real-time requirements.In these scenarios,the signal components to be estimated are often short signals.We hope to get the accurate estimation of frequency and initial phase based on the short signals of two or three cycles or even shorter,and it is a big challenge.At the same time,most of the existing methods of signal parameter estimation are to estimate one of the frequency and phase first and then estimate the other on the basis of the former estimation,or to estimate the two separately,independent of each other,so it is not easy to obtain the most accurate estimation.This article focuses on the joint estimation of frequency and initial phase of components in short signal.The main work is as follows:(1)A joint estimator of frequency and initial phase of signal components based on spectral comparison is proposed.A series of sinusoidal signals are constructed with frequency and initial phase as variables,and the two-dimensional parameter optimization of frequency and initial phase is carried out with the optimization goal of the maximum similarity or the minimum difference between the constructed signal and the signal to be estimated.In this way,the parameter estimation problem is transformed into an optimization problem.Using the conjugate gradient algorithm and the 0.618 algorithm to solve the optimization problem,the final optimization results are the estimated value of frequency and initial phase.A large number of simulation experiments show that the estimation accuracy of this method is better than that of other existing frequency-domain parameter estimation methods,whether it is frequency estimation or initial phase estimation,and it can better approach the theoretical lower limit Cramer-Rao Lower Bound(CRLB)of signal parameter estimation at any signalto-noise ratio level.(2)A joint estimator of frequency and phase of signal components based on waveform comparison in time-domain is proposed.A large number of simulation experiments have verified that under the same conditions,both frequency and initial phase estimation can achieve the same accuracy as the frequency-domain joint estimation,and is better than other non-frequency-domain parameter estimation methods.At the same time,analysis and experiments show that the time-domain joint estimation is better than the frequency-domain joint estimation in anti-interference and computational complexity.