Research On Frequency Estimation Algorithms Of Real Sinusoidal Signals Based On DFT

Frequency estimation of real sinusoidal signals is widely used in various engineering and technical fields,including radar systems which can be used in military applications and in the production and life of people and their important power systems,traditional metal processing systems,communication systems and voice systems which have developed rapidly in recent years.The parameters estimation of damped sinusoidal signal is applied to mechanical spectrometer,proton magnetometer and other instruments to improve measurement accuracy.DFT has been widely used in frequency estimation due to its low computational complexity.A large number of frequency estimation algorithms for sinusoidal signals based on DFT have been proposed.However,when DFT is applied to real sinusoidal signals,there will be picket fence effect and spectrum leakage from negative frequency.Most existing algorithms cannot eliminate the influence of spectrum leakage completely.Therefore,when the positive and negative frequency distribution is close,that is,when the number of acquired sine-wave cycles is small,the estimation accuracy of DFT will be reduced sharply.A novel high-precision interpolated DFT(IpDFT)based frequency estimation algorithm of real sinusoidal signals is proposed in this paper.The algorithm solves the picket fence effect by IpDFT,and the spectral superposition between positive and negative frequency is considered and calculated to eliminate the spectrum leakage from negative frequency.The theoretical performance of the algorithm is analyzed in the case of additive white Gaussian noise.The simulation results match well the theoretical values.In addition,computer simulation results show that the algorithm improves the accuracy and stability of frequency estimation,and has good noise suppression capability,especially when the distribution of positive and negative frequency is close.At the same time,accuracy of the proposed estimator when there is harmonics interference is analyzed by simulation.Two new and high-precision IpDFT based parameters estimators are proposed for damped sinusoidal signals.The two algorithms take both positive and negative frequencies into account in order to completely eliminate the influence of spectrum leakage from negative frequency.The simulation results show that,the proposed algorithms improve the estimation accuracy of frequency and damping factor compared with other algorithms,and the stability and anti-noise performance are better.In addition,when the frequency changes with time,the simulation results show that the proposed algorithm can achieve frequency tracking.In a word,the IpDFT based frequency estimation algorithm of real sinusoidal signal and the IpDFT based parameter estimators of damped sinusoidal signal are proposed in this paper,which solve the problem of negative frequency spectrum leakage.They improve the estimation accuracy and stability,especially when the number of acquired sine-wave cycles is small,compared with other existing algorithms.