Research On LFM Signal Parameter Estimation Based On FRFT

With the development of signal research field,researchers find that most of the signals in practical application are non-stationary signals,such as biomedical signal,speech signal,radar and sonar signal,seismic wave signal,wireless communication signal and so on.The non-stationary signals have the performance of time-frequency localization and time-frequency coupling,so it is widely used in radar,communication,biomedicine,geologic examination and so on.The parameters of LFM signal contain the position information of the target.Therefore,the parameters of LFM signal are detected and estimated effectively has important theoretical and practical significance for many fields.Fractional Fourier transform can give the fractional domain information between time domain and frequency domain,and it will not be disturbed by cross-term when processing multi-component LFM signals.Therefore,FRFT is widely used in LFM signal detection,parameter estimation,denoising and other branches.At present,the commonly used LFM signal parameter estimation methods mostly start with mathematical model and search method,which have the problems of poor noise robustness and low search efficiency.Aiming at above shortcomings,two methods for parameter estimation of LFM signal are proposed in thesis.The main work as follows:1)LFM signal parameter estimation method based on FRFT andl~p-norm.Firstly,thel~p-norm of LFM signal fractional spectrum is calculated,and thel~p-norm as optimization model in parameter estimation.Then,the model is optimized by Harris Hawks Optimization algorithm.Simulation results show that,thel~p-norm method has high noise robustness and parameter estimation stability.Finally,The curvature radius of lens is measured byl~p-norm method.2)LFM signal parameter estimation method based on FRFT and Rényi entropy.Firstly,the Rényi entropy distribution of LFM signal fractional spectrum is established,and based on above the optimization model of parameter estimation is established.Then,the model is searched by Harris Hawks Optimization algorithm.Simulation results show that,the parameter estimation performance of this method is stable,the noise is robust.Finally,The Rényi entropy method is applied in the separation of mixed interference fringes,and the cross fringes is separated with high precision.The research results of thesis not only enrich the theoretical system of LFM signal processing,but also expand the application of FRFT in the field of optics.