Design Of Time-frequency Analysis Module For Mixed Domain Oscilloscope

With the rapid development of modern communication and integrated circuit technology,signal modulation and coding methods have become more complex and diverse,traditional analysis instruments cannot meet the requirements of accurate measurement of modern electrical signals.In order to detect and analyze signals with high bandwidth,fast transmission,multiple complex modulations and high transient rates from multiple angles,mixed-domain oscilloscopes have an important development position in the entire measurement industry.This thesis is based on the design and implementation of the time-frequency analysis module of the mixed domain oscilloscope.Three main aspects are studied with regard to the analysis of multi-domain time-dependent measurements based on digital down converters,time-frequency analysis algorithms and three-dimensional waveform drawing techniques.(1)A multi-domain time-correlated measurement analysis based on digital downconversion is carried out.First,the three modules of digital down-conversion,mixing,anti-aliasing filter and decimation generate mutually orthogonal in-phase signal and quadrature signal,and then based on the in-phase and quadrature baseband signal flow.The instantaneous frequency estimation algorithm based on the central finite difference method is used to obtain the relationship between amplitude,phase and time through calculations.Finally,the central finite difference method is used to calculate the instantaneous frequency,and the phase unwinding is used to solve the phase winding problem.(2)Two time-frequency analysis algorithms,short-time Fourier transform and wavelet transform are used to replace Fourier transform for frequency analysis of complex signals.Short-time Fourier transform is realized based on discrete time-domain signal,which is multiplied by the selected window sequence of the same length,and then the fast Fourier transform is carried out to get the STFT result.fast wavelet transform based on linear frequency modulation z-transform is proposed.When a is an integer,it can be calculated by DFT.When a is a fraction or irrational number,CZT is used to calculate the Fourier transform of the wavelet basis function.(3)The OpenGL-based 3D waveform drawing is realized.First,the data processed by the time-frequency analysis are transformed by view coordinates and opaque calculation to form specific graphics elements stored in memory.Then,the environment and other attribute variables are set to construct the final model to be displayed.Finally,the final 3D waveform display is realized by projection transformation,3D cropping and viewpoint transformation.Finally,the 3D waveform is displayed by projection transformation,3D cropping,and viewpoint transformation,and the edge of the image is softened by anti-aliasing algorithm.Finally,the function of the time-frequency analysis module implemented in this paper is verified on the built verification platform,and the correctness and effectiveness of the function of the time-frequency analysis module are finally verified.