Font Size: a A A

Research On Time-Frequency Domain Based High Dynamic Gnss Signal Precision Tracking Technology

Posted on:2018-11-28Degree:DoctorType:Dissertation
Country:ChinaCandidate:X XiaFull Text:PDF
GTID:1368330590455301Subject:Navigation, Guidance and Control
Abstract/Summary:PDF Full Text Request
The high dynamic tracking problem of a GNSS(Global Navigation Satellite System)signal is essentially a non-stationary signal parameter estimation problem.The frequency resolution and the time resolution are two basic performance requirements of this parameter estimation problem.But by the uncertainty principle,bandwidth and signal-to-noise ratio,there is a triangular trade-off between frequency resolution,time resolution and sensitivity.Therefore,the time-frequency analysis method is introduced into the signal processing field systematically in this article,and the following research has been done:1.An accurate and comprehensive mathematical description of the high dynamic GNSS signal has been set up by the time-frequency analysis theory.Limitations of the traditional tracing method,vector tracking(and strap down inertial navigation)and spectral estimation method in high dynamic tracking,has been analyzed.The root cause of these limitations comes from the uncertainty principle.2.Based on the time-frequency representations,a time-frequency image based parameter estimation method and a short-time fractional Fourier transform based parameter estimation method has been proposed.The corresponding uncertainty of these two methods has been analyzed in this paper.They both decrease the Cramér-Rao lower bound by increasing the number of observations and change observed variable respectively,thus improving the signal parameters estimation precision.3.The time-frequency domain based GNSS signal tracking loops has been designed.The various properties of tracking loops have been analyzed in detail,including the unbiasedness,gain,error,dynamic performance,etc.The static experiments have confirmed the feasibility of the time-frequency domain based tracking loop.They have also compared the advantages and disadvantages between the time-frequency domain tracking method,the traditional tracing method,the vector method and the spectrum estimation method.The performance of time-frequency domain tracking loop has been studied through dynamic simulation.4.Taking the time-frequency image based parameter estimation method as the foundation,the feasibility of the vehicle speed and acceleration estimation method via GNSS signal has been discussed.This method is not affected by the interference of gravitational field changing and can directly estimate the instantaneous value of vehicle speed and acceleration.Its performance,the advantages and disadvantages and application prospect have been discussed.5.A time-frequency domain BD signal positioning receiver has been designed and manufactured.It can track the BD signal in high dynamic environment,based on the short-time fractional Fourier transform.The simulation and the experiment show that it works reliable and stable.Due to the fact that the energy of the signal in the time-frequency domain is concentrated,the uncertainty of the signal can be reduced.Consequently,high dynamics no longer reduces the processing gain.On the other hand it makes the instantaneous frequency estimation more accurate.Experiment and simulation show that with a time resolution of 1 ms,high dynamics barely affect the tracking accuracy and processing gain.Therefore,the time-frequency domain tracking method does not only balance the trade-off between the dynamic performance and tracking accuracy,but also balance the trade-off between the sensitivity and dynamic performance.Because the acceleration estimation range only depends on the correlation integral time and the length of the signal and Kalman tracking filter can automatically adjust tracking bandwidth according to the acceleration,the time-frequency domain GNSS signal tracking method achieved better tracking performance than other methods,as high as hundreds of times the acceleration of gravity.What's more,the time-frequency domain GNSS signal parameter estimation method can be used for motion estimation,which expands the application of this method.Without complicated loop filter design,and without external inertial navigation,the tracking loop design of time-frequency domain GNSS signal tracking method is simpler than other methods.
Keywords/Search Tags:Time-frequency analysis, High dynamic, GNSS, Beidou, GPS, Signal tracking, Uncertainty principle, Instantaneous frequency, Parameter estimation, Wigner-Ville distribution, Fractional Fourier transform, Spectral estimation, Doppler frequency
PDF Full Text Request
Related items