Fast Ambiguity Resolution Of Medium And Long Baseline Based On Wavelet Transform

Integer ambiguity resolution is very important to achieve high-precision GNSS positioning.Fast ambiguity resolution under medium and long baselines is also the key problem and technical difficulty to achieve high-precision positioning.The appearance of tri-frequency observations is helpful to achieve quick solution of ambiguity.In this thesis,the corresponding research on the fast fixation of medium-long baseline ambiguity under GPS tri-frequency observation is carried out.Based on the measured GPS data,the fixed effect of traditional TCAR algorithm is analyzed,and an improved model is proposed by combining wavelet transform to improve the success rate of single epoch fixing of integer ambiguity.The main tasks as follows:(1)The domestic and international research status of medium and long baseline ambiguity resolution in recent years is analyzed.The theory of GNSS three-frequency combination observation and its error characteristics are discussed,and the error effects in the fast resolution of double-difference ambiguity under medium and long baselines are analyzed.The relationship between the number of lanes under combined observations and the wavelength and combined observations is analyzed.Under the constraints of ionospheric factor,noise factor and Ratio of wavelength to noise,the selection of the optimal combination coefficients for the combination of ultra-wide lane,wide lane and narrow lane is discussed.(2)The double-difference ionospheric delay under medium-long baseline is solved and its solution accuracy is analyzed.The denoising of ionospheric delay is carried out by using wavelet denoising.The quality evaluation index of wavelet denoising is introduced,and the suitable wavelet basis functions under different baseline lengths are obtained under the guidance of the quality evaluation index.(3)The optimal decomposition and reconstruction scale of wavelet denoising is analyzed by means of multi-indicator fusion.The characteristics of the optimal decomposition and reconstruction scales varying with the baseline length are discussed.Under the optimal denoising effect,accurate double-difference ionospheric delay information is obtained.Finally,the ionospheric information under different baselines is discussed.(4)The research on GF-TCAR,GIF-TCAR and GB-TCAR models has been carried out,and combined with wavelet denoising,the combination of very wide lane,wide lane and narrow lane ambiguity in each model has been improved.For the wide-lane ambiguity,the refined pseudo-pitch method is used to improve the solution success rate.A single epoch weak ionospheric K-value method is proposed to improve the resolution success rate of wide roadway ambiguity under long baseline based on the relationship between ionospheric delay and extracted noise information.On the basis of multi-index fusion denoising,the optimal scale denoising analysis of floating point solution of narrow roadway under GIF model is carried out.Finally,using the measured data to compare the traditional model and the improved model,the results show that each improved method has a certain improvement on the fixed ambiguity fixed success rate.The solution based on geometric model is better than the non-geometric model.The improved GB-TCAR has the best improvement effect in both wide lane and narrow lane combination,and the longer the baseline,the larger the ionospheric delay and the more obvious the improvement effect.