Research On Stand-alone Real-time GNSS Dynamic Instantaneous Velocimetry Technology

Instantaneous velocities are the important navigational parameter for different vehicles,whether on land or in the ocean,atmosphere or space,and it is also suitable for diverse applications,such as earthquake monitoring and early warning,airborne gravimetry,navigation,control and guidance,etc.The instantaneous velocities can be achieved using different technologies,different instruments or sensors.Compared to others,GNSS(Global Navigation Satellite Systems)velocimetry technology has been increasingly received due to its low cost,high accuracy and globally availability.With the demand of users for the accuracy,reliability and real-time performance of velocimetry,the existing velocimetry methods have been unable to meet the needs of users.Therefore,it is very important to develop relevant theories and techniques of real-time instantaneous velocity estimation based on the existing velocimetry methods,which has important theoretical and practical significance for complementing and perfecting the GNSS velocimetry theories and methods and improving the velocimetry accuracy.In this paper,velocimetry models are established by using TDCP or TDCP/Doppler measurements to obtain real-time and high-precision instantaneous velocities as the aim of the research.Meanwhile,the GNSS instantaneous velocimetry technology is studied in-depth and in-detail.The main research contents and achievements of this paper are as follows:1)According to the principle that the integral of velocity polynomial between adjacent epochs is the interval displacement,the single-channel instantaneous velocimetry coordinate domain model is presented by using the moving window technology.The proposed model has the characteristics of finite impulse response(FIR),which is essentially an FIR filtering problem.In the single-channel coordinate domain method,three independent FIR filters are used to solve the velocity of the three components separately,hence this kind of method is considered as the single-channel FIR filtering method.According to the different measurements,the single-channel TDCP coordinate domain algorithm and the single-channel TDCP/Doppler coordinate domain algorithm are derived respectively.To solve the difficulty of the inverse of a high-dimensional matrix in the process of modeling,an efficient Cholesky rank one updating algorithm is adopted,which reduces the computational amount,and makes the model more suitable for real-time application.Experimental results show that the proposed algorithm can effectively improve the velocimetry accuracy.Moreover,the stability of the algorithm can be effectively improved by combining the Doppler measurements.2)Using an FIR filter to solve all the three components of the 3D velocity vector simultaneously,the multi-channel TDCP/Doppler coordinate domain algorithm is presented.Compared with the single-channel coordinate domain algorithm,the multi-channel coordinate domain algorithm takes into account the cross-correlations among the three components,simplifies the computational complexity and improves the computational efficiency.The experiment results show that the velocimetry performance of the proposed algorithm is similar to that of the single-channel TDCP/Doppler coordinate domain algorithm.3)Based on the multi-channel TDCP/Doppler coordinate domain algorithm,a tighter multi-channel TDCP/Doppler observation domain algorithm is proposed by directly using the original measurements.The proposed algorithm can still provide a continuous instantaneous velocity solution for users in the scenario of short-time satellite signal loss,which is equivalent to the multi-channel TDCP/Doppler coordinate domain algorithm in the scenario of enough visible satellites.This algorithm has wider applicability than the multi-channel TDCP/Doppler coordinate domain algorithm.To deal with the potential cycle slips contained in TDCP,an observation domain algorithm with prediction and update steps similar to KF filtering is proposed in this paper.The results show that the proposed algorithm can enhance the processing capacities in the environment of short-time satellite signal loss,can detect and repair cycle slips with a high success rate.4)Combined TDCP/Doppler adaptive filtering models are designed,and the dynamic state-space model is used to characterize the uncertainty of the system state noise.Because the state noise predefined by the system is often inaccurate and time-varying,the adaptive filtering method is usually used to adjust it online.In this paper,we introduce two adaptive strategies,which are fading memory adaptive filtering and adaptive filtering based on variance component estimation.In order to refine the stochastic model of TDCP,a more accurate generalized stochastic model of TDCP is derived.The experimental results show that the adaptive filter model can effectively improve the accuracy and stability of the system.