Search

CN-121995421-A - DPE/INS quasi-deep combination method and system in GNSS navigation domain

CN121995421ACN 121995421 ACN121995421 ACN 121995421ACN-121995421-A

Abstract

A method for combining DPE/INS quasi-deep in GNSS navigation domain includes collecting GNSS digital intermediate frequency signal by hardware platform radio frequency front end of GNSS receiver, inputting local signal and digital intermediate frequency signal into correlator after determining parameter searching range of local copy signal, obtaining PVT estimation result and covariance of DPE receiver by maximum likelihood estimation, calculating PVT estimation result and covariance of INS at current moment by inertial navigation calculation, obtaining PVT estimation result and covariance of DPE receiver and INS combined system by combining and filtering to determine parameter searching center, searching range and searching step length of local copy signal of DPE receiver, and further determining a series of local copy signal. The invention utilizes the characteristic that INS is not interfered by environment to determine the function of INS sensor in the navigation domain of DPE receiver, establishes the interaction mechanism of INS auxiliary information in the relevant domain and the navigation domain, and names the relevant domain as quasi-deep combination according to the characteristics of the architecture, thereby effectively helping the DPE receiver to reduce the searching range and improving the precision.

Inventors

  • YANG RONG
  • GAO WEI
  • HUANG JIHONG
  • WANG YINGDONG
  • ZHAN XINGQUN

Assignees

  • 上海交通大学

Dates

Publication Date
20260508
Application Date
20241101

Claims (10)

  1. 1. A DPE/INS quasi-deep combination method in a GNSS navigation domain is characterized in that a hardware platform radio frequency front end of a GNSS receiver is used for processing and collecting GNSS digital intermediate frequency signals, after a parameter search range of a local copy signal is determined, the local signal and the digital intermediate frequency signals are input into a correlator, PVT estimation results and covariance of the DPE receiver are obtained through maximum likelihood estimation, PVT estimation results and covariance of the INS at the current moment are calculated through inertial navigation calculation, PVT estimation results and covariance of the DPE receiver and the INS are obtained through combination filtering, and the PVT estimation results and covariance of the DPE/INS combination system are used for determining a parameter search center, a search range and a search step length of the local copy signal of the DPE receiver, and a series of local copy signals are further determined.
  2. 2. The DPE/INS quasi-deep combination method in GNSS navigation domain according to claim 1, wherein the radio frequency front-end processing means that signals of all visible GNSS satellites are received through a GNSS antenna, filtered and amplified by a pre-filter and a pre-amplifier, mixed with a sine wave local oscillation signal generated by a local oscillator to be down-converted into an intermediate frequency signal, and finally converted into a discrete-time digital intermediate frequency signal through an analog-to-digital converter.
  3. 3. The DPE/INS quasi-deep combining method in the GNSS navigation domain of claim 1, wherein the locally replicated signals are periodic oscillating signals including pseudo codes and carriers by selecting candidate Doppler shift and code phase parameters.
  4. 4. The DPE/INS quasi-deep combining method in GNSS navigation domain as claimed in claim 1, wherein said determining the parameter search range of the local replica signal comprises Wherein PVT parameters to be estimated in a DPE receiver The middle position vector is r , the clock difference is δt, the speed vector is v and the speed vector is Zhong Piao Each dimension corresponds to gamma 1 ～γ 8 ;S i , is a search space corresponding to each PVT parameter dimension, α is a search center of the parameter, is determined by INS calculation results, n is the number of search grids, beta is a search step length, and the gamma 1 ～γ 8 ;S i and the gamma are determined by covariance calculated by the INS.
  5. 5. The DPE/INS quasi-deep combination method in the GNSS navigation domain according to claim 1, wherein the maximum likelihood estimation is to determine PVT candidate parameters determined by the local replica signal corresponding to the maximum correlation value as the estimation result of PVT parameters according to the correlation result of the digital intermediate frequency signal and the local replica signal by taking the correlation value as the maximum basis, specifically: Wherein, superscript-PVT solution as candidate, superscript-estimation result of DPE receiver, And M is the total number of visible satellites, and is the correlation value of the local signal and the digital intermediate frequency signal of the m th satellite under the current candidate solution.
  6. 6. The DPE/INS quasi-deep combination method in the GNSS navigation domain according to claim 1, wherein the inertial navigation calculation is to determine the PVT estimated value of the next moment by kinematic relation integration accumulation according to the PVT estimated value of the current moment and the specific force and angular velocity information output by the INS measuring unit, and specifically comprises the following steps: Wherein: And The specific force and angular velocity values output by INS, And The rotational angular velocity of the earth and the gravitational acceleration of the earth are respectively, The carrier pose described for the quaternion, And v is the carrier position and carrier velocity respectively, The upper label is a coordinate transfer matrix from the body coordinate system b system to the ECEF coordinate system e system, the three-dimensional vector is converted into a corresponding quaternion, the upper label is a value related to inertial navigation solution, and the lower label k is a time epoch.
  7. 7. The DPE/INS quasi-deep combination method in the GNSS navigation domain according to claim 1, wherein the combined filtering means for determining the estimated value and the covariance after the multi-sensor fusion by using the Kalman filtering method according to the estimated values of different sensors and the corresponding estimated covariance comprises the following steps: K k+N ＝P k+N/k H T (HP k+N/k H T +R k+N ) -1 ,P k+N ＝(I-K k+N H)P k+N/k , The state quantity x is defined as PVT parameter Γ and expands the vector attitude q, the quantity measurement z is the difference between PVT estimated by the DPE receiver and INS, the superscript is the state estimated value output by the combined system, the INS is the state value estimated by integral iteration, the subscript k+N/k is the transition matrix or the prediction covariance from k time to k+N time, the matrices phi and H are the linearized and discretized filter state transition matrix and the measurement matrix respectively, the filter parameter P, Q, R is the estimated covariance, the process noise variance and the measurement noise variance of the current state respectively, and I is the identity matrix. The filter gain K for each epoch accounts for both the uncertainty of the DPE receiver estimate and the uncertainty of the INS estimate.
  8. 8. The DPE/INS quasi-deep combining method in a GNSS navigation domain according to claim 1, wherein the correlator mixes the digital intermediate frequency signal with a local replica signal, and correlates the mixed result with a locally replicated ranging code, so as to output a correlation result in a code phase and a doppler dimension, and the correlation result is specifically: Wherein τ and f d are the code phase and Doppler shift of the current satellite channel, respectively, superscript-is the candidate code phase or Doppler shift, And The equivalent code phase and doppler error for the current candidate PVT solution, For the correlation values of the local signal and the digital intermediate frequency signal of the m th satellite under the current candidate solution, the correlation value subscripts τ and f are the correlation value components contributed by the code phase and doppler, For noise in the correlation domain, sinc (x) =sin (x)/x is a sine function, and T c is the coherent integration time of the local signal and the digital intermediate frequency signal.
  9. 9. A DPE/INS navigation domain quasi-deep combination system for realizing the method of any one of claims 1-8 is characterized by comprising an INS resolving module, a DPE/INS combination module, a navigation domain module and a correlation domain module, wherein the INS resolving module carries out inertial navigation resolving on PVT parameter updating epoch t k ～t k+N of each DPE receiver, superimposes the IMU data during the inertial navigation resolving on an initial value to obtain PVT resolving results of the INS, outputs the INS resolving results to the DPE/INS combination module at a next DPE receiver parameter updating time point t k+N , the DPE/INS combination module obtains a candidate DPE noise variance and DPE estimation result obtained by the correlation domain module and an equivalent PVT solution and a corresponding covariance outputted by the INS resolving module, obtains a state metric and a covariance metric for compensating an estimated value through a combined navigation filter, takes the state metric outputting resolving module as an initial value, and the navigation domain takes the INS resolving results and the correlation domain as a correlation domain, and a candidate DPE noise variance and a correlation domain is obtained by taking the latest error of the DPE and a space of the INS and a correlation domain as a candidate DPE and a correlation domain, and a DPE error and a space is obtained by using a correlation domain, and a DPE error and a space is obtained by a fuzzy search result, and a DPE error is obtained by a correlation domain.
  10. 10. The DPE/INS navigation domain quasi-deep combination system of claim 9, wherein the initial value is obtained by correcting an INS estimation result output by the module through a state estimator in the DPE/INS combination module at a last DPE receiver parameter updating time point t k ; The inertial navigation resolving unit in the INS resolving module calculates to obtain a resolving result according to the initial value and IMU data; The combined filter unit in the DPE/INS combined module calculates to obtain covariance estimation and state estimation according to the DPE noise variance, the DPE estimation result and the INS estimation value in the navigation domain module, and then performs difference processing according to the state estimation and the INS estimation value to obtain an initial value in the INS calculation module; The navigation domain module directly obtains a search center alpha according to state estimation in the DPE/INS combination module, accumulates and divides grids according to covariance estimation and DPE noise variance in the DPE/INS combination module to obtain a search range n beta and a search step beta, and then jointly determines candidate PVT solutions through the search center alpha, the search range n beta and the search step beta; the search center alpha, the search range n beta, the search step beta and the candidate PVT solution jointly form PVT parameters; and the correlation domain module generates a code phase and a Doppler frequency shift according to the candidate PVT solution calculation in the navigation domain module, then assembles the code phase and the Doppler frequency shift to obtain a local replication signal, and carries out correlation calculation on the local replication signal and the digital intermediate frequency signal through a correlator to obtain maximum likelihood estimation for generating a DPE noise variance and a DPE estimation result in the navigation domain module.

Description

DPE/INS quasi-deep combination method and system in GNSS navigation domain Technical Field The invention relates to a technology in the field of satellite navigation, in particular to a Direct Position Estimation (DPE)/Inertial Navigation System (INS) quasi-deep combination method and system in the navigation domain of a global satellite navigation system (GNSS). Background Direct Position Estimation (DPE), which is one of new architectures for receiver design in global satellite navigation systems (GNSS) is distinguished from traditional acquisition tracking receiver architecture, aiming at establishing direct projection from the relevant domain to the navigation domain in the receiver, with a hopeful increase in accuracy and robustness. However, DPE receivers still face significant challenges in urban environments, such as multipath, non-line of sight (NLOS), and interference. To solve this problem, it may be combined with other navigation methods to achieve the complementary purpose of advantage, such as Inertial Navigation System (INS), and the operation of the DPE receiver in the relevant domain and navigation domain is different from that of the conventional receiver, but is related and interdependent, and the combination of the DPE receiver and INS corresponding to the operation is not related to the technology and implementation process. Disclosure of Invention Aiming at the problems that the existing single DPE receiver is difficult to deal with the blocking of GNSS signals caused by buildings, trees and the like in urban challenging environments, multipath effects caused by non-direct signals generated by reflection of GNSS signals on building walls and other areas, radio frequency interference on GNSS signals caused by artificial factors such as broadcast radio signals and the like, the invention provides a DPE/INS quasi-deep combination method and system in a GNSS navigation domain, which utilize the characteristic that INS is not interfered by the environment to determine the function of an INS sensor in the navigation domain of the DPE receiver, establish an interaction mechanism of INS auxiliary information in the relevant domain and the navigation domain, and name the interaction mechanism as quasi-deep combination according to the characteristics of architecture of the interaction mechanism, thereby effectively helping the DPE receiver to narrow the search range and improving the precision. The invention is realized by the following technical scheme: The invention relates to a DPE/INS quasi-deep combination method in a GNSS navigation domain, which comprises the steps of processing and collecting GNSS digital intermediate frequency signals through a hardware platform radio frequency front end of a GNSS receiver, inputting the local signals and the digital intermediate frequency signals into a correlator after determining a parameter search range of local copy signals, obtaining PVT estimation results and covariance of the DPE receiver through maximum likelihood estimation, calculating PVT estimation results and covariance of the INS at the current moment by utilizing inertial navigation calculation, obtaining PVT estimation results and covariance of the DPE receiver and the INS by combining and filtering, and determining a parameter search center, a search range and a search step length of the local copy signals of the DPE receiver. The radio frequency front-end processing means that signals of all visible GNSS satellites are received through a GNSS antenna, filtered and amplified by a pre-filter and a pre-amplifier, mixed with a sine wave local oscillation signal generated by a local oscillator to be down-converted into an intermediate frequency signal, and finally the intermediate frequency signal is converted into a discrete-time digital intermediate frequency signal through an analog-to-digital converter. The local replica signal is a periodic oscillation signal comprising pseudo codes and carriers is output by selecting candidate Doppler frequency shift and code phase parameters. The parameter searching range for determining the local copy signal is specifically that the whole candidate solution spaceWherein PVT parameters to be estimated in a DPE receiverThe middle position vector is r, the clock difference is δt, the speed vector is v and the speed vector is Zhong PiaoEach dimension corresponds to gamma 1～γ8;Si, is a search space corresponding to each PVT parameter dimension, α is a search center of the parameter, is determined by INS calculation results, n is the number of search grids, beta is a search step length, and the gamma 1～γ8;Si and the gamma are determined by covariance calculated by the INS. The maximum likelihood estimation is to determine that the estimated result of the PVT parameter is the PVT candidate parameter determined by the local copy signal corresponding to the maximum correlation value according to the correlation result of the digital intermediate frequency