DPE / INS quasi-depth combination method and system in GNSS navigation domain

Abstract

The invention discloses a DPE / INS quasi-deep combination method and system in a GNSS navigation domain, and the method comprises the steps: collecting a GNSS digital intermediate frequency signal through a hardware platform radio frequency front end of a GNSS receiver, inputting a local signal and the digital intermediate frequency signal into a correlator after determining a parameter search range of a local copied signal, and obtaining a PVT estimation result and covariance of a DPE receiver through maximum likelihood estimation; meanwhile, the PVT estimation result and covariance of the INS at the current moment are calculated through inertial navigation calculation, and the PVT estimation result and covariance of a DPE / INS combined system are obtained through combined filtering of the PVT estimation result and covariance of the DPE receiver and the INS and used for determining the parameter search center, the search range and the search step length of a local copy signal of the DPE receiver; a series of local replication signals are then determined. According to the method, the function of the INS sensor in the navigation domain of the DPE receiver is determined by utilizing the characteristic that the INS is not interfered by the environment, an interaction mechanism of the INS auxiliary information in the related domain and the navigation domain is established, and the interaction mechanism is named as a quasi-deep combination according to the characteristics of the architecture, so that the DPE receiver is effectively helped to narrow the search range and improve the precision.

Description

Technical Field

[0001] This invention relates to a technology in the field of satellite navigation, specifically a method and system for combining direct position estimation (DPE) and quasi-depth in the navigation domain of a Global Navigation Satellite System (GNSS). Background Technology

[0002] Direct Position Estimation (DPE), distinct from traditional acquisition-track receiver architectures, is a novel receiver design architecture in Global Navigation Satellite Systems (GNSS). It aims to establish a direct projection from the correlation domain to the navigation domain within the receiver, potentially improving accuracy and robustness. However, DPE receivers still face significant challenges in urban environments, such as multipath propagation, non-line-of-sight (NLOS), and interference. To address this, it can be combined with other navigation methods to achieve complementary advantages, such as Inertial Navigation Systems (INS). While the operation of DPE receivers in both the correlation and navigation domains differs from that of traditional receivers, these domains are interconnected and interdependent. Consequently, the combination of DPE receivers with INS lacks relevant technologies and implementation processes. Summary of the Invention

[0003] This invention addresses the challenges of existing single DPE receivers in handling GNSS signal obstruction caused by buildings and trees in challenging urban environments; multipath effects caused by non-direct signals reflected from building walls; and radio frequency interference caused by human factors such as broadcast radio signals. It proposes a quasi-depth combination method and system for DPE / INS in the GNSS navigation domain. Utilizing the environmentally unaffected nature of INS, the method determines the role of the INS sensor in the DPE receiver's navigation domain, establishes an interaction mechanism between the relevant domain and the navigation domain for INS auxiliary information, and names this method "quasi-depth combination" based on its architectural characteristics. This effectively helps the DPE receiver narrow its search range and improve accuracy.

[0004] This invention is achieved through the following technical solution:

[0005] This invention relates to a quasi-depth combination method for DPE / INS in the GNSS navigation domain. It involves acquiring GNSS digital intermediate frequency (IF) signals through the RF front-end processing of a GNSS receiver hardware platform. After determining the parameter search range of the locally replicated signal, the local signal and the IF signal are input to a correlator. Maximum likelihood estimation is used to obtain the PVT estimation result and covariance of the DPE receiver. Simultaneously, inertial navigation system (INS) calculations are used to calculate the PVT estimation result and covariance of the INS at the current time. The PVT estimation results and covariance of the DPE receiver and INS are then combined and filtered to obtain the PVT estimation result and covariance of the DPE / INS combined system. This is used to determine the parameter search center, search range, and search step size of the locally replicated signal of the DPE receiver, thereby determining a series of locally replicated signals.

[0006] The aforementioned radio frequency front-end processing refers to: receiving signals from all visible GNSS satellites through a GNSS antenna, filtering and amplifying them through a pre-filter and a pre-amplifier, then mixing them with the sinusoidal local oscillator signal generated by the local oscillator to downconvert them into an intermediate frequency signal, and finally converting the intermediate frequency signal into a discrete-time digital intermediate frequency signal through an analog-to-digital converter.

[0007] The locally replicated signal refers to a periodic oscillation signal containing a pseudo-code and a carrier wave, which is output by selecting candidate Doppler frequency shift and code phase parameters.

[0008] The parameter search range for determining the locally replicated signal specifically refers to: the entire candidate solution space. Wherein: the PVT parameters to be estimated in the DPE receiver The position vector is r The clock difference is δt, and the velocity vector is... v Zhong Piaowei Each dimension corresponds to γ1 to γ8; S i The search space corresponding to each PVT parameter dimension; α The search center for this parameter is determined by the INS solution. n β is the number of search grids and β is the search step size, both determined by the covariance calculated by INS.

[0009] The maximum likelihood estimation refers to determining the PVT candidate parameters based on the correlation results between the digital intermediate frequency signal and the locally replicated signal, using the maximum correlation value as the criterion. Specifically: Where: superscript ~ represents a candidate PVT solution, and superscript · represents the estimation result of the DPE receiver. For the first m The correlation values ​​of the local signal and digital intermediate frequency signal of each satellite under the current candidate solution, where M is the total number of visible satellites.

[0010] The inertial navigation calculation refers to determining the PVT estimate for the next moment by integrating and accumulating the kinematic relationship based on the current PVT estimate and the specific force and angular velocity information output by the INS measurement unit. Specifically: in: and These are the specific force and angular velocity values ​​output by the INS, respectively. and These are the Earth's rotational angular velocity and the Earth's gravitational acceleration, respectively. The carrier attitude described by quaternions. and These are the carrier position and the carrier velocity, respectively. This is the coordinate transformation matrix from the body coordinate system b to the ECEF coordinate system e. The superscript ' indicates the conversion of a three-dimensional vector into the corresponding quaternion, the superscript is the value related to the inertial navigation solution, and the subscript k is the time epoch.

[0011] The aforementioned combined filtering refers to a method of determining the estimated value and covariance after multi-sensor fusion using Kalman filtering based on the estimated values ​​and corresponding estimated covariances from different sensors. Specifically: K k+N =P k+N / k H T HP k+N / k H T +R k+N ) -1 P k+N =(IK k+N H)P k+N / k , Where: the state variable x is defined as the PVT parameter Γ and the extended-dimensional carrier attitude q, and the quantity is measured. z Let K be the difference between the PVT estimates from the DPE receiver and the INS. The superscript ^ represents the state estimate output by the combined system, and represents the state value estimated by the INS through integral iteration. The subscript k+N / k represents the transition matrix or prediction covariance from time k to time k+N. Matrices Φ and H are the linearized and discretized filter state transition matrix and measurement matrix, respectively. The filter parameters P, Q, and R are the estimated covariance, process noise variance, and measurement noise variance of the current state, respectively. I is the identity matrix. The filter gain K at each epoch simultaneously considers the uncertainties of the DPE receiver estimate and the INS estimate.

[0012] The correlator mixes the digital intermediate frequency signal with the locally replicated signal, and then correlates the mixing result with the locally replicated ranging code to output the correlation results in the code phase and Doppler dimension. Specifically: in:τ and f d These represent the code phase and Doppler shift of the current satellite channel, respectively. The superscript ~ indicates candidate code phases or Doppler shifts. and These represent the equivalent code phase and Doppler error of the current candidate PVT solution, respectively. For the first m The correlation values ​​of the local signal and digital intermediate frequency signal of the satellite under the current candidate solution, with the correlation value subscript. τ f and f are the correlation components contributed by code phase and Doppler, respectively. For noise in the correlation domain, sinc(x) = sin(x) / x is the sigma function, and T c The coherent integration time is the time between the local signal and the digital intermediate frequency signal. Technical effect

[0013] This invention uses a DPE receiver as the GNSS receiver framework to realize GNSS / INS integrated navigation, and constructs a DPE / INS quasi-depth integration method and system with the GNSS navigation domain as the core. Under this framework, the key role of INS in finding PVT solutions and constructing the candidate search space of the DPE receiver is determined, which effectively helps the DPE receiver to narrow the search range and improve accuracy, and enables the integrated system to achieve a balance between noise performance and dynamic performance. It solves the problems caused by multipath, non-line-of-sight and interference in challenging urban environments that existing single DPE receivers cannot handle. Attached Figure Description

[0014] Figure 1 This is a flowchart of the present invention;

[0015] Figure 2 This is a schematic diagram of the system of the present invention;

[0016] Figures 3-6 The image shown is a rendering of an example. Detailed Implementation

[0017] like Figure 1 As shown, this embodiment illustrates a quasi-depth combination method for the DPE / INS navigation domain, comprising:

[0018] Step 1: Use IMU data to perform inertial navigation calculations and obtain the corresponding PVT solution and covariance;

[0019] Step 2: Input the digital intermediate frequency signal and the locally replicated signal into the correlator to obtain the maximum likelihood estimate solution for the DPE receiver;

[0020] Step 3: Input the PVT solution of INS and the maximum likelihood estimate solution of DPE receiver into the combined filter to obtain the combined estimate and the corresponding covariance.

[0021] Step 4: Correct the inertial navigation solution obtained in Step 1 with the combined estimated value, determine the parameter search range of the local replicated signal using the combined estimated value and covariance, generate the local replicated signal in Step 2, return to Step 1 and repeat Step 1-Step 4.

[0022] like Figure 2 As shown, the DPE / INS navigation domain quasi-depth combination system implementing the above method in this embodiment includes: an INS calculation module 200, a DPE / INS combination module 210, a navigation domain module 220, and a related domain module 230, wherein: the INS calculation module 200 updates the PVT parameters of each DPE receiver for epoch t. k ~t k+N The IMU data 202 during the period is processed by inertial navigation calculation 203 and superimposed on the initial value 201 to obtain the INS PVT calculation result 204, and then updated at the next DPE receiver parameter update time point t. k+N The INS solution result 204 is output to the DPE / INS combination module 210; the DPE / INS combination module 210, based on the DPE noise variance 226 and DPE estimation result 227 obtained from the correlation domain module 230, and the equivalent PVT solution 212 and corresponding covariance output from the INS solution module, obtains the state estimate 214 and covariance estimate 213 used to compensate for the INS estimate 212 after passing through the combined navigation filter 211. The state estimate 214 is then output to the INS solution module 200 as the initial value 201 for INS solution; the navigation domain module 2 20. Using the DPE covariance 226 and the covariance estimate 213 as the basis, determine the search range nβ and search step size β of the DPE receiver. Using the state estimate 214 as the search center α, calculate the candidate solution space 225 of the DPE receiver. The correlation domain module 230 uses the candidate solution space 225 to construct the code phase 231 and Doppler frequency shift 232 of the local copy signal, and then obtains the local copy signal 233. After inputting it and the digital intermediate frequency signal 234 into the correlator 235, the maximum likelihood estimation 236 is performed to obtain the noise variance 226 and PVT estimation result 227 of the DPE.

[0023] The initial value 201 is determined by the previous DPE receiver parameter update time point t. k The INS estimation result 212 output by this module is obtained by correcting the state estimate 214 in the DPE / INS combined module.

[0024] The inertial navigation solution unit 203 in the INS solution module 200 calculates the solution result 204 based on the initial value 201 and IMU data 202.

[0025] The combined filter unit 211 in the DPE / INS combined module 210 calculates the covariance estimate 213 and the state estimate 214 based on the DPE noise variance 226, the DPE estimation result 227 and the INS estimate 212 in the navigation domain 220 module. Then, it performs a subtraction operation on the state estimate 214 and the INS estimate 212 to obtain the initial value 201 in the INS solution module 200.

[0026] The navigation domain module 220 directly obtains the search center α222 based on the state estimation 214 in the DPE / INS combination module 210. It accumulates and divides the covariance estimation 213 and DPE noise variance 226 in the DPE / INS combination module 210 to obtain the search range nβ223 and search step size β224. Then, it determines the candidate PVT solution 225 by combining the search center α222, the search range nβ223, and the search step size β224.

[0027] The search center α222, search range nβ223, search step size β224, and candidate PVT solution 225 together constitute the PVT parameter 221.

[0028] The correlation domain module 230 calculates and generates code phase 231 and Doppler frequency shift 232 based on the candidate PVT solution 225 in the navigation domain module 220. Then, it assembles a local copy signal 233 based on the code phase 231 and Doppler frequency shift 232. The local copy signal 233 and the digital intermediate frequency signal are correlated by the correlator 235 to obtain a maximum likelihood estimate 236, which is used to generate the DPE noise variance 226 and DPE estimation result 227 in the navigation domain module 220.

[0029] like Figure 3 As shown, in the simulated vehicle scenario, the vehicle speed is 19.63 m / s, the vehicle trajectory is a playground-shaped trajectory, the GNSS uses L1C / A signal, the carrier-to-noise ratio C / N0 = 35 dB-Hz, and the coherence integration time T is... c =20ms, INS noise parameters are gyroscope zero bias stability 0.01° / s (high precision), 0.1° / s (low precision), gyroscope noise density Accelerometer zero bias stability 1.6mg (high precision), 16mg (low precision), accelerometer noise density Under these conditions, the combined system corresponding to the two precision INS reduces position and velocity errors by more than 70% compared to the pure DPE receiver. The combined system with the high precision IMU further reduces velocity error by 25% compared to the combined system with the low precision IMU.

[0030] like Figures 4-6 As shown, in a real-world urban vehicle scenario, the vehicle trajectory is near Lujiazui in Shanghai. Figure 4 The trajectory shown uses an L1C / A GNSS signal, with a runtime of 400 seconds, a signal sampling rate of 30.68 MHz, a quantization bit width of 2 bits, and a coherence integration time T. c =20ms, INS noise parameters are gyroscope zero bias stability 8° / h, gyroscope noise density Accelerometer zero-bias stability 0.015mg, accelerometer noise density

[0031] like Figure 5 As shown, in typical urban multipath, non-line-of-sight, and interference challenging environments within the 40-70s range, compared to a pure DPE receiver, the maximum position error of the DPE / INS combined system is reduced from 250m to 50m, and the maximum velocity error is reduced from 30m / s to 4m / s, representing a performance improvement of over 80%.

[0032] like Figure 6 As shown, around 55s in the simulation time, the parameter search range of the locally replicated signal in the correlator changes for the pure DPE receiver and the DPE / INS combined system. Compared to the pure DPE receiver, the parameter search range of the locally replicated signal in the DPE / INS combined system is reduced by 70% in both the position and velocity dimensions.

[0033] Compared with existing technologies, this invention utilizes the prior information provided by INS, which plays a crucial role in both finding PVT solutions and constructing the candidate search space for the DPE receiver. This effectively helps the DPE receiver narrow its search range and improve accuracy. Vehicle-mounted experiments conducted in simulated in-vehicle scenarios and urban environments demonstrate that, compared to a standalone DPE receiver, the DPE / INS combined system maintains more stable and accurate positioning performance and exhibits robustness in multipath, non-line-of-sight, and interference scenarios.

[0034] The above-described specific implementations can be partially adjusted by those skilled in the art in different ways without departing from the principles and purpose of the present invention. The scope of protection of the present invention is defined by the claims and is not limited to the above-described specific implementations. All implementation schemes within the scope of the claims are bound by the present invention.

Claims

1. A method for combining DPE / INS quasi-depth in a GNSS navigation domain, characterized in that, The GNSS digital intermediate frequency (IF) signal is acquired through the RF front-end processing of the GNSS receiver hardware platform. After determining the parameter search range of the locally replicated signal, the local signal and the IF signal are input into a correlator, and the PVT estimation result and covariance of the DPE receiver are obtained through maximum likelihood estimation. At the same time, the PVT estimation result and covariance of the INS at the current time are calculated using inertial navigation system (INS) calculation. The PVT estimation results and covariance of the DPE receiver and INS are combined and filtered to obtain the PVT estimation result and covariance of the DPE / INS combined system. This is used to determine the parameter search center, search range, and search step size of the locally replicated signal of the DPE receiver, and thus determine a series of locally replicated signals.

2. The DPE / INS quasi-depth combination method in the GNSS navigation domain according to claim 1, characterized in that, The aforementioned radio frequency front-end processing refers to: receiving signals from all visible GNSS satellites through a GNSS antenna, filtering and amplifying them through a pre-filter and a pre-amplifier, then mixing them with the sinusoidal local oscillator signal generated by the local oscillator to downconvert them into an intermediate frequency signal, and finally converting the intermediate frequency signal into a discrete-time digital intermediate frequency signal through an analog-to-digital converter.

3. The DPE / INS quasi-depth combination method in the GNSS navigation domain according to claim 1, characterized in that, The locally replicated signal refers to a periodic oscillation signal containing a pseudo-code and a carrier wave, which is output by selecting candidate Doppler frequency shift and code phase parameters.

4. The DPE / INS quasi-depth combination method in the GNSS navigation domain according to claim 1, characterized in that, The parameter search range for determining the locally replicated signal specifically refers to: the entire candidate solution space. Wherein: the PVT parameters to be estimated in the DPE receiver The position vector is r The clock difference is δt, and the velocity vector is... v Zhong Piaowei Each dimension corresponds to γ1 to γ8; S i The search space corresponding to each PVT parameter dimension; α The search center for this parameter is determined by the INS solution. n β is the number of search grids and β is the search step size, both determined by the covariance calculated by INS.

5. The DPE / INS quasi-depth combination method in the GNSS navigation domain according to claim 1, characterized in that, The maximum likelihood estimation refers to determining the PVT candidate parameters based on the correlation results between the digital intermediate frequency signal and the locally replicated signal, using the maximum correlation value as the criterion. Specifically: Where: superscript ~ represents a candidate PVT solution, and superscript · represents the estimation result of the DPE receiver. For the first m The correlation values ​​of the local signal and digital intermediate frequency signal of each satellite under the current candidate solution, where M is the total number of visible satellites.

6. The DPE / INS quasi-depth combination method in the GNSS navigation domain according to claim 1, characterized in that, The inertial navigation calculation refers to determining the PVT estimate for the next moment by integrating and accumulating the kinematic relationship based on the current PVT estimate and the specific force and angular velocity information output by the INS measurement unit. Specifically: in: and These are the specific force and angular velocity values ​​output by the INS, respectively. and These are the Earth's rotational angular velocity and the Earth's gravitational acceleration, respectively. The carrier attitude described by quaternions. v and v represent the carrier position and carrier velocity, respectively. This is the coordinate transformation matrix from the body coordinate system b to the ECEF coordinate system e. The superscript ' indicates the conversion of a three-dimensional vector into the corresponding quaternion, the superscript is the value related to the inertial navigation solution, and the subscript k is the time epoch.

7. The DPE / INS quasi-depth combination method in the GNSS navigation domain according to claim 1, characterized in that, The aforementioned combined filtering refers to a method of determining the estimated value and covariance after multi-sensor fusion using Kalman filtering based on the estimated values ​​and corresponding estimated covariances from different sensors. Specifically: K k+N =P k+N / k H T HP k+N / k H T +R k+N ) -1 P k+N =(IK k+N H)P k+N / k , Where: state variables x Defined as PVT parameter Γ and extended dimension carrier attitude q, measurement z Let K be the difference between the PVT estimates from the DPE receiver and the INS. The superscript ^ represents the state estimate output by the combined system, and represents the state value estimated by the INS through integral iteration. The subscript k+N / k represents the transition matrix or prediction covariance from time k to time k+N. Matrices Φ and H are the linearized and discretized filter state transition matrix and measurement matrix, respectively. The filter parameters P, Q, and R are the estimated covariance, process noise variance, and measurement noise variance of the current state, respectively. I is the identity matrix. The filter gain K at each epoch simultaneously considers the uncertainties of the DPE receiver estimate and the INS estimate.

8. The DPE / INS quasi-depth combination method in the GNSS navigation domain according to claim 1, characterized in that, The correlator mixes the digital intermediate frequency signal with the locally replicated signal, and then correlates the mixing result with the locally replicated ranging code to output the correlation results in the code phase and Doppler dimension. Specifically: in: τ and f d These represent the code phase and Doppler shift of the current satellite channel, respectively. The superscript ~ indicates candidate code phases or Doppler shifts. and These represent the equivalent code phase and Doppler error of the current candidate PVT solution, respectively. For the first m The correlation values ​​of the local signal and digital intermediate frequency signal of the satellite under the current candidate solution, with the correlation value subscript. τ f and f are the correlation components contributed by code phase and Doppler, respectively. For noise in the correlation domain, sinc(x) = sin(x) / x is the sigma function, and T c The coherent integration time is the time between the local signal and the digital intermediate frequency signal.

9. A DPE / INS navigation domain quasi-depth combination system for implementing the method of any one of claims 1-8, characterized in that, include: The system comprises an INS calculation module, a DPE / INS combination module, a navigation domain module, and a related domain module. Specifically, the INS calculation module updates the PVT parameters of each DPE receiver in epoch t. k ~t k+N The IMU data during the period is processed by inertial navigation and superimposed on the initial values ​​to obtain the PVT calculation result of the INS. This result is then applied at the next DPE receiver parameter update time point t. k+N The INS solution is output to the DPE / INS combined module. The DPE / INS combined module, based on the DPE noise variance and DPE estimation results obtained from the correlation domain module, and the equivalent PVT solution and corresponding covariance output from the INS solution module, uses a combined navigation filter to obtain the state estimate and covariance estimate used to compensate for the INS estimate. The state estimate is then output to the INS solution module as the initial value for the INS solution. The navigation domain module uses the DPE covariance and covariance estimate as the basis to determine the search range nβ and search step size β of the DPE receiver, and uses the state estimate as the search center α to calculate the candidate solution space of the DPE receiver. The correlation domain module uses the candidate solution space to construct the code phase and Doppler frequency shift of the locally replicated signal. Then, the locally replicated signal is input into the correlator along with the digital intermediate frequency signal and maximum likelihood estimation is performed to obtain the noise variance of DPE and the PVT estimation results.

10. The DPE / INS navigation domain quasi-depth combination system according to claim 9, characterized in that, The initial value is determined by the previous DPE receiver parameter update time point t. k The INS estimation results output by this module are obtained by correcting the state estimators in the DPE / INS combined module; In the INS solution module, the inertial navigation solution unit calculates the solution result based on the initial value and IMU data; In the DPE / INS combined module, the combined filter unit calculates the covariance estimate and state estimate based on the DPE noise variance, DPE estimation result and INS estimate in the navigation domain module, and then performs subtraction on the state estimate and INS estimate to obtain the initial value in the INS solution module. The navigation domain module directly obtains the search center α based on the state estimation in the DPE / INS combined module. It obtains the search range nβ and search step size β by summing the covariance estimation in the DPE / INS combined module and the DPE noise variance and dividing the grid. Then, the candidate PVT solution is determined by the search center α, the search range nβ, and the search step size β. The search center α, search range nβ, search step size β, and candidate PVT solutions together constitute the PVT parameters; The correlation domain module calculates the code phase and Doppler frequency shift based on the candidate PVT solution in the navigation domain module, and then assembles the local copy signal based on the code phase and Doppler frequency shift. The correlation module performs correlation calculation on the local copy signal and the digital intermediate frequency signal to obtain the maximum likelihood estimate, which is used to generate the DPE noise variance and DPE estimation results in the navigation domain module.

Images