Inertial auxiliary method for reliable fixing of Beidou-3 multi-frequency ambiguity

Abstract

The application discloses a method for reliably fixing multi-frequency ambiguity of Beidou-3 with inertial assistance, and belongs to the field of high-precision data processing of Beidou / inertial navigation combination navigation. The application adopts a tight combination mode to perform information fusion on the level of Beidou-3 multi-frequency pseudo-range and carrier phase observation values, and uses the position predicted by INS to assist ambiguity resolution of Beidou-3 ultra-wide lane, wide lane and L1 frequency points. When the ultra-wide lane ambiguity is assisted by INS, a non-geometric distance model is adopted, and when the wide lane ambiguity is assisted by INS, a geometric correlation model is adopted. The fixed ultra-wide lane observation value, the position observation value predicted by the inertial navigation and the wide lane observation value are jointly used to solve the floating point wide lane ambiguity, and the LAMBDA method is adopted to perform ambiguity fixing. The ambiguity parameter of the L1 frequency point is solved by using the fixed wide lane observation value, and the position constraint predicted by the INS is not adopted, so that the influence of the system deviation of the INS on the L1 ambiguity fixing is avoided. The application can improve the success rate and reliability of ambiguity fixing.

Description

Technical Field

[0001] This invention relates to a method for reliably fixing multi-frequency ambiguities in BeiDou-3 with inertial assistance, belonging to the field of high-precision data processing for BeiDou / inertial navigation integrated navigation. Background Technology

[0002] The BeiDou-3 Global Navigation Satellite System was officially launched at the end of July 2020, publicly broadcasting four frequency signals (B1C, B1I, B3I, and B2a). High-precision centimeter-level positioning relies on the correct resolution of carrier phase integer ambiguities. BeiDou-3 multi-frequency observations contribute to the rapid and reliable fixation of ambiguities. However, in complex observation environments such as urban areas, satellite signals are inevitably blocked and interfered with. In such cases, observations are easily contaminated by multipath errors, severely impacting navigation and positioning accuracy. Compared to satellite navigation systems, inertial navigation systems (INS) are fully autonomous systems with advantages such as high short-term accuracy. INS are often combined with satellite navigation systems to complement each other, thereby improving overall performance. Specifically, the BeiDou-3 / INS tight combination based on carrier phase observations can utilize the short-term high-precision position information of INS to assist in multi-frequency ambiguity resolution, improving the success rate and reliability of ambiguity fixation. After successful ambiguity fixation, the carrier phase observations can assist INS in reducing error drift and achieving high-precision positioning and attitude determination.

[0003] The key to achieving centimeter-level high-precision positioning using the BeiDou-3 / INS tight combination is INS-assisted multi-frequency ambiguity fixing. Traditional INS-assisted multi-frequency ambiguity fixing methods use the position information predicted by the INS as constraints to sequentially fix the ultra-wide lane, wide lane, and L1 ambiguities. However, the wavelength of the L1 frequency point is only about 20cm. If the INS state experiences a systematic deviation (such as accepting incorrectly fixed ambiguities and using them for tight combination filter updates), the INS positioning constraints will hinder the fixing of L1 ambiguities and may even cause the tight combination filter to diverge, resulting in inaccurate positioning. Summary of the Invention

[0004] To overcome the shortcomings of traditional INS-assisted multi-frequency ambiguity fixing methods when fixing L1 frequency ambiguity, the main objective of this invention is to provide a reliable method for fixing multi-frequency ambiguities in BeiDou-3 using inertial assistance. This method employs a tight combination approach to fuse information at the level of BeiDou-3 multi-frequency pseudorange and carrier phase observations. It utilizes INS-predicted position to assist in the ambiguity resolution of BeiDou-3's ultra-wide lane, wide lane, and L1 frequency points. It jointly uses fixed ultra-wide lane observations, INS-predicted position observations, and wide lane observations to resolve floating-point wide lane ambiguity, and then uses the LAMBDA method for fixing. After fixing the wide lane ambiguity, it can be used to assist in fixing L1 frequency ambiguity. INS-assisted fixing of BeiDou-3's ultra-wide lane ambiguity also reduces the use of pseudorange observations susceptible to gross errors, improving the success rate and reliability of ambiguity fixing. This, in turn, enhances the reliability and availability of the BeiDou-3 / INS tight combination, achieving high-precision positioning and attitude determination.

[0005] The objective of this invention is achieved through the following technical solution:

[0006] This invention discloses a method for reliably fixing multi-frequency ambiguities in BeiDou-3 using inertial-assisted navigation. It employs a tight-combination approach to fuse information at the level of BeiDou-3 multi-frequency pseudorange and carrier phase observations, utilizing INS-predicted position to assist in ambiguity resolution for the ultra-wide lane, wide lane, and L1 frequency points. An INS-assisted ultra-wide lane ambiguity model without geometry is used, while an INS-assisted wide lane ambiguity model employs geometry-dependent ambiguity. The method jointly uses fixed ultra-wide lane observations, INS-predicted position observations, and wide lane observations to calculate floating-point wide lane ambiguity, and then employs the LAMBDA (Least-Squares AM Biguity Decorrelation Adjustment) method for ambiguity fixing. The fixed wide lane ambiguity is then used to assist in fixing L1 frequency ambiguity, thus achieving reliable fixing of multi-frequency ambiguities in BeiDou-3 using inertial-assisted navigation.

[0007] The fixed wide-lane ambiguity is used to assist in fixing the L1 frequency ambiguity. The ambiguity parameters of the L1 frequency point are solved using the fixed wide-lane observation values, instead of using the position constraints predicted by INS. The satellite-to-Earth distance is calculated using the INS-predicted position and the satellite position obtained through ephemeris. This distance replaces the pseudorange observations for calculating the ultra-wide lane ambiguity (B1C-B1I and B3I-B2a), and the ultra-wide lane ambiguity is obtained by rounding. Using the INS-predicted position observations and the fixed ultra-wide lane observations as constraints, a system of equations is constructed together with the wide lane observations (B1I-B3I) to solve for the floating-point wide lane ambiguity and its variance-covariance matrix. Using the fixed wide lane observations as constraints, a system of equations is constructed together with the L1 observations to solve for the L1 floating-point ambiguity and its variance-covariance matrix, and the LAMBDA method is used to fix the wide lane ambiguity. Using the fixed wide lane observations as constraints, a system of equations is constructed together with the L1 observations to solve for the L1 floating-point ambiguity and its variance-covariance matrix, and then the LAMBDA method is used to fix the L1 floating-point ambiguity. Since fixing the ambiguity of the wide lane can provide a high-precision positioning constraint to assist in fixing the L1 ambiguity, the prior position constraint of the INS is not used, thus avoiding the influence of systematic deviations in the INS that hinder the fixing of the L1 ambiguity.

[0008] The implementation of the BeiDou-3 / INS tight combination is as follows: the position information predicted by INS is used to assist in fixing the multi-frequency ambiguity of BeiDou-3. The carrier phase observation value after ambiguity fixing is fused with the inertial measurement value to estimate the inertial device error and navigation error online and output high-precision position, velocity and attitude information. That is, the difference between the carrier phase observation value after fixing L1 ambiguity and the satellite-to-ground distance observation value predicted by INS is used as the measurement to correct the cumulative error of INS and to compensate the inertial device error online. After the BeiDou signal is lost, the compensated inertial device data is used to maintain high-precision mechanical arrangement calculation and output position, velocity and attitude information.

[0009] The method for fixing the ambiguity of the ultra-wide alleyway using a geometric distance-free model is as follows:

[0010] The satellite-to-ground distance is calculated using the INS-predicted position and the satellite position obtained through ephemeris. This satellite-to-ground distance replaces the pseudorange observation value and is used to calculate the ultra-wide lane ambiguity. The ultra-wide lane ambiguity is obtained by rounding.

[0011] The implementation method of fixing the wide-lane ambiguity using a geometric correlation model is as follows:

[0012] Using the INS-predicted location observations and the previously fixed ultra-wide aisle observations as constraints, a system of equations is constructed by combining the wide aisle observations to solve the floating-point wide aisle ambiguity and its variance-covariance matrix. Then, the LAMBDA method is used to fix the wide aisle ambiguity.

[0013] The implementation method of fixing L1 ambiguity using a geometric correlation model is as follows:

[0014] Using fixed wide-lane observations as constraints, a system of equations is constructed in conjunction with L1 observations to solve the L1 floating-point ambiguity and its variance-covariance matrix. Then, the LAMBDA method is used to fix the L1 ambiguity.

[0015] The ambiguity check is implemented as follows:

[0016] When using a geometric correlation model to fix the width lane and L1 ambiguity, the data-driven index ratio value and the model-driven index BootStrapping success rate are used to verify whether the ambiguity fixation is correct. At the same time, the least squares post-hoc residual of the phase observation values ​​of the fixed ambiguity is checked to further determine whether the searched ambiguity is correct.

[0017] Furthermore, the state estimation methods for fusing BeiDou-3 multi-frequency observations with inertial device observations include, but are not limited to, Extended Kalman Filter (EKF) or Graph Optimization.

[0018] Beneficial effects:

[0019] 1. This invention discloses a method for reliably fixing multi-frequency ambiguities in BeiDou-3 with inertial assistance. It utilizes prior position information predicted by the inertial navigation system (INS) to assist in fixing the ultra-wide lane ambiguity and wide lane ambiguity of BeiDou-3, and then uses fixed wide lane observations to assist in fixing the ambiguity of the L1 frequency point. Specifically, when fixing the ultra-wide lane ambiguity with INS assistance, a geometric distance-free model is directly used; when fixing the wide lane and L1 ambiguities, a geometric correlation model is used; and when fixing the L1 ambiguity, a fixed wide lane observation constraint is used instead of the INS prior position constraint. Since a systematic deviation in the L1 wavelength magnitude in the INS prior position will hinder the fixing of the L1 frequency point ambiguity, and the wavelength of the ultra-wide lane and wide lane combination is much larger than the L1 frequency point, the impact of this systematic deviation on fixing the ultra-wide lane and wide lane ambiguities is very limited. This avoids the INS having an existing bias state that hinders the fixing of the L1 ambiguity. This invention can effectively solve the problem of hindering the fixation of L1 frequency ambiguity when there is a system deviation in the inertial navigation position in the existing inertial navigation-assisted multi-frequency ambiguity fixation method. At the same time, the inertial navigation-assisted fixation of ambiguity of Beidou-3 four-frequency signals (B1C, B1I, B3I and B2a) can effectively improve the success rate and reliability of ambiguity fixation, thereby improving the reliability and availability of Beidou-3 / INS tight combination and realizing high-precision positioning and attitude determination.

[0020] 2. The present invention discloses a method for reliable fixation of multi-frequency ambiguity of Beidou-3 with inertial assistance. Since the fixation of wide-lane ambiguity can provide high-precision positioning constraints to assist in the fixation of L1 ambiguity, it does not use INS prior position constraints, thereby avoiding the influence of INS system deviation on the fixation of L1 ambiguity.

[0021] 3. The present invention discloses a method for reliable fixation of multi-frequency ambiguity of Beidou-3 with inertial assistance. In the process of solving the wide lane and L1 ambiguity, a partial ambiguity fixation method is adopted or the observation values ​​of multiple epochs are used for batch processing to solve the ambiguity, thereby further improving the success rate and reliability of ambiguity fixation. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of a method for reliably fixing multi-frequency ambiguities of BeiDou-3 satellites with inertial assistance, as disclosed in this invention. Specific implementation methods

[0023] To facilitate the implementation of this invention, the following will describe the invention in further detail through embodiments. It should be noted that the specific embodiments described herein are merely illustrative examples of this invention and should not be considered as limiting the scope of protection of this invention. The scope of protection of this invention should be determined by the appended claims.

[0024] This embodiment discloses a method for reliably fixing multi-frequency ambiguities in BeiDou-3 assisted by INS. It employs a tight combination approach to fuse information at the level of BeiDou-3 multi-frequency pseudorange and carrier phase observations, utilizing INS-predicted position to assist in ambiguity resolution for BeiDou-3 ultra-wide lane, wide lane, and L1 frequencies. A geometrically distance-free model is used when INS assists in ultra-wide lane ambiguity, while a geometric correlation model is used when assisting in wide lane ambiguity. Floating-point wide lane ambiguity is calculated by jointly using fixed ultra-wide lane observations, INS-predicted position observations, and wide lane observations, and then fixed using the least squares downcorrelation adjustment method. The fixed wide lane ambiguity can be used to assist in fixing L1 frequency ambiguity. Here, to avoid the influence of INS system bias hindering L1 ambiguity fixing, prior INS position constraints are not used, as the fixed wide lane ambiguity provides a higher-precision positioning constraint to assist in L1 ambiguity fixing.

[0025] The implementation method of fixing the ambiguity of ultra-wide alleys using a geometric distance-free model is as follows:

[0026] The satellite-to-ground distance is calculated using the INS-predicted position and the satellite position obtained through ephemeris. This satellite-to-ground distance replaces the pseudorange observation value in the calculation of the ultra-wide alley ambiguity, which can be obtained by rounding.

[0027] The implementation method of fixing the wide-lane ambiguity using a geometric correlation model is as follows:

[0028] Using the INS-predicted location observations and the previously fixed ultra-wide aisle observations as constraints, a system of equations is constructed by combining the wide aisle observations to solve the floating-point wide aisle ambiguity and its variance-covariance matrix, and then the LAMBDA (Least-SquaresAMBiguity Decorrelation Adjustment) method is used to fix it.

[0029] The implementation method of fixing L1 ambiguity using a geometric correlation model is as follows:

[0030] Using fixed wide-lane observations as constraints, a system of equations is constructed in conjunction with L1 observations to solve the L1 floating-point ambiguity and its variance-covariance matrix, and then the LAMBDA method is used to fix it.

[0031] The ambiguity check is implemented as follows:

[0032] When using a geometric correlation model to fix the width lane and L1 ambiguity, the data-driven index ratio value and the model-driven index BootStrapping success rate are used to verify whether the ambiguity fixation is correct. At the same time, the least squares post-hoc residual of the phase observation values ​​of the fixed ambiguity is checked to further determine whether the searched ambiguity is correct.

[0033] The implementation method of the BeiDou-3 / INS tight combination is as follows:

[0034] The difference between the carrier phase observation value with fixed L1 ambiguity and the satellite-to-ground distance observation value predicted by INS is used as a measurement to correct the cumulative error of INS and to compensate for the error of inertial devices online. After the BeiDou signal is lost, the compensated inertial device data can be used to maintain high-precision mechanical arrangement calculation and output information such as position, velocity and attitude.

[0035] like Figure 1 As shown in this embodiment, a method for reliably fixing multi-frequency ambiguities in inertial-assisted BeiDou-3 satellites includes the following steps:

[0036] Step 1 involves preprocessing the BeiDou-3 multi-frequency data acquired by the base station and rover receivers. This process utilizes position information derived from mechanically arranged inertial navigation systems for assistance. The preprocessing includes cycle slip detection and gross error removal.

[0037] Step 2: Solve the ultra-wide alley ambiguity using a geometry-free distance model. The satellite-to-ground distance is calculated using the INS-predicted position and the satellite position obtained through ephemeris. This distance replaces the pseudorange observations in the calculation of the ultra-wide alley ambiguity. The ultra-wide alley ambiguity is obtained by rounding. There are two ultra-wide alley ambiguities for BeiDou-3 four-frequency data, namely N... B1C-B1I and N B3I-B2a .

[0038] Step 3: Solve the wide-lane ambiguity using a geometric correlation model. Using the INS-predicted position observations and the previously fixed ultra-wide-lane observations as constraints, a system of equations is constructed using the combined wide-lane observation values ​​to solve for the floating-point wide-lane ambiguity and its variance-covariance matrix. Then, the LAMBDA method is used to fix the wide-lane ambiguity NB1. I-B3I .

[0039] Step 4: Solve the L1 ambiguity using a geometric correlation model. Using fixed wide-lane observations as constraints, a system of equations is constructed using the L1 observations to solve the L1 floating-point ambiguity and its variance-covariance matrix. Then, the LAMBDA method is used to fix the L1 ambiguity.

[0040] Step 5, Ambiguity Confirmation and Verification. For the wide lane and L1 ambiguities fixed using the geometric correlation model, the correctness of the ambiguity fixation is verified using the data-driven index ratio value and the model-driven index BootStrapping success rate. At the same time, the least squares post-hoc residuals of the phase observations of the fixed ambiguities are checked to further determine whether the searched ambiguities are correct.

[0041] Step 6: Fuse the L1 carrier phase observations with fixed ambiguity with the inertial measurement values, estimate the inertial sensor error and the position, velocity and attitude errors of the carrier in real time online, and output the final high-precision position, velocity and attitude results using closed-loop correction. The data fusion algorithm includes, but is not limited to, Extended Kalman Filter (EKF), Unscented Kalman Filter (UKF) or Graph Optimization.

[0042] In practical implementation, the technical solution of this embodiment can be automatically run by a computer program. The following is a detailed explanation using extended Kalman filtering as the fusion algorithm for BeiDou-3 multi-frequency data and inertial data.

[0043] This embodiment discloses a method for reliably fixing multi-frequency ambiguities in inertial-assisted BeiDou-3 navigation satellites, comprising the following steps:

[0044] Step 1: BeiDou-3 / INS tight combination solution.

[0045] The BeiDou-3 / INS compact combination based on Kalman filtering inputs the carrier phase observation values ​​(fixed or floating-point solution) at the L1 frequency point into the Kalman filter to estimate inertial sensor errors (zero bias and scaling factor, etc.) and position, velocity, and attitude errors online. Closed-loop correction is then used to provide feedback correction for inertial sensor errors and navigation parameter errors. The BeiDou-3 / INS compact combination model includes a state model and an observation model. The navigation coordinate system is selected as the Earth-Centered Earth-Fixed (ECEF) coordinate system. The error state model of the compact combination can be expressed as:

[0046]

[0047] in, and These are position error, velocity error, and attitude error, respectively. and These are the time derivatives of the corresponding quantities, f b The specific force output by the accelerometer. Let be the rotation matrix from the vehicle coordinate system to the navigation system. Let δg be the angular velocity of Earth's rotation. e Due to gravity error, δb represents the angular velocity error of the gyroscope output. g δb a These are the zero bias errors of the gyroscope and accelerometer, respectively. and These are the time derivatives of the corresponding quantities. These are the first-order Gaussian Markov correlation times corresponding to the zero bias errors of the gyroscope and accelerometer, respectively, w g w a These are the driving white noises for the gyroscope and accelerometer, respectively.

[0048] The observation model of the BeiDou-3 / INS compact system establishes a functional relationship between observed data and system state parameters. The BeiDou-3 / INS compact system primarily uses L1 frequency carrier phase observations after stepwise ambiguity fixing. The specific formula is given below using a double-difference positioning model as an example. For each satellite, the following observation equation applies:

[0049]

[0050] in, This is a double difference operator, where the subscripts b and r represent the base station and the rover station, respectively, the superscript j represents the reference satellite, and k represents the non-reference satellite; Carrier phase observation; ρ is the geometric distance from the receiver to the satellite; T and I are the tropospheric and ionospheric delays, respectively; λ and N are the carrier wavelength and carrier phase integer ambiguity, respectively. This includes measurement noise of the carrier phase and other unmodeled errors.

[0051] Based on the observation equations formed by the single satellite and the reference satellite, the observation model for a certain observation epoch k is established as follows:

[0052] Z k =H k δx k +η k (3)

[0053] In the formula H k For compact combination design matrix; Z k For the observed data updated by filtering; η k This is the measurement noise vector.

[0054] After establishing the system state model and observation model of the BeiDou-3 / INS tightly integrated system, the extended Kalman filter is used to fuse the two and estimate the inertial sensor error and navigation parameter error online. After each filter update, the inertial sensor error and navigation error are compensated and the integrated navigation result is output.

[0055] Step 2: BeiDou-3 / INS tightly coupled multi-frequency ambiguity resolution. Step 1 corresponds to the specific implementation methods of steps 2 to 5.

[0056] In the BeiDou-3 / INS tight integration, inertial sensor errors can be estimated online through filtering and fusion. INS provides high positional accuracy in the short term, which can help to quickly and reliably fix ambiguities. In this embodiment, the prior position predicted by inertial navigation is used to assist in the resolution of ultra-wide aisle and wide aisle ambiguities when resolving BeiDou-3 multi-frequency ambiguities.

[0057] Based on the sensor error obtained in step one, the ultra-wide lane and wide lane ambiguities are calculated; then, a fixed wide lane ambiguity constraint is used to assist in solving the L1 ambiguity. The specific implementation steps are as follows:

[0058] Step 2.1: Solve the ambiguity of the ultrawide alley using a geometric distance-free model. Step 2.1 corresponds to Step 2.

[0059] The double-difference geometric distance between the satellite and the receiver is obtained by using the prior position information predicted by INS and the satellite position calculated by satellite ephemeris. For short baselines, ionospheric and tropospheric errors are ignored. Therefore, according to equation (2), the two ultrawide ambiguities of the BeiDou-3 four-frequency observations are:

[0060]

[0061]

[0062] Where Round is the rounding operator; The ultrawide lane ambiguity for frequency points B1C and B1I; The ultrawide lane ambiguity for frequency points B3I and B2a; The ultra-wide alleyway observations, measured in weeks, are formed by frequency points B1C and B1I. The ultra-wide alleyway observations, measured in weeks, are formed by frequency points B1I and B2a. λ represents the double-difference distance observations predicted by INS. B1C-B1I With λ B3I-B2a These are the wavelengths of the two ultra-wide aisle observations mentioned above.

[0063] Step 2.2: Solve the wide-lane ambiguity using a geometric correlation model. Step 2.2 corresponds to Step 3.

[0064] The wide-lane ambiguity is solved using the prior location information predicted by INS and two fixed ultra-wide-lane carrier phase observations through the following equation:

[0065]

[0066] In the formula, ε is the residual of the observation equation, I is the identity matrix, A is the design matrix consisting of the line-of-sight vectors between the satellite and the receiver, and λ B1I-B3I Let δX be the wavelength corresponding to the wide aisle observation. These are the coordinate corrections and the width-lane ambiguity to be solved, respectively. and These are two observations of the ultra-wide alleyway (in meters) after the ambiguity has been fixed. For the wide-lane observations for which ambiguity needs to be solved, This is the approximate satellite-to-ground distance calculated using the INS-predicted location and satellite positions.

[0067] Step 2.3: Solve the L1 ambiguity using the geometric correlation model. Step 2.3 corresponds to Step 4.

[0068] Using fixed wide-lane observations as constraints instead of prior inertial navigation system (INS) positions minimizes the impact of INS bias on L1 ambiguity fixation, thereby improving the success rate and reliability of ambiguity fixation. The equations for L1 ambiguity resolution using fixed wide-lane observations are as follows:

[0069]

[0070] In the formula, λ B1I For the wavelength of the BeiDou B1I frequency point, N B1I Let B1I ambiguity parameters be the ones to be solved. These are the wide-lane observations after the ambiguity has been fixed. For the B1I phase observations for which ambiguity needs to be resolved.

[0071] In the process of solving the ambiguity of the wide lane and L1, it is preferable to use a partial ambiguity fixation method or to use observations from multiple epochs for batch processing to further improve the success rate and reliability of ambiguity fixation.

[0072] Step 2.4: Ambiguity Confirmation and Verification. Step 2.4 corresponds to Step 5.

[0073] Successful and correct ambiguity fixation is a prerequisite for the BeiDou-3 / INS compact system to achieve high-precision positioning and attitude determination based on multi-frequency carrier phase observations. After ambiguity fixation, carrier observations can be used as high-precision range observations to correct INS errors. Conversely, incorrect ambiguity fixation will severely affect the positioning and attitude determination accuracy of the BeiDou-3 / INS compact system. Therefore, it is necessary to confirm and verify the ambiguities of the searched wide lane and L1 frequency points.

[0074] In this embodiment, the method for ambiguity confirmation is to comprehensively utilize the data-driven indicator ratio value and the model-driven indicator BootStrapping success rate to confirm whether the searched ambiguity is correct. Specifically, the LAMBDA method is used to fix the ambiguity, the ratio value is required to be greater than 3, and the BootStrapping success rate is greater than 0.99.

[0075] It should be understood that the above-described specific embodiments are quite detailed, but this should not be considered as a limitation on the scope of protection of this invention. Under the guidance of this invention, those skilled in the art can make appropriate adjustments without departing from the scope of protection of the claims of this invention, and all such adjustments shall fall within the scope of protection of this invention. The scope of protection of this invention shall be determined by the appended claims.

Claims

1. A method for reliably fixing multi-frequency ambiguities in inertial-assisted BeiDou-3 navigation satellites, characterized in that: A compact combination approach is adopted to perform information fusion at the level of BeiDou-3 multi-frequency pseudorange and carrier phase observations. The position predicted by INS is used to assist in the ambiguity resolution of BeiDou-3 ultra-wide lane, wide lane, and L1 frequency points. When INS assists in ultra-wide lane ambiguity, a geometric distance-free model is used, while when assisting in wide lane ambiguity, a geometric correlation model is used. The floating-point wide lane ambiguity is solved by jointly using fixed ultra-wide lane observations, INS-predicted position observations, and wide lane observations. Then, the LAMBDA method is used for ambiguity fixation. The fixed wide lane ambiguity is used to assist in the fixation of L1 frequency ambiguity, that is, to achieve reliable fixation of inertial-assisted BeiDou-3 multi-frequency ambiguity. The fixed wide-lane ambiguity is used to assist in fixing the L1 frequency ambiguity. The ambiguity parameters of the L1 frequency are solved using fixed wide-lane observations instead of INS-predicted position constraints. Specifically, the satellite-to-ground distance is calculated using the INS-predicted position and the satellite position obtained through ephemeris. This distance replaces the pseudorange observations in calculating the ultra-wide-lane ambiguities B1C-B1I and B3I-B2a. The ultra-wide-lane ambiguity is obtained by rounding. The INS-predicted position observations and the fixed ultra-wide-lane observations are used as... Constraints are applied, and equations are constructed using the combined wide-lane observations B1I-B3I to solve for the floating-point wide-lane ambiguity and its variance-covariance matrix. Using fixed wide-lane observations as constraints, equations are constructed using the combined L1 observations to solve for the L1 floating-point ambiguity and its variance-covariance matrix. The LAMBDA method is then used to fix the wide-lane ambiguity.

2. The method for reliably fixing multi-frequency ambiguities of inertial-assisted BeiDou-3 as described in claim 1, characterized in that: The implementation of the BeiDou-3 / INS tight combination is as follows: the position information predicted by INS is used to assist in fixing the multi-frequency ambiguity of BeiDou-3. The carrier phase observation value after ambiguity fixing is fused with the inertial measurement value to estimate the inertial device error and navigation error online and output high-precision position, velocity and attitude information. That is, the difference between the carrier phase observation value after fixing L1 ambiguity and the satellite-to-ground distance observation value predicted by INS is used as the measurement to correct the cumulative error of INS and to compensate the inertial device error online. After the BeiDou signal is lost, the compensated inertial device data is used to maintain high-precision mechanical arrangement calculation and output position, velocity and attitude information.

3. The method for reliably fixing multi-frequency ambiguities of inertial-assisted BeiDou-3 as described in claim 1, characterized in that: The method for fixing the ambiguity of the ultra-wide alleyway using a geometric distance-free model is as follows: The satellite-to-ground distance is calculated using the INS-predicted position and the satellite position obtained through ephemeris. This satellite-to-ground distance will replace the pseudorange observation value in the calculation of the ultra-wide alley ambiguity. The ultra-wide alley ambiguity is obtained by rounding. The implementation method of fixing the wide-lane ambiguity using a geometric correlation model is as follows: Using the INS-predicted location observations and the previously fixed ultra-wide aisle observations as constraints, a system of equations is constructed by combining the wide aisle observations to solve the floating-point wide aisle ambiguity and its variance-covariance matrix. Then, the LAMBDA method is used to fix the wide aisle ambiguity. The implementation method of fixing L1 ambiguity using a geometric correlation model is as follows: Using fixed wide-lane observations as constraints, a system of equations is constructed in conjunction with L1 observations to solve the L1 floating-point ambiguity and its variance-covariance matrix. Then, the LAMBDA method is used to fix the L1 ambiguity.

4. The method for reliably fixing multi-frequency ambiguities of inertial-assisted BeiDou-3 as described in claim 1, characterized in that: The ambiguity check is implemented as follows: When using a geometric correlation model to fix the width lane and L1 ambiguity, the data-driven index ratio value and the model-driven index BootStrapping success rate are used to verify whether the ambiguity fixation is correct. At the same time, the least squares post-hoc residual of the phase observation values ​​of the fixed ambiguity is checked to further determine whether the searched ambiguity is correct.

5. The method for reliably fixing multi-frequency ambiguities of inertial-assisted BeiDou-3 as described in claim 1, characterized in that: Includes the following steps, Step 1: Preprocess the BeiDou-3 multi-frequency data acquired by the base station and rover receivers. In this process, the position information recursively arranged by the inertial navigation system is used for assistance. The preprocessing includes cycle slip detection and gross error removal. Step 2: Solve the ultra-wide alley ambiguity using a geometry-free distance model; calculate the satellite-to-ground distance using the INS-predicted position and the satellite position obtained through ephemeris. This satellite-to-ground distance will replace the pseudorange observation value in calculating the ultra-wide alley ambiguity. The ultra-wide alley ambiguity is obtained by rounding. There are two ultra-wide alley ambiguities for BeiDou-3 four-frequency data, namely N. B1C-B1I and N B3I-B2a ; Step 3: Solve the wide-lane ambiguity using a geometric correlation model; using the INS-predicted position observations and the previously fixed ultra-wide-lane observations as constraints, construct a system of equations based on the combined wide-lane observations to solve the floating-point wide-lane ambiguity and its variance-covariance matrix, and then use the LAMBDA method to fix the wide-lane ambiguity N. B1I-B3I ; Step 4: Solve the L1 ambiguity using a geometric correlation model; use fixed wide aisle observations as constraints, combine L1 observations to form a system of equations to solve the L1 floating-point ambiguity and its variance-covariance matrix, and then use the LAMBDA method to fix the L1 ambiguity. Step 5, Ambiguity Confirmation and Verification: For the wide lane and L1 ambiguities fixed using the geometric correlation model, the correctness of ambiguity fixing is verified by using the data-driven index ratio value and the model-driven index BootStrapping success rate. At the same time, the least squares post-hoc residual of the phase observation values ​​of the fixed ambiguity is checked to further determine whether the searched ambiguity is correct. Step 6: Fuse the L1 carrier phase observations with fixed ambiguity with the inertial measurement values, estimate the inertial sensor error and the position, velocity and attitude errors of the carrier in real time online, and output the final high-precision position, velocity and attitude results using closed-loop correction. The data fusion algorithm includes extended Kalman filtering, unscented Kalman filtering or graph optimization.

6. The method for reliably fixing multi-frequency ambiguities of inertial-assisted BeiDou-3 as described in claim 5, characterized in that: The BeiDou-3 / INS compact combination based on Kalman filtering inputs carrier phase observations at the L1 frequency point into a Kalman filter to estimate inertial sensor errors and position, velocity, and attitude errors online. Closed-loop correction is then used to provide feedback correction for inertial sensor and navigation parameter errors. The BeiDou-3 / INS compact combination model includes a state model and an observation model. The navigation coordinate system is selected as the geocentric-ground-fixed coordinate system ECEF. The error state model of the compact combination is expressed as follows: in, and These are position error, velocity error, and attitude error, respectively. and These are the time derivatives of the corresponding quantities, f b The specific force output by the accelerometer. Let be the rotation matrix from the vehicle coordinate system to the navigation system. Let δg be the angular velocity of Earth's rotation. e Due to gravity error, δb represents the angular velocity error of the gyroscope output. g δb a These are the zero bias errors of the gyroscope and accelerometer, respectively. and These are the time derivatives of the corresponding quantities. These are the first-order Gaussian Markov correlation times corresponding to the zero bias errors of the gyroscope and accelerometer, respectively, w g w a The white noise driving the gyroscope and accelerometer are respectively. The observation model of the BeiDou-3 / INS compact system establishes a functional relationship between the observed data and the system state parameters. The BeiDou-3 / INS compact system primarily uses the L1 frequency carrier phase observations after stepwise ambiguity fixing. For each satellite, the following observation equation applies: in, This is a double difference operator, where the subscripts b and r represent the base station and the rover station, respectively, the superscript j represents the reference satellite, and k represents the non-reference satellite; Carrier phase observation; ρ is the geometric distance from the receiver to the satellite; T and I are the tropospheric and ionospheric delays, respectively; λ and N are the carrier wavelength and carrier phase integer ambiguity, respectively. This includes measurement noise of the carrier phase and other unmodeled errors; Based on the observation equations formed by a single satellite and a reference satellite, the observation model for a certain observation epoch k is established as follows: Z k =H k δx k +η k (3) In the formula H k For compact combination design matrix; Z k For the observed data updated by filtering; η k This is the measurement noise vector.

7. The method for reliably fixing multi-frequency ambiguities of inertial-assisted BeiDou-3 as described in claim 6, characterized in that: The ambiguity of the ultra-wide lane and wide lane is calculated based on the sensor error; then, a fixed wide lane ambiguity constraint is used to assist in the calculation of the L1 ambiguity. The specific implementation steps are as follows. The method for solving the ambiguity of ultra-wide alleys using a geometric distance-free model described in step 2 is as follows: The double-difference geometric distance between the satellite and the receiver is obtained by using the prior position information predicted by INS and the satellite position calculated by satellite ephemeris. For short baselines, ionospheric and tropospheric errors are ignored. Therefore, according to equation (2), the two ultrawide ambiguities of the BeiDou-3 four-frequency observations are: Where Round is the rounding operator; The ultrawide lane ambiguity for frequency points B1C and B1I; The ultrawide lane ambiguity for frequency points B3I and B2a; The ultra-wide alleyway observations, measured in weeks, are formed by frequency points B1C and B1I. The ultra-wide alleyway observations, measured in weeks, are formed by frequency points B1I and B2a. λ represents the double-difference distance observations predicted by INS. B1C-B1I With λ B3I-B2a These are the wavelengths of the two ultra-wide alleyway observations mentioned above; Step 3 describes the method for solving the wide-lane ambiguity using a geometric correlation model; The wide-lane ambiguity is solved using the prior location information predicted by INS and two fixed ultra-wide-lane carrier phase observations through the following equation: In the formula, ε is the residual of the observation equation, I is the identity matrix, A is the design matrix consisting of the line-of-sight vectors between the satellite and the receiver, and λ B1I-B3I Let δX be the wavelength corresponding to the wide aisle observation. These are the coordinate corrections and the width-lane ambiguity to be solved, respectively. and These are two ultra-wide alleyway observations after the ambiguity has been fixed. For the wide-lane observations for which ambiguity needs to be solved, This is the approximate satellite-to-ground distance calculated using the INS-predicted location and satellite positions; The method for solving L1 ambiguity using the geometric correlation model described in step 4 is as follows: Using fixed wide-lane observations as constraints instead of inertial navigation prior positions minimizes the impact of inertial navigation bias on L1 ambiguity fixation, thereby improving the success rate and reliability of ambiguity fixation. The equations for L1 ambiguity resolution using fixed wide-lane observations are as follows: In the formula, λ B1I N is the wavelength of the BeiDou B1I frequency point. B1I Let B1I ambiguity parameters be the ones to be solved. These are the wide-lane observations after the ambiguity has been fixed. For B1I phase observations where ambiguity needs to be resolved; The ambiguity confirmation and verification method described in step 5 is as follows: By combining data-driven metrics such as the ratio value and model-driven metrics such as the BootStrapping success rate, we can confirm whether the fuzziness of the search is correct.

8. The method for reliably fixing multi-frequency ambiguities of inertial-assisted BeiDou-3 as described in claim 7, characterized in that: In the process of solving the ambiguity of the wide lane and L1, a partial ambiguity fixation method or batch processing using observations from multiple epochs is adopted to further improve the success rate and reliability of ambiguity fixation.

Images