A GNSS / INS combined positioning method based on common frequency signal difference between systems

Abstract

The present invention relates to a GNSS / INS combined positioning method based on the difference of common frequency signals between systems. The invention combines the development trend of compatibility and interoperability of GNSS technology, utilizes the common frequency signals between different satellite systems, and proposes a new GNSS / INS combined positioning method based on the excellent stability of differential inter-system biases (DISB) between the same frequencies of different systems in the time domain. Firstly, the prior information of DISB parameters is obtained through ultra-short baseline static experiments; using the prior DISB information, GNSS / INS combined positioning is realized based on the GNSS tight combination RTK based on the difference of common frequency signals between systems. Compared with the traditional GNSS / INS combined positioning algorithm, this method can fully mine the information of multi-system GNSS and significantly increase the amount of available information, thereby effectively improving the positioning performance of the GNSS / INS combined system in complex environments.

Description

Technical Field

[0001] The present invention relates to the field of GNSS / INS combined positioning data processing, and in particular to a GNSS / INS combined positioning method for inter-system common frequency signal differentiation. Background Art

[0002] High-precision positioning technology has driven the rapid development of a series of industries and is the premise for the successful realization of beautiful visions such as unmanned driving, smart transportation, and smart agriculture. The GNSS / INS combination technology based on the global navigation satellite system (GNSS) positioning technology and the inertial navigation system (INS) technology is undoubtedly the most widely used high-precision positioning method in outdoor scenes. Although this technology has been put into practical engineering use for nearly 40 years, there are still some urgent problems to be solved. Especially for the civil GNSS / INS combination system, due to its heavy reliance on GNSS signals, GNSS needs to provide sufficient information to suppress the accumulation of INS errors. When the GNSS signal is interfered or blocked, the INS error will accumulate rapidly, which will cause the combination system to be unable to maintain high-precision positioning and attitude for a long time in such scenes. Scholars have proposed many ways to improve the positioning performance of the GNSS / INS combination system in this case, such as using vehicle carrier information constraints, adding other types of sensor assistance, etc. These methods are essentially to introduce new observation information to replace GNSS observations to suppress the divergence of INS errors. The introduction of new information must take into account the cost of new equipment and the issue of information fusion, which will further increase the difficulty of implementing a GNSS / INS combined system.

[0003] On the one hand, there are currently four major satellite navigation systems, namely China's BDS, the United States' GPS, the European Union's GALILEO and Russia's GLONASS, which can provide PNT services to global users, and Japan's QZSS and India's NAVIC, two regional navigation systems, which can provide PNT services to Asia-Pacific users. In order to increase compatibility and interoperability with other satellite systems, each GNSS system currently intentionally broadcasts signals that overlap with some other satellite systems. Considering that satellite signals of the same frequency can share the RF front end to reduce the hardware cost of the receiver, as well as the excellent compatibility and interoperability between signals of the same frequency, it can be foreseen that supporting overlapping frequencies of multiple systems will be the development trend of GNSS receivers in the future.

[0004] On the other hand, the current commercial GNSS / INS combined positioning algorithm is usually based on the combination of (RealTime kinematic, RTK) + INS. In this type of RTK algorithm, each satellite system will use a reference satellite separately. This type of RTK is also called GNSS between-system loose combination RTK (GNSS between-system Loose combination RTK, LRTK). Correspondingly, there is another type of RTK mode called GNSS between-system tight combination RTK (GNSS between-system Tightcombination RTK, TRTK), in which all satellite systems will share a reference satellite. Of course, due to the differences in wavelength, hardware delay and initial phase deviation corresponding to the observation values ​​of different satellite systems, TRTK is more troublesome to implement than LRTK. TRTK based on overlapping frequency only needs to consider the influence of differential hardware delay (Differential Inter-system Bias, DISB) between different satellite systems. DISB has very good stability in the time domain, which enables TRTK based on overlapping frequency to make better use of GNSS observation information than LRTK. Summary of the invention

[0005] The purpose of the present invention is to provide a GNSS / INS combined positioning method for differentiating common frequency signals between systems, in order to improve the positioning performance of the GNSS / INS combined system in complex environments, in order to address the pain points of the current GNSS / INS combined positioning algorithm.

[0006] In order to achieve the above object, the technical solution of the present invention is as follows:

[0007] A GNSS / INS combined positioning method for common frequency signal difference between systems includes the following steps:

[0008] Step S1: Place the device to be tested and the reference station at the two ends of the ultra-short baseline respectively, keep them stationary and synchronously collect common frequency data for 10-60 minutes;

[0009] Step S2: single epoch estimation of DISB, and obtaining the prior value of DISB based on the DISB time series;

[0010] Step S3: receiving and decoding the IMU observations and GNSS observations generated by the integrated navigation module, and the GNSS corrections broadcast by the reference station;

[0011] Step S4: Using the DISB information obtained by the pre-correction, according to the combined positioning algorithm embedded in the integrated navigation module, perform GNSS / INS loose combination or tight combination solution, filter solution of the parameters to be estimated, and fix and transfer the ambiguity parameters.

[0012] The step S1 comprises:

[0013] Step S11: Arrange the device to be tested and the reference station as an ultra-short baseline;

[0014] Step S12: receiving and decoding GNSS common frequency observation values ​​of the device under test for 10-60 minutes;

[0015] Step S13: Receive and decode the GNSS common frequency observation values ​​of the reference station for 10-60 minutes.

[0016] The step S2 comprises:

[0017] Step S21: pre-processing GNSS observation data;

[0018] Data preprocessing includes but is not limited to monitoring and eliminating abnormal data, detecting and repairing cycle slips, single-point positioning of monitoring stations and reference stations, setting of satellite cutoff elevation angle, setting of data cutoff signal-to-noise ratio, setting of maximum time difference of asynchronous differential, setting of reference satellite selection strategy, and weighting according to elevation angle;

[0019] The height angle weighting formula used is:

[0020]

[0021] Among them, θ i represents the altitude angle of the i-th satellite, σ0 is the prior unit weighted mean error, and σ is the prior standard deviation of the i-th satellite observation value;

[0022] Step S22: constructing a double difference observation equation;

[0023] First, the known GNSS raw observations are written as

[0024]

[0025]

[0026] Among them, the subscripts s, K, j, and r represent the satellite, the satellite system, the frequency, and the receiver respectively; φ and p represent the carrier phase and pseudorange observation values, respectively, in meters; ρ, τ, and ι represent the satellite-to-earth distance, tropospheric delay, and ionospheric delay, respectively; in represents frequency j; dt r , They represent the receiver clock error and satellite clock error respectively; Indicates the phase hardware delay of the receiver end and the phase hardware delay of the satellite end; represents the pseudorange hardware delay at the receiver end and the pseudorange hardware delay at the satellite end; λ j Indicates wavelength; Indicates ambiguity; and Respectively represent the observation noise of phase and pseudorange; the upper and lower scripts of the parameters reflect the factors that affect the parameters;

[0027] If there is a satellite v belonging to system Q at this time, and satellite v also broadcasts a signal on frequency j, then the double difference observation between satellites s and v is:

[0028]

[0029]

[0030] For the convenience of derivation, the dual-system single-frequency short baseline case is used as an example for derivation, that is, the influence of ionospheric and tropospheric delays is ignored, and it is assumed that n and m satellites in systems Q and K are observed respectively; then the single-epoch TRTK observation equation is written as:

[0031]

[0032] in,

[0033] With a similar structure, is the coordinate correction value to be estimated;

[0034] Step S23: estimating parameter floating point solution;

[0035] For formula (4), use

[0036]

[0037] Make equivalent changes to obtain the relationship between TRTK and LRTK

[0038]

[0039] in

[0040]

[0041] That is, the observation equation of a single epoch of LRTK. Compared with LRTK, TRTK introduces two new observation values ​​and three parameters to be estimated. Since the DISBs parameter has very good stability in the time domain, and the ambiguity parameter also maintains a constant characteristic in the time domain when there is no cycle slip, as the observation time accumulates, TRTK will have more and more redundant observation information than LRTK. However, due to the and In order to correctly calculate the value of phase DISB, and Merge parameters; use To replace the combined parameters, equation (5) can be written as

[0042]

[0043] Finally, the floating point values ​​of all parameters in each epoch are calculated using least squares;

[0044] Step S24: Obtain DISB parameters;

[0045] The DISB parameter is defined as and However, due to the parameter estimation and Cannot be separated, It is impossible to estimate, so the usual practice is to The fractional part Considered as the equivalent phase DISB; first calculate the and In order to reduce the impact of observation noise on DISB, the final DISB is obtained by taking an average over a long period of time.

[0046] The step S3 comprises:

[0047] Step S31: Receive and decode IMU observations and provide them to S4;

[0048] Step S32: receiving and decoding GNSS observation values ​​in the combined system and providing them to S4;

[0049] Step S33: Receive and decode the reference station GNSS correction data and provide it to S4.

[0050] The step S4 comprises:

[0051] Step S41: pre-processing IMU data, monitoring and eliminating abnormal data, unit conversion, coordinate conversion in sequence; providing to S42;

[0052] At the same time, the GNSS observation data is pre-processed, and abnormal data monitoring and elimination, cycle slip detection and repair, single-point positioning of monitoring stations and reference stations, satellite cutoff elevation angle setting, data cutoff signal-to-noise ratio setting, asynchronous differential maximum time difference setting, reference satellite selection strategy setting, and weighting according to elevation angle are performed in sequence; and provided to S42;

[0053] Step S42: Initialize INS: calculate the approximate position, velocity and attitude of the combined system through GNSS SPP, set the initial variance covariance matrix and process noise matrix of the filter system, and enter S43;

[0054] Step S43: Perform INS mechanical arrangement and enter S44;

[0055] Step S44: Combine the fixed ambiguity parameters to construct the GNSS / INS loose / tight combination equation, including:

[0056] According to the relationship between space vectors, we get:

[0057]

[0058] The subscripts b and e represent the IMU's own coordinate system and the Earth-centered Earth-fixed coordinate system, respectively. represents the coordinates of the phase center of the GNSS antenna in the e system, Represents the coordinates of the IMU center of mass in the e system, represents the coordinate transformation matrix from b system to e system, l b represents the spatial distance between the GNSS antenna phase center and the IMU centroid in the IMU coordinate system, which is a constant. Equation (8) describes the relationship between the position solved by GNSS and the position solved by INS under ideal conditions. Since the solution process will inevitably introduce errors in practice, the error perturbation analysis of Equation (8) is performed to obtain the relationship between the position error solved by GNSS and the position error solved by INS:

[0059]

[0060] Where δ represents the error and φ represents the attitude error; then the GNSS / INS tight combination equation based on TRTK is written as

[0061]

[0062] The loose combination equation of GNSS / INS based on TRTK is written as

[0063]

[0064] Where (~) represents the observed value, represents the speed error, bω ,b f Represent the zero bias of the gyroscope and accelerometer, s ω ,s f Represent the scale factors of the gyroscope and accelerometer respectively;

[0065] Go to step S45;

[0066] Step S45: Filter and solve the estimated parameters, output the final position, speed, and attitude settlement results, and provide the settlement results to S46 for further iteration:

[0067] Different random walk models are established according to the properties of the parameters, and Kalman filtering is used to solve all the parameters to be estimated;

[0068] Go to step S46;

[0069] Step S46: Fixing and transferring fuzziness parameters:

[0070] The LAMBDA method is used to fix the ambiguity parameters. After the ambiguity is correctly fixed, the fixed ambiguity is obtained. At the same time, the fixed ambiguity parameters are recorded for use in subsequent epochs, and S44 is executed until all data are resolved.

[0071] In summary, the GNSS / INS combined positioning method of common frequency signal difference between systems of the present invention can mine the information of multi-system GNSS and create new observation information for the GNSS / INS combined system without additional enhancement cost, thereby effectively improving the positioning performance of the combined system in complex environments.

[0072] Specifically, compared with the prior art, the present invention has the following advantages:

[0073] The existing GNSS / INS combined positioning algorithm is heavily dependent on GNSS information. When the GNSS signal is interfered with or blocked, it will not be able to suppress the accumulation of INS errors, resulting in a decrease in the positioning performance of the combined system. The existing GNSS / INS combined positioning improvement technology requires the introduction of new external sensors, which on the one hand increases the hardware cost of the combined system, and on the other hand further increases the difficulty of implementing the fusion algorithm. The present invention makes use of the superior stability of DISB in the time domain, fully exploits the information of multi-system GNSS, significantly increases the amount of available information of the combined system, and can effectively improve the positioning performance of the current GNSS / INS combined system in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 Flow chart of the GNSS / INS combined positioning method based on the difference of common frequency signals between systems; DETAILED DESCRIPTION

[0075] The specific embodiments of the present invention will be described in more detail below in conjunction with the schematic diagram. The advantages and features of the present invention will become more apparent from the following description and claims.

[0076] refer to Figure 1 In a preferred embodiment of the present invention, a GNSS / INS combined positioning method for inter-system common frequency signal difference includes:

[0077] Step S1: Place the device to be tested (which can be replaced by a device of the same model as the device to be tested) and the reference station at the two ends of the ultra-short baseline, and keep them stationary and synchronously collect common frequency data for 10-60 minutes;

[0078] Specifically, step S1 includes:

[0079] Step S11: Arrange the device to be tested and the reference station as an ultra-short baseline;

[0080] Step S12: receiving and decoding GNSS common frequency observation values ​​of the device under test for 10-60 minutes;

[0081] Step S13: receiving and decoding GNSS common frequency observation values ​​of the reference station for 10-60 minutes;

[0082] Step S2: single epoch estimation of DISB, obtaining a priori value of DISB according to the DISB time series; specifically, step S2 includes:

[0083] Step S21: pre-processing GNSS observation data;

[0084] Data preprocessing includes but is not limited to monitoring and eliminating abnormal data, detecting and repairing cycle slips, single-point positioning of monitoring stations and reference stations, setting of satellite cutoff elevation angle, setting of data cutoff signal-to-noise ratio, setting of maximum time difference of asynchronous differential, setting of reference satellite selection strategy, and weighting according to elevation angle.

[0085] Preferably, the height angle weighting formula used is:

[0086]

[0087] Among them, θ i represents the altitude angle of the i-th satellite, σ0 is the a priori unit weighted mean error, and σ is the a priori standard deviation of the i-th satellite observation value.

[0088] Step S22: constructing a double difference observation equation;

[0089] First, it is known that the raw GNSS observations can be written as

[0090]

[0091]

[0092] Among them, the subscripts s, K, j, and r represent the satellite, the satellite system, the frequency, and the receiver respectively; φ and p represent the carrier phase and pseudorange observation values, respectively, in meters; ρ, τ, and ι represent the satellite-to-earth distance, tropospheric delay, and ionospheric delay, respectively; in represents frequency j; dt r , They represent the receiver clock error and satellite clock error respectively; Indicates the phase hardware delay of the receiver end and the phase hardware delay of the satellite end; represents the pseudorange hardware delay at the receiver end and the pseudorange hardware delay at the satellite end; λ j Indicates wavelength; Indicates ambiguity; and Represent the observation noise of phase and pseudorange respectively. It should be emphasized that the subscripts and subscripts of the parameters reflect the factors that affect the parameters.

[0093] If there is a satellite v belonging to system Q at this time, and satellite v also broadcasts a signal on frequency j, then the double difference observation between satellites s and v is:

[0094]

[0095]

[0096] For the convenience of derivation, we take the case of dual-system single-frequency short baseline as an example, that is, ignoring the influence of ionospheric and tropospheric delays, and assuming that n and m satellites in systems Q and K are observed respectively. Then the single-epoch TRTK observation equation can be written as:

[0097]

[0098] in, With a similar structure, is the coordinate correction value to be estimated.

[0099] Step S23: estimating parameter floating point solution;

[0100] For the above formula (4), we can use

[0101]

[0102] Make equivalent changes to obtain the relationship between TRTK and LRTK

[0103]

[0104] in

[0105]

[0106] This is the observation equation of a single epoch of LRTK. It can be seen that TRTK introduces three parameters to be estimated while introducing two new observation values ​​compared to LRTK. It seems that TRTK has fewer redundant observations, but fortunately, since the DISBs parameters have very good stability in the time domain, and the ambiguity parameters also maintain constant characteristics in the time domain when there is no cycle slip. Therefore, as the observation time accumulates, TRTK will have more and more redundant observation information than LRTK. However, due to the and Therefore, these two parameters cannot be estimated correctly at the same time. In order to correctly calculate the value of phase DISB, we may as well and Merge parameters. To replace the combined parameters, equation (5) can be written as

[0107]

[0108] Finally, the floating point values ​​of all parameters in each epoch can be calculated using least squares.

[0109] Step S24: Obtain DISB parameters;

[0110] The DISB parameter is defined as and However, due to the parameter estimation and Cannot be separated, It is impossible to estimate, so the usual practice is to The fractional part First, calculate the equivalent phase DISB in each epoch. and In order to reduce the impact of observation noise on DISB, the final DISB can be obtained by averaging over a long period of time.

[0111] Step S3: receiving and decoding the IMU observations and GNSS observations generated by the integrated navigation module, and the GNSS corrections broadcast by the reference station; specifically, step S3 includes:

[0112] Step S31: Receive and decode IMU observations and provide them to S4.

[0113] Step S32: Receive and decode GNSS observations in the combined system and provide them to S4.

[0114] Step S33: Receive and decode the reference station GNSS correction data and provide it to S4.

[0115] Step S4: using the DISB information obtained by the pre-calibration, according to the combined positioning algorithm embedded in the combined navigation module, perform GNSS / INS loose combination or tight combination solution, and enter S45; specifically, step S4 includes:

[0116] Perform data preprocessing on GNSS and IMU data, INS initialization, INS mechanical arrangement, construction of GNSS / INS loose / tight combination equations, filtering and solving of estimated parameters, and ambiguity parameter fixing and transfer.

[0117] Step S41: pre-processing IMU data, sequentially performing (including but not limited to) monitoring and eliminating abnormal data, unit conversion, and coordinate conversion; and providing to S42;

[0118] At the same time, the GNSS observation data is preprocessed, and the following operations are performed in sequence (including but not limited to) monitoring and elimination of abnormal data, detection and repair of cycle slips, single-point positioning of monitoring stations and reference stations, setting of satellite cutoff altitude angle, setting of data cutoff signal-to-noise ratio, setting of maximum time difference of asynchronous differential, setting of reference satellite selection strategy, and weighting according to altitude angle; and provided to S42.

[0119] Step S42: Initialize INS and enter S43:

[0120] The approximate position, velocity and attitude of the combined system are calculated through GNSS SPP, and the initial variance-covariance matrix and process noise matrix of the filtering system are set.

[0121] Step S43: Perform INS mechanical arrangement and enter S44. This technology is the basic content of GNSS / INS integrated navigation technology and can be found in existing literature and open source programs, so it will not be repeated here.

[0122] Step S44: Combine the ambiguity parameters fixed in S46 to construct the GNSS / INS loose / tight combination equation, including:

[0123] According to the relationship between space vectors, we can get:

[0124]

[0125] The subscripts b and superscripts e represent the IMU's own coordinate system and the Earth-centered Earth-fixed coordinate system, respectively. represents the coordinates of the phase center of the GNSS antenna in the e system, Represents the coordinates of the IMU center of mass in the e system, represents the coordinate transformation matrix from b system to e system, l b represents the spatial distance between the GNSS antenna phase center and the IMU centroid in the IMU coordinate system, which can be regarded as a constant. The above formula describes the relationship between the position solved by GNSS and the position solved by INS under ideal conditions. Since the solution process will inevitably introduce errors in practice, the error perturbation analysis of the above formula (8) can be used to obtain the relationship between the position error solved by GNSS and the position error solved by INS.

[0126]

[0127] Where δ represents the error and φ represents the attitude error. Then the GNSS / INS tight combination equation based on TRTK can be written as

[0128]

[0129] Similarly, the loosely combined equation of GNSS / INS based on TRTK can be written as

[0130]

[0131] Where (~) represents the observed value, represents the speed error, b ω ,b f Represent the zero bias of the gyroscope and accelerometer, s ω ,s f Represent the scale factors of the gyroscope and accelerometer respectively.

[0132] Step S45: Filter and solve the estimated parameters, output the final position, speed, and attitude settlement results, and provide the settlement results to S46 for further iteration:

[0133] Different random walk models are established according to the properties of the parameters, and Kalman filtering is used to solve all the parameters to be estimated.

[0134] Step S46: Fixing and transferring fuzziness parameters:

[0135] The LAMBDA method is used to fix the ambiguity parameters. After the ambiguity is correctly fixed, the fixed ambiguity is obtained. At the same time, the fixed ambiguity parameters are recorded and used in subsequent epochs until all data are resolved.

[0136] In summary, the GNSS / INS combined positioning method of the present invention based on the difference of common frequency signals between systems can effectively improve the positioning performance of the combined system in complex environments by deeply mining the information of multi-system GNSS and creating new observation information for the GNSS / INS combined system.

[0137] Specifically, compared with the prior art, the present invention has the following advantages:

[0138] The existing GNSS / INS combined positioning algorithm is heavily dependent on GNSS information. When the GNSS signal is interfered or blocked, it will not be able to suppress the accumulation of INS errors, resulting in a decrease in the navigation performance of the combined system. The existing GNSS / INS combined positioning improvement technology requires the introduction of new external sensors, which on the one hand increases the hardware cost of the combined system, and on the other hand further increases the difficulty of implementing the fusion algorithm. The present invention fully exploits the information of multi-system GNSS, utilizes the superior stability of DISB in the time domain, significantly increases the amount of available information of the combined system, and can effectively improve the positioning performance of the GNSS / INS combined system in complex environments.

[0139] The above is only a preferred embodiment of the present invention and does not limit the present invention in any way. Any technician in the relevant technical field, without departing from the scope of the technical solution of the present invention, makes any form of equivalent replacement or modification to the technical solution and technical content disclosed in the present invention, which does not depart from the content of the technical solution of the present invention and still falls within the protection scope of the present invention.

Claims

1. A GNSS / INS combined positioning method based on common frequency signal difference between systems, characterized in that: The steps include: Step S1: Place the device to be tested and the reference station at the two ends of the ultra-short baseline respectively, keep them stationary and synchronously collect common frequency data for 10-60 minutes; Step S2: single epoch estimation of DISB, and obtaining the prior value of DISB based on the DISB time series; Step S3: receiving and decoding the IMU observations and GNSS observations generated by the integrated navigation module, and the GNSS corrections broadcast by the reference station; Step S4: using the DISB information obtained by the pre-calibration, according to the combined positioning algorithm embedded in the integrated navigation module, perform GNSS / INS loose combination or tight combination solution, filter solution of the parameters to be estimated, and fix and transfer the ambiguity parameters; The step S2 comprises: Step S21: pre-processing GNSS observation data; Data preprocessing includes but is not limited to monitoring and eliminating abnormal data, detecting and repairing cycle slips, single-point positioning of monitoring stations and reference stations, setting of satellite cutoff elevation angle, setting of data cutoff signal-to-noise ratio, setting of maximum time difference of asynchronous differential, setting of reference satellite selection strategy, and weighting according to elevation angle; The height angle weighting formula used is: Among them, θ i represents the altitude angle of the i-th satellite, σ0 is the prior unit weighted mean error, and σ is the prior standard deviation of the i-th satellite observation value; Step S22: constructing a double difference observation equation; First, the known GNSS raw observations are written as The subscripts s, K, j, and r represent the satellite, the satellite system, the frequency, and the receiver, respectively; φ and p represent the carrier phase and pseudorange observation values, respectively, in meters; ρ, τ, and ι represent the satellite-to-earth distance, tropospheric delay, and ionospheric delay, respectively; μ j =(f1 2 ) / (f j 2 ), where f j 2 represents frequency j; dt r , They represent the receiver clock error and satellite clock error respectively; Indicates the phase hardware delay of the receiver end and the phase hardware delay of the satellite end; represents the pseudorange hardware delay at the receiver end and the pseudorange hardware delay at the satellite end; λ j Indicates wavelength; Indicates ambiguity; and Respectively represent the observation noise of phase and pseudorange; the upper and lower scripts of the parameters reflect the factors that affect the parameters; If there is a satellite v belonging to system Q at this time, and satellite v also broadcasts a signal on frequency j, then the double difference observation between satellites s and v is: Ignoring the effects of ionospheric and tropospheric delays, and assuming that n and m satellites in systems Q and K are observed respectively, the single epoch TRTK observation equation is written as: in, With a similar structure, is the coordinate correction value to be estimated; Step S23: estimating parameter floating point solution; For formula (4), use Make equivalent changes to obtain the relationship between TRTK and LRTK in This is the observation equation of a single epoch of LRTK. Compared with LRTK, TRTK introduces two new observation values ​​and also introduces three parameters to be estimated. As the observation time accumulates, TRTK will have more and more redundant observation information than LRTK. However, due to the and In order to correctly calculate the value of phase DISB, and Merge parameters; use To replace the combined parameters, equation (5) can be written as Finally, the floating point values ​​of all parameters in each epoch are calculated using least squares; Step S24: Obtain DISB parameters; The DISB parameter is defined as and However, due to the parameter estimation and Cannot be separated, It is impossible to estimate, so the usual practice is to The fractional part is considered as the equivalent phase DISB; first calculate the and In order to reduce the influence of observation noise on DISB, the final DISB is obtained by taking an average over a long period of time; The step S4 comprises: Step S41: pre-processing IMU data, monitoring and eliminating abnormal data, unit conversion, coordinate conversion in sequence; providing to S42; At the same time, the GNSS observation data is pre-processed, and abnormal data monitoring and elimination, cycle slip detection and repair, single-point positioning of monitoring stations and reference stations, satellite cutoff elevation angle setting, data cutoff signal-to-noise ratio setting, asynchronous differential maximum time difference setting, reference satellite selection strategy setting, and weighting according to elevation angle are performed in sequence; and provided to S42; Step S42: Initialize INS: calculate the approximate position, velocity and attitude of the combined system through GNSS SPP, set the initial variance covariance matrix and process noise matrix of the filter system, and enter S43; Step S43: Perform INS mechanical arrangement and enter S44; Step S44: Combine the fixed ambiguity parameters to construct the GNSS / INS loose / tight combination equation, including: According to the relationship between space vectors, we get: The subscripts b and e represent the IMU's own coordinate system and the Earth-centered Earth-fixed coordinate system, respectively. represents the coordinates of the phase center of the GNSS antenna in the e system, Represents the coordinates of the IMU center of mass in the e system, represents the coordinate transformation matrix from b system to e system, l b represents the spatial distance between the GNSS antenna phase center and the IMU centroid in the IMU coordinate system, which is a constant. Equation (8) describes the relationship between the position solved by GNSS and the position solved by INS under ideal conditions. Error perturbation analysis is performed on equation (8) to obtain the relationship between the position error solved by GNSS and the position error solved by INS: Where δ represents the error and φ represents the attitude error; then the GNSS / INS tight combination equation based on TRTK is written as The loose combination equation of GNSS / INS based on TRTK is written as Where (~) represents the observed value, represents the speed error, b ω ,b f Represent the zero bias of the gyroscope and accelerometer, s ω ,s f Represent the scale factors of the gyroscope and accelerometer respectively; Go to step S45; Step S45: Filter and solve the estimated parameters, output the final position, speed, and attitude settlement results, and provide the settlement results to S46 for further iteration: Different random walk models are established according to the properties of the parameters, and Kalman filtering is used to solve all the parameters to be estimated; Go to step S46; Step S46: Fixing and transferring fuzziness parameters: The LAMBDA method is used to fix the ambiguity parameters. After the ambiguity is correctly fixed, the fixed ambiguity is obtained. At the same time, the fixed ambiguity parameters are recorded, and S44 is executed in subsequent epochs until all data are resolved.

2. The GNSS / INS combined positioning method of inter-system common frequency signal difference according to claim 1, characterized in that: The step S1 comprises: Step S11: Arrange the device to be tested and the reference station as an ultra-short baseline; Step S12: receiving and decoding GNSS common frequency observation values ​​of the device under test for 10-60 minutes; Step S13: Receive and decode the GNSS common frequency observation values ​​of the reference station for 10-60 minutes.

3. The GNSS / INS combined positioning method of inter-system common frequency signal difference according to claim 1, characterized in that: The step S3 comprises: Step S31: Receive and decode IMU observations and provide them to S4; Step S32: receiving and decoding GNSS observation values ​​in the combined system and providing them to S4; Step S33: Receive and decode the reference station GNSS correction data and provide it to S4.

Images