Navigation methods and navigation devices

Abstract

This disclosure relates to navigation methods and navigation devices. The navigation device includes an inertial navigation system (101) coupled to a satellite navigation system (100), wherein information supplied by the satellite positioning system is used to adjust the inertial navigation system. The device is characterized by further including means (100) for measuring signals from base stations of a wireless cellular network. The satellite navigation system and the means for measuring signals from base stations of the wireless cellular network are coupled to each other via tight coupling (100) to form a satellite navigator and / or a base station navigator, thereby realizing an estimator with simultaneous positioning and mapping. The satellite navigation system and the means for measuring signals from base stations of the wireless cellular network are coupled to the inertial navigation system via loose coupling (102) to adjust the inertial navigation system.

Description

Technical Field

[0001] This invention relates to navigation methods and systems. Background Technology

[0002] Currently, the use of inertial navigation systems (known as INS systems) is commonplace, especially in the fields of air and sea navigation. Therefore, INS systems are routinely installed in vehicles such as aircraft or ships. INS systems use various sensors, such as accelerometers or gyroscopes. These sensors provide measurements (measurements of rotation, acceleration, etc.), and the integration of these sensors makes it possible to determine the vehicle's attitude, velocity, and position during movement.

[0003] However, these sensors are imperfect and have inherent errors or measurement biases that are prone to change during movement. Furthermore, these sensors are susceptible to measurement noise. One consequence of these imperfections is that INS systems may prove accurate in the short term, but will experience significant drift in the longer term.

[0004] To overcome these drifts, some INS systems are coupled with GNSS-type satellite positioning systems (Global Navigation Satellite Systems, hereinafter referred to as GNSS systems), such as GPS (Global Positioning System), GLONASS (Russian Satellite Navigation System), Galileo (the future European Satellite Navigation System), or Beidou (the future Chinese Satellite Navigation System). This results in what is known as a GNSS / INS hybrid system, as described in document EP2245479.

[0005] In GNSS / INS hybrid systems, information supplied by the GNSS system is typically used to adjust the INS system. This adjustment is usually performed using a Kalman filter, which optimizes navigation performance by estimating the corresponding errors of the GNSS and INS systems.

[0006] Then, distinguish between two types of coupling: loose coupling and tight coupling.

[0007] In loose coupling, in addition to receiving the navigation solution (position, velocity) from the INS system, the Kalman filter also receives the navigation solution (position, velocity) supplied by the GNSS system as input.

[0008] In tight coupling, instead of the aforementioned navigation solution, the GNSS system supplies input of raw data (e.g., pseudo-range or pseudo-velocity) to the Kalman filter.

[0009] Currently, most GNSS / INS hybrid navigation systems use loose coupling because it is simple to implement.

[0010] However, in order to provide a navigation solution, a GNSS system must acquire signals from at least a predetermined number (usually equal to four) of satellites (referred to as GNSS satellites). In some environments (such as urban environments), satellite signal reception is perturbed, which can at least temporarily prevent the GNSS system from acquiring the predetermined number of GNSS satellites necessary to determine a navigation solution. In particular, these perturbations involve GNSS satellite shielding, multipath propagation, and interference. When a GNSS system encounters such perturbations, it may be unable to provide a navigation solution at the input of the Kalman filter. Without the assistance of the GNSS system, the GNSS / INS hybrid system subsequently becomes a simple INS system.

[0011] One advantage of tightly coupled GNSS / INS hybrid systems is that even when the number of GNSS satellites available to the GNSS system is less than the predetermined number, the GNSS / INS hybrid system can benefit from the availability of raw data supplied by the GNSS system to determine improved navigation solutions (compared to an INS-only system). However, the management of raw GNSS data (e.g., calculations for clock corrections of GNSS satellites, and calculations for position and velocity corrections from ephemeris data) makes tightly coupled GNSS / INS hybrid systems fundamentally more complex than loosely coupled GNSS / INS hybrid systems.

[0012] It is desired to overcome these shortcomings of existing technologies. In particular, it is desired to propose a method and system that enables the overcoming of the deficiencies of loosely coupled GNSS / INS hybrid navigation systems while maintaining less complexity than tightly coupled GNSS / INS hybrid navigation systems. Summary of the Invention

[0013] According to a first aspect, the present invention relates to a navigation device comprising an inertial navigation system coupled to a satellite navigation system, wherein information supplied by the satellite positioning system is used to adjust the inertial navigation system, characterized in that the device further comprises means for measuring signals from base stations of a wireless cellular network, wherein the satellite navigation system and the means for measuring signals from base stations of the wireless cellular network are coupled to each other by tight coupling to form a satellite navigator and / or a base station navigator, thereby realizing an estimator with simultaneous positioning and mapping, and wherein the satellite navigation system and the means for measuring signals from base stations of the wireless cellular network are coupled to the inertial navigation system by loose coupling to adjust the inertial navigation system.

[0014] The present invention also relates to a navigation method comprising an inertial navigation system coupled to a satellite navigation system, wherein information supplied by the satellite positioning system is used to adjust the inertial navigation system, characterized in that the method comprises the following steps:

[0015] - Measure the signal from the base station of the wireless cellular network;

[0016] - By tightly coupling the satellite navigation system and the means for measuring signals from the base station of the wireless cellular network to form a satellite navigator and / or a base station navigator, an estimator with simultaneous positioning and mapping is realized;

[0017] - The satellite navigation system and the means for measuring signals from the base station of the wireless cellular network are coupled to the inertial navigation system by loose coupling to adjust the inertial navigation system.

[0018] Therefore, the present invention makes it possible to potentially have the maximum available information (GNSS and / or 5G) within a hybrid architecture based on loosely coupled pseudo-range information of adjusted inertial navigation and relying on a tightly coupled GNSS / 5G structure, which allows for the merging of raw information from the two satellite sources (via GNSS and 5G receivers) to be optimized to the greatest extent possible.

[0019] According to a specific embodiment of the present invention, the satellite navigator and / or base station navigator operates in three operating modes: a first mode (referred to as the mapping mode) is characterized in that pseudo-range, pseudo-velocity, and carrier-to-noise ratio of signals from the base station of the wireless cellular network and pseudo-range, pseudo-velocity, and carrier-to-noise ratio of signals from the satellite are available; a second mode (referred to as the 5G mode) is characterized in that pseudo-range, pseudo-velocity, and carrier-to-noise ratio of signals from the base station of the wireless cellular network are available; and a third mode (referred to as the standard mode) is characterized in that pseudo-range, pseudo-velocity, and carrier-to-noise ratio of signals from the satellite are available.

[0020] Therefore, the highest quality of continuous navigation is achieved when only GNSS is available, or when only cellular networks are available, or when both GNSS and cellular networks are available.

[0021] According to a specific embodiment of the present invention, the device further includes a matrix generation module for forming a plurality of matrices, the plurality of matrices including a transition matrix, a Jacobian matrix, a measurement noise covariance matrix, and a process noise covariance matrix.

[0022] Therefore, according to a specific embodiment of the present invention, for each operating mode, the matrix generation module forms a transition matrix, a Jacobian matrix, a measurement noise covariance matrix, and a process noise covariance matrix that are different from those formed for other modes.

[0023] According to a specific embodiment of the present invention, the navigation device includes a Kalman filter.

[0024] According to a specific embodiment of the present invention, the method includes the following steps: determining the operating mode of the satellite navigator and / or the base station navigator in three operating modes, wherein the first mode is a mode in which pseudo-range, pseudo-velocity, and carrier-to-noise ratio of the signal from the base station of the wireless cellular network and pseudo-range, pseudo-velocity, and carrier-to-noise ratio of the signal from the satellite are available; the second mode is a mode in which pseudo-range, pseudo-velocity, and carrier-to-noise ratio of the signal from the base station of the wireless cellular network are available; and the third mode is a mode in which pseudo-range, pseudo-velocity, and carrier-to-noise ratio of the signal from the satellite are available.

[0025] According to a specific embodiment of the present invention, the method further includes a matrix generation step of forming a plurality of matrices, the plurality of matrices including a transition matrix, a Jacobian matrix, a measurement noise covariance matrix, and a process noise covariance matrix. Attached Figure Description

[0026] The features of the invention mentioned above, as well as other features, will become clearer from the following description of exemplary embodiments, which is made with reference to the accompanying drawings, in which:

[0027] Figure 1 An example is shown of the environment in which the invention is implemented;

[0028] Figure 2 A hybrid navigation system with loose coupling between an INS system and a tightly coupled GNSS / 5G system, according to the present invention, is illustrated schematically.

[0029] Figure 3 A tightly coupled GNSS / 5G system is illustrated schematically;

[0030] Figure 4 A GNSS system is illustrated schematically;

[0031] Figure 5 The 5G system is illustrated schematically;

[0032] Figure 6 A GNSS / 5G detector is illustrated schematically;

[0033] Figure 7 A GNSS / 5G navigator is illustrated schematically;

[0034] Figure 8 The INS navigation system is illustrated schematically.

[0035] Figure 9 A loosely coupled system is illustrated schematically;

[0036] Figure 10This schematically illustrates a loosely coupled system between an INS system and a tightly coupled GNSS / 5G system;

[0037] Figure 11 An example of the hardware architecture of the processing module is illustrated schematically;

[0038] Figure 12 The method for selecting the navigator's operating mode, implemented by the processing module, is illustrated schematically.

[0039] Figure 13 A method for implementing a navigator using a loosely coupled system between an INS system and a tightly coupled GNSS / 5G system is illustrated schematically. Detailed Implementation

[0040] The invention is described below in a marine environment, wherein the navigation system implementing the method according to the invention is transported by a ship-type vehicle or transport vehicle. The invention can also be applied to different environments (such as aerial or road environments), and the vehicle transporting the navigation system according to the invention can be an aircraft, drone, car, or truck.

[0041] Figure 1 An environment in which the invention is implemented is illustrated. Figure 1 In this context, vessel 1 moves in an environment that includes satellites of a GNSS satellite constellation (satellites 3A, 3B, and 3C are shown here) and base stations of a fifth-generation cellular wireless communication system, referred to as 5G base stations (two 5G base stations are shown here). Vessel 1 includes a hybrid navigation system 10, which has loose coupling between an INS system and a tightly coupled GNSS / 5G hybrid system, hereinafter referred to as the GNSS / 5G+INS hybrid navigation system.

[0042] In one implementation, the base station may also be a base station for a fourth-generation telephone system (referred to as a 4G base station, or a combination of 5G base stations and 4G base stations).

[0043] Figure 2 A GNSS / 5G+INS hybrid navigation system 10 is illustrated schematically.

[0044] Figure 2 The GNSS / 5G+INS hybrid navigation system 10 includes a tightly coupled GNSS / 5G system 100, an INS system 101, and a loosely coupled system 102. The various couplings are described below.

[0045] In one implementation, the loosely coupled system 102 receives calibrated measurement results from the INS system 101 and a navigation solution from the GNSS / 5G tightly coupled system 100.

[0046] In another embodiment, the loosely coupled system 102 receives measurement results from the INS system 101 and navigation solutions from the GNSS / 5G tightly coupled system 100, and generates residual errors as input measurement results based on this information.

[0047] Figure 3 A tightly coupled GNSS / 5G system 100 is schematically illustrated.

[0048] The GNSS / 5G tightly coupled system 100 includes a GNSS sensor 1000, a 5G sensor 1001, a GNSS / 5G detector 1002, and a GNSS / 5G navigator 1003.

[0049] The GNSS / 5G detector 1002 determines whether the GNSS / 5G tightly coupled system 100 can use only GNSS data, only 5G data, or both 5G and GNSS data based on the output data it receives. Therefore, the GNSS / 5G detector 1002 defines a mapping operation mode, a 5G operation mode, or a standard operation mode.

[0050] The GNSS / 5G navigator 1003 is based on a SLAM Kalman filter, which is configured according to the operating mode. Therefore, the GNSS / 5G navigator 1003 defines the state and observation vectors, the Kalman filter matrix, and more specifically their dimensions for each mode.

[0051] The main advantage of the GNSS / 5G tightly coupled system 100 is its tight coupling between the GNSS sensor 1000 and the 5G sensor 1001, thus allowing the deployment of a SLAM (Simultaneous Localization and Mapping) estimator in a navigator shared by the GNSS sensor 1000 and the 5G sensor 1001 (i.e., the GNSS / 5G navigator 1003). The SLAM estimator is capable of simultaneously estimating the state of the 5G base station (i.e., position and clock state) and the navigation solution of the GNSS / 5G tightly coupled system 100 (i.e., position, velocity, and clock state). Therefore, as long as the GNSS / 5G tightly coupled system 100 captures at least a predetermined number N... sat GNSS satellites (generally N sat =4), which is sufficient to estimate the state (location and clock state) of the 5G base station. Therefore, the GNSS / 5G tightly coupled system 100 does not require external additional devices for determining the location and clock state of the 5G base station. Once the state of the 5G base station is known, when the number of satellites acquired by the GNSS / 5G tightly coupled system 100 is less than N, sat At that time, the tightly coupled GNSS / 5G system can provide a navigation solution by using measurement results from the 5G sensor 1001 to compensate for the absence of GNSS signals.

[0052] Next, the navigation solution determined by the tightly coupled GNSS / 5G system 100 is combined with the measurement results from the INS system 101 in the loosely coupled system 102 to produce an improved navigation system and correct any errors in the INS system 101.

[0053] Figure 4 A GNSS sensor 1000 is schematically illustrated. The GNSS sensor 1000 receives radio frequency (RF) signals 1000A from a GNSS satellite. For each RF signal of sufficient quality received from the GNSS satellite, the GNSS sensor 1000 generates output data 1000B (referred to as GNSS output data) and an estimate of the position of the GNSS satellite 1000C. The output data 1000B includes pseudo-range and pseudo-velocity (referred to as GNSS pseudo-range and pseudo-velocity) and carrier-to-noise ratio C / N0 (referred to as GNSS C / N0 ratio) corresponding to the RF signals.

[0054] Figure 5 A 5G sensor 1001 is schematically illustrated. The 5G sensor 1001 receives RF signals 1001A (referred to as 5G signals) from a 5G base station. For each RF signal of sufficient quality received from the 5G base station, the 5G sensor 1001 generates output data (referred to as 5G output data), which includes estimates 1001B of pseudo-range and pseudo-velocity (referred to as 5G pseudo-range and pseudo-velocity) and carrier-to-noise ratio C / N0 (referred to as 5G C / N0 ratio) corresponding to the RF signals.

[0055] Figure 6 The GNSS / 5G detector 1002 is illustrated schematically.

[0056] GNSS / 5G detector 1002 receives GNSS and 5G output data, the output data including:

[0057] -Estimation results of GNSS pseudo range and pseudo velocity and GNSS C / N0 ratio 1000B;

[0058] - Estimated positions of GNSS satellites (1000C); and

[0059] -5G pseudo range and pseudo speed, and 5G C / N0 ratio estimation results 1001B.

[0060] The GNSS / 5G detector 1002 then performs two roles. Its first role is to synchronize GNSS output data with 5G output data when both types of data are available. For example, if a number of 5G output data points are collected between two collections of GNSS output data, the 5G output data are averaged to reflect the average 5G output data over the time interval corresponding to the two collections of GNSS output data. The second role of the GNSS / 5G detector 1002 is to determine whether the tightly coupled GNSS / 5G system 100 can use only GNSS data, only 5G data, or both 5G and GNSS data, based on the output data it receives. The types of data used allow for defining the operating mode of the GNSS / 5G+INS hybrid system 10 across multiple possible operating modes. These multiple operating modes include a so-called mapping mode, a so-called 5G mode, and a so-called standard (GNSS) mode. In the mapping mode, a sufficient number of GNSS and 5G pseudo-range, pseudo-velocity, and C / N0 ratios are available.

[0061] In 5G mode, only 5G pseudo-distance, pseudo-speed, and C / N0 ratio are available.

[0062] In standard mode, only GNSS pseudo range, pseudo velocity, and C / N0 ratio are available.

[0063] Then, the GNSS / 5G detector 1002 supplies the following as outputs: output measurement results 1002A of GNSS and / or 5G pseudo velocity and pseudo range of the GNSS / 5G tightly coupled system 100, estimated results 1002B of GNSS and / or 5G / N0 ratio corresponding to measurement results 1002A, positions of GNSS satellites corresponding to measurement results 1002A 1002C, and identifiers 1002D indicating the data (GNSS or 5G or GNSS and 5G) to be used by the GNSS / 5G tightly coupled system 100.

[0064] Figure 7 The GNSS / 5G navigator 1003 is illustrated schematically.

[0065] The use of GNSS / 5G navigator 1003 will be referenced. Figure 12 The algorithm described.

[0066] As mentioned above Figure 2 As mentioned, the GNSS / 5G navigator 1003 includes a SLAM estimator. An extended Kalman filter (EKF) (hereinafter referred to as an EKF filter) is typically used to solve the SLAM problem. Therefore, the GNSS / 5G navigator 1003 includes a module (referred to as EKF-SLAM module 10030) that solves the SLAM problem of this invention.

[0067] The EKF-SLAM module 10030 includes a matrix generation module 10031, a correction module 10032 called the EKF correction module, and a prediction module 10033 called the EKF prediction module.

[0068] The matrix generation module 10031 is responsible for generating and using the satellite's position, the position and velocity of the GNSS / 5G tightly coupled system 100, and the position of the 5G base station. Multiple matrices are used in the navigation method described below. These multiple matrices include:

[0069] - Transition matrix F;

[0070] - Jacobian matrix H;

[0071] -Measurement noise covariance matrix R;

[0072] -Process-noise covariance matrix Q.

[0073] The matrix generation module 10031 receives the following as inputs: the estimated result of the GNSS and / or 5G C / N0 ratio 1002B, the position of the GNSS satellite 1002C, and the corrected estimate result of the GNSS and / or 5G pseudo velocity and pseudo range of the GNSS / 5G tightly coupled system 100 supplied by the EKF correction module 10032.

[0074] Based on the data received as input, the matrix generation module 10031 supplies the Jacobian matrix H at the input of the EKF correction module 10032, and supplies matrices F, H, R, and Q at the input of the EKF prediction module 10033. Figure 7 (Ref. c)

[0075] EKF correction module 10032 receives the following as inputs: GNSS and / or 5G pseudo-velocity and pseudo-range estimation results 1002A of the GNSS / 5G tightly coupled system 100, an identifier 1002D indicating the data to be used by the GNSS / 5G tightly coupled system 100, Jacobian matrix H, prediction f of the error state vector of the GNSS / 5G tightly coupled system 100 from EKF prediction module 10033, and prediction g of the error covariance matrix associated with the state vector.

[0076] Based on this data, the EKF correction module 10032 supplies the EKF prediction module 10033 with a posterior estimate d of the error state vector of the GNSS / 5G tightly coupled system (i.e., after correction) and a posterior estimate e of the error covariance matrix associated with the state vector. Furthermore, the EKF correction module 10032 generates an identifier 1003A indicating the operating mode of the GNSS / 5G tightly coupled system 100 and a navigation solution 1003B (referred to as the GNSS / 5G navigation solution).

[0077] The EKF prediction module 10033 receives the following as inputs: matrices F, H, R, and Q (refer to c), a posterior estimate d of the error state vector of the tightly coupled GNSS / 5G system, and a posterior estimate e of the error covariance matrix associated with the state vector.

[0078] Based on this data, the EKF prediction module 10033 supplies the EKF correction module 10032 with a prediction f of the error state vector of the GNSS / 5G tightly coupled system 100 and a prediction g of the error covariance matrix associated with the state vector.

[0079] Figure 8 The INS navigation system 101 is shown schematically.

[0080] INS Navigation System 101 usage will refer to Figure 13 The algorithm described.

[0081] The INS system 101 includes an INS sensor 1010 and an INS navigator 1011. The INS sensor 1010 includes, for example, at least one accelerometer and at least one gyroscope, each of which provides measurement results to the INS navigator 1011.

[0082] Similar to a conventional INS navigator, the INS navigator 1011 receives measurement results from the INS sensor 1010 and generates a navigation solution 101A (referred to as the INS navigation solution). However, the INS navigator 1011 also receives position and velocity correction results 101D, attitude correction results 101C, and (deviation, accelerometer, and gyroscope scaling factor) correction results 101B from the loosely coupled system 102, to apply to the measurement results supplied by the INS sensor 1010.

[0083] Figure 9 A loosely coupled system 102 is illustrated schematically.

[0084] The loosely coupled system 102 receives an identifier 1003A indicating the operating mode of the GNSS / 5G tightly coupled system 100, as well as a GNSS / 5G hybrid navigation solution 1003B supplied by the GNSS / 5G tightly coupled system 100. The loosely coupled system also receives an INS navigation solution 101A supplied by the INS system 101.

[0085] Based on this data, the loosely coupled system 102 supplies the correction results 101B, 101C, and 101D to the INS system 101 and generates a navigation solution 102A (referred to as the GNSS / 5G+INS hybrid navigation solution). The GNSS / 5G+INS hybrid navigation solution is the navigation solution supplied by the GNSS / 5G+INS hybrid navigation system 10.

[0086] Figure 10 The details of the loosely coupled system 102 are illustrated schematically.

[0087] The loosely coupled system 102 includes a synchronization module 1020 and a navigator 1021 (referred to as a GNSS / 5G+INS navigator). The synchronization module 1020 is responsible for supplying navigation solution 102B to the GNSS / 5G+INS navigator 1021. Navigation solution 102B is obtained by the synchronization module 1020 based on navigation solutions 101A and 1003B from the INS system 101 and the GNSS / 5G navigator 1003, respectively. However, the INS system 101 and the tightly coupled GNSS / 5G system 100 typically do not generate navigation solutions with the same frequency. The synchronization module 1020 is then responsible for synchronizing the INS navigation solution with the GNSS / 5G navigation solution. For example, when the INS navigation solution is supplied at a higher frequency than the GNSS / 5G navigation solution, the synchronization module 1020 calculates the GNSS / 5G solution between the two GNSS / 5G solutions supplied by the tightly coupled GNSS / 5G system 100, each intermediate GNSS / 5G solution being time-aligned with the INS solution supplied by the INS system 101. Based on these synchronized INS and GNSS / 5G navigation solutions, the GNSS / 5G+INS navigator 1021 generates correction results 101B, 101C, and 101D, as well as a GNSS / 5G+INS hybrid navigation solution 102A.

[0088] Figure 11 An example of the hardware architecture of the processing module 11 is illustrated schematically.

[0089] The modules included in the GNSS / 5G+INS hybrid navigation system 10 (including GNSS / 5G detector 1002, matrix generation module 10031, EKF correction module 10032, EKF prediction module 10033, synchronization module 1020, and GNSS / 5G+INS navigator 1021) all include a processing module 11.

[0090] according to Figure 11The hardware architecture shown as an example includes a processing module 11 that then comprises, connected by a communication bus 110: a processor or CPU (central processing unit) 111; random access memory (RAM) 112; read-only memory (ROM) 113; a storage unit (such as a hard disk) or storage medium reader (such as an SD (Secure Digital) card reader) 114; and at least one communication interface 115 that enables the processing module 11 to communicate with other modules in the GNSS / 5G+INS hybrid navigation system 10.

[0091] Processor 111 is capable of executing instructions loaded into RAM 112 from ROM 113, external memory (not shown), storage media (such as an SD card), or a communication network. When the GNSS / 5G+INS hybrid navigation system 10 is powered on, processor 111 is capable of reading instructions from RAM 112 and executing those instructions. These instructions form a computer program that causes processor 111 to perform or partially perform methods executed by the GNSS / 5G+INS hybrid navigation system module 10, which includes the processing module 11.

[0092] These methods can be implemented in software by executing a set of instructions by a programmable machine (such as a DSP (Digital Signal Processor) or a microcontroller), or in hardware by a machine or a dedicated component (such as a FPGA (Field Programmable Gate Array) or an ASIC (Application-Specific Integrated Circuit)).

[0093] For example, when the present invention is implemented on a vessel during port approach or side-by-side arrival, the measurement results determine the operating mode of the GNSS / 5G+INS hybrid receiver 10. For example, when the GNSS measurement results are partially masked and / or partially subjected to multipath, the standard mode is selected.

[0094] Select 5G mode when GNSS measurements are completely or very strongly masked and / or completely subjected to multipath.

[0095] Select INS / GNSS mode when 5G measurement results are completely and extremely masked and / or completely subjected to extreme multipath.

[0096] In the mapping mode, GNSS and 5G pseudorange are available. The state vector of the EKF-SLAM estimator is obtained based on the position, velocity, offset, and clock drift of the hybrid receiver and the position, offset, and clock drift of the 5G station.

[0097] Between corrections, the EKF-SLAM estimator generates state predictions using the receiver's transition model and clock model. and prediction error covariance When pseudo-range and pseudo-velocity measurements are available, the EKF-SLAM estimator generates an update to the state estimate. Update P with the associated estimation error covariance k .

[0098] In 5G mode, only 5G pseudo-range is available. The estimator's state vector is obtained based on the receiver's position and velocity, as well as the difference between the receiver's offset and clock drift on one side and the offset and clock drift of each 5G station on the other side.

[0099] The 5G mode considered here is a restricted mode because it uses the location of the 5G base station as well as the receiver bias and clock drift estimated in the mapping mode. Once the GNSS measurements become available again, the 5G station bias and clock drift, along with the location, are reintroduced into the state vector, and the mapping mode is entered again, allowing for a more refined determination of the 5G base station location.

[0100] Between corrections, the EKF-SLAM estimator generates state predictions using the transition model and clock model of the 5G base station. and prediction error covariance

[0101] In the GNSS mode known as the standard mode, only GNSS pseudorange is available. 5G measurement results are unavailable under environmental disturbances or disturbances related to the absence of a 5G network.

[0102] The state vector of the EKF-SLAM estimator is derived from the receiver's position, velocity, bias, and clock drift. Between corrections, the EKF-SLAM estimator generates state predictions using the receiver's transition model and clock model. and prediction error covariance When GNSS and / or 5G pseudorange and pseudovelocity measurements are available, the EKF-SLAM estimator generates an update to the state estimate. Update P with the associated estimation error covariance k .

[0103] Outside of the pattern, each 5G signal originates from a spatially fixed base station, and its state vector includes its three-dimensional location coordinates r. s =[x,y,z] T and its clock state Where c is velocity, δt s It is the clock skew of the 5G base station, and It's clock drift in 5G base stations.

[0104] The dynamic range of 5G signals is described by the following model: (x s )k+1 =F s (x s ) k +(w s ) k ,

[0105] in F s =diag[I 3x3 ,F Hor ], w s It is modeled as having zero mean and covariance Q. s =diag[0 3*3 ,c 2 Q hor,s The white noise process noise, And T is a constant sampling interval.

[0106] item and These are the process noise spectra of 5G station deviation and clock drift, respectively.

[0107] item and Can be compared with power-law coefficients Relatedly, these power-law coefficients were shown to be based on... The power spectral density is used to characterize the fractional frequency difference y(t) of the oscillator relative to the nominal frequency. A typical approximation uses only parameters h0 and h... -2 And therefore and

[0108] The state vector x of a tightly coupled GNSS / 5G system 100 r Including the three-dimensional position coordinates r of the system 100 r =[X,Y,Z] T The three-dimensional velocity of the system 100 and the clock state of the receiver

[0109]

[0110] The receiver's state changes over time in the following manner:

[0111] (x r ) k+1 =F r (x r ) k +(w r ) k ,

[0112] in T is the constant sampling interval of the measurement results, w r It is modeled as having zero mean and covariance Q. r =diag[Q pv ,c 2 Q hor,r The process noise vector model of white noise, where diag represents the diagonal cascade of blocks of matrix.

[0113]

[0114] S p It is the velocity noise power spectrum. and These are the power spectra of the receiver bias and the process noise of clock drift, respectively. and For example, for a clock using a quartz oscillator, h0 = 2 -19 seconds and h -2 =2 -20 Hz.

[0115] The pseudo-range observation of the nth 5G base station at time k, made by the GNSS / 5G receiver, is correlated with the state of the receiver and the state of the 5G base station:

[0116]

[0117] Where r r =[X,Y,Z] T It is the location of the receiver. It is the position of the nth 5G gNodeB. It is modeled as having zero mean and variance. Gaussian white noise.

[0118] The pseudo-velocity (δ-distance) observation of the nth 5G base station at time k, made by the GNSS / 5G receiver, is correlated with the state of the receiver and the state of the 5G base station:

[0119]

[0120] in It is the LOS vector of the 5G base station at the receiver. It is the receiver's predicted speed. It represents the zero speed of the 5G transmitter (because it is stationary within this framework), where c is the speed. It is the receiver's clock drift and This refers to the clock drift of the 5G base station, where λ is the wavelength of the nominal carrier frequency of the 5G base station. It has zero mean and zero variance. Gaussian white noise.

[0121] After correcting for ionospheric and tropospheric delays, the pseudo-range observation results of the m-th GNSS satellite taken by the GNSS / 5G receiver are as follows:

[0122]

[0123] in δt iono and δt tropo These are the ionospheric and tropospheric delays, respectively. It is an uncorrected pseudo-distance, and Modeled as having zero mean and variance Gaussian white noise.

[0124] The pseudo-velocity (δ-distance) observation of the m-th GNSS satellite at time k by the GNSS / 5G receiver is correlated to the receiver's state using the following formula:

[0125]

[0126] in It is the LOS vector of the satellite at the receiver. It is the receiver's predicted speed. It is the satellite's speed, where c is the velocity. It is the receiver's clock drift and It's the satellite's clock drift, λ L It is the wavelength of the satellite's nominal carrier frequency. It has zero mean and zero variance Gaussian white noise.

[0127] The navigation algorithm uses an extended Kalman filter as an estimator, which has increased complexity related to the fact that the base station's state... It is estimated simultaneously with the status of the GNSS / 5G receiver.

[0128] In the mapping mode, the extended Kalman filter (EKF) produces x. k Estimate Estimation of error covariance in And k is the final moment when the measurement is performed.

[0129] In the implementation of the EKF estimator, the receiver's position, velocity, offset, and clock drift errors, as well as the 5G base station's position, offset, and clock drift errors, are estimated. Subsequently, the state is updated at time k by adding the errors estimated at time k-1 to the state estimated at time k-1.

[0130]

[0131] The state vector of the EKF estimator at time k (x is the true vector of the parameters to be estimated, and) The estimate of the parameter vector is given by the following equation:

[0132]

[0133] Position, velocity, and clock state are defined using the usual added error. To simplify the notation, we should consider representing the estimated error state vector as a function of the posterior estimate. Or for prior estimates Instead of Δx k Then, the posterior estimate of the error state vector of the EKF estimator at time k is expressed as follows:

[0134]

[0135] In 5G mode, the measurement result z sv Unavailable. The receiver's clock state estimate is no longer being calculated. Location status of 5G base stations Instead, it estimates the deviation of the 5G clock relative to the receiver clock. (where n = 1 to N).

[0136] At time t0, the state is... and Removed from the estimator, and the estimate relative to the clock state. It is initialized in the following way:

[0137] Where n = 1 to N, and

[0138] The difference between the receiver's clock state vector and the base station's clock state vector According to the equation And change, of which w hor =w hor,r -w hor,s For having zero mean and covariance White noise.

[0139] Error when estimating the new state vector x' and the corresponding error covariance matrix Initialize as follows: Δx' = DΔx k And P' k =DP k D T ,

[0140] Where D is Δx k Transform to The matrix.

[0141] The new error state vector is represented as Make If vector Having dimension U*1 and vector If the dimension is V*1, then the D matrix has a dimension of V*U. D is calculated based on the number of available 5G base stations for each mode change.

[0142] Between the two correction times, the estimator uses the previously described dynamic range model to propagate the state estimate. or This provides a prediction of the corresponding error covariance matrix, and the observation model is used.

[0143] In the mapping mode, state prediction is performed:

[0144] in

[0145] in

[0146] Notice, Where F s =diag[I 3×3 ,F hor ],

[0147] Similarly, the error covariance prediction matrix is ​​therefore given by the following equation:

[0148]

[0149] In 5G mode, the following state predictions are made:

[0150]

[0151] in

[0152]

[0153] Similarly, the error covariance prediction matrix It has the same form as in the equation, except that F k By F' k Replace and Q k Depend on Replacement, among which

[0154]

[0155] S p It is the velocity noise power spectrum.

[0156]

[0157] The EK-SLAM estimator corrects for state errors using the following EKF correction equation:

[0158]

[0159] in It is a measurement result estimated based on an observation model. Jacobi H can also be used k and prior state prediction To estimate.

[0160] The covariance matrix R of the measurement noise k The dimension depends on the number of GNSS satellites M being processed and / or the number of 5G base stations receiving their signals. L = N + M is the total number of available transmitters. R k It is a 2L x 2L dimensional matrix, constrained by the following formula:

[0161]

[0162] in:

[0163] DLL loop that follows GNSS or 5G signals, FLL loop band following GNSS or 5G signals, T n : coherent integration time of the GNSS or 5G DLL loop of interest, C / N0: estimate of the carrier-to-noise ratio of the GNSS or 5G channel of interest, and d: chip spacing between the early and late correlators of the GNSS or 5G DLL of interest.

[0164] In the mapping mode, In this case, the corresponding Jacobian matrix is ​​obtained in the following way:

[0165] in, and Correction produces posterior estimates and the corresponding error covariance matrix P k The posterior estimate is obtained by estimating the posterior error. The estimate of the true state at time k-1 is added to update the true state (position, velocity, clock) at time k (i.e., ...). ).

[0166] In 5G mode, only 5G pseudo-distance is available, where z = z s And the Jacobian form is given by the following equation:

[0167]

[0168] in and Correction produces posterior estimates and the corresponding error covariance matrix P' k The posterior estimate is obtained by estimating the posterior error. The estimate of the true state at time k-1 is added to update the true state (position, velocity, clock) at time k (i.e., ...). ).

[0169] In 5G mode, let N be the number of 5G base stations receiving signals:

[0170] X represents 3+3+2*N states, and Z represents N position difference observations and N velocity difference observations.

[0171]

[0172]

[0173] in

[0174] Matrix H k It has the following forms:

[0175]

[0176] In GNSS mode (standard mode), if M is the number of satellites receiving its signal, X represents 3+3+2 states independent of the number of states, and Z represents M position difference observations and M velocity difference observations.

[0177]

[0178] (x r ) k+1 =F r (x r ) k +(w r ) k

[0179]

[0180]

[0181] Matrix H k It has the following forms:

[0182]

[0183] In the mapping mode, X represents 3+3+2+(3+2)*N states, and X is not a cascade of individual 5G and GNSS states. Z represents M position difference observations followed by M velocity difference observations + N position difference observations and N velocity difference observations.

[0184]

[0185]

[0186] Matrix H k It has the following forms:

[0187]

[0188] in

[0189] and

[0190]

[0191]

[0192] Figure 12 The method for selecting the navigator's operating mode, implemented by the processing module, is illustrated schematically.

[0193] Step E120 corresponds to an operating mode in which only GNSS signals are measured, and the navigator's operating mode is the standard GNSS mode. This mode, for example, corresponds to the situation where the vehicle is on open ocean.

[0194] In step E121, the processing module checks whether the signal transmitted by the 5G base station is being measured.

[0195] If yes, the processing module proceeds to step E122. If no, the processing module returns to step E120.

[0196] Step E122 corresponds to the operating mode in which GNSS and 5G signals are measured, and the navigator's operating mode is mapping mode. This mode corresponds, for example, to the situation where the vehicle is approaching a port.

[0197] In step E123, the processing module checks whether the GNSS signal cannot be used.

[0198] If yes, the processing module proceeds to step E124; otherwise, the processing module returns to step E122.

[0199] Step E124 corresponds to the 5G operating mode. This mode, for example, corresponds to the situation where the vehicle is docked.

[0200] In step E125, the processing module checks whether the 5G signal is unusable.

[0201] If yes, the processing module returns to step E120. If no, the processing module returns to step E124.

[0202] Figure 13 A method for implementing a navigator using a loosely coupled system between an INS system and a tightly coupled GNSS / 5G system is illustrated schematically.

[0203] In step E130, an initial estimate of the state vector is obtained. Initial estimates of the error covariance.

[0204] In the next step E131, the navigator's operating mode and measurement result Z are obtained. k Z' k or Z” k .

[0205] In step E132, the Jacobian matrix and the process noise covariance matrix are obtained.

[0206] If the operation mode is mapping mode, then matrix F is obtained. k H k R k and Q k .

[0207] If the operating mode is 5G mode, then obtain matrix F' k H' k 、R' k and Q' k .

[0208] If the operating mode is standard GNSS mode, then matrix F” is obtained. k H”k 、R” k and Q” k .

[0209] In steps E133 to E135, the obtained matrix is ​​used for correction.

[0210] More precisely, the Kalman gain is calculated at step E133, the correction to the prior estimate of the new measurement result is calculated at step E134, and the estimated error covariance matrix is ​​calculated.

[0211] Therefore, when the pattern is a mapping pattern, the following matrix is ​​calculated:

[0212]

[0213] When the mode is 5G, the following matrix is ​​calculated:

[0214]

[0215] When the mode is the standard GNSS mode, the following matrix is ​​calculated:

[0216]

[0217] At the end of the calibration, predictions are performed in steps E136 to E138. In step E136, the prediction result of the state vector is calculated. In step E137, the measurement result vector is predicted, and in step E138, the prediction result of the error covariance matrix is ​​calculated.

[0218] Therefore, when the pattern is a mapping pattern, the following prediction results are calculated:

[0219]

[0220] When the mode is 5G, the following prediction results are calculated:

[0221]

[0222] When the mode is a standard GNSS mode, the following prediction results are calculated:

[0223]

Claims

1. A navigation device comprising an inertial navigation system coupled to a satellite navigation system, wherein information supplied by the satellite positioning system is used to adjust the inertial navigation system, characterized in that, The navigation device further includes means for measuring signals from base stations of a wireless cellular network, and tightly coupling the satellite navigation system and the means for measuring signals from base stations of the wireless cellular network to each other to form a satellite navigator and / or a base station navigator, thereby realizing an estimator with simultaneous positioning and mapping, and loosely coupling the satellite navigation system and the means for measuring signals from base stations of the wireless cellular network to the inertial navigation system to adjust the inertial navigation system; The satellite navigator and / or base station navigator operate in three operating modes: the first mode is characterized by the availability of pseudo-range, pseudo-velocity, and carrier-to-noise ratio (CNR) of signals from the base station of the wireless cellular network and signals from the satellite; the second mode is characterized by the availability of pseudo-range, pseudo-velocity, and CNR of signals from the base station of the wireless cellular network; and the third mode is characterized by the availability of pseudo-range, pseudo-velocity, and CNR of signals from the satellite.

2. The navigation device according to claim 1, characterized in that, The navigation device also includes a matrix generation module for forming multiple matrices, including a transition matrix, a Jacobian matrix, a measurement noise covariance matrix, and a process noise covariance matrix.

3. The navigation device according to claim 2, characterized in that, For each operating mode, the matrix generation module forms a transition matrix, Jacobian matrix, measurement noise covariance matrix, and process noise covariance matrix that are different from those formed for other modes.

4. The navigation device according to any one of claims 1 to 3, characterized in that, The navigation device includes a Kalman filter.

5. A navigation method comprising an inertial navigation system coupled to a satellite navigation system, wherein information supplied by a satellite positioning system is used to adjust the inertial navigation system, characterized in that, The navigation method includes the following steps: Measure signals from base stations in wireless cellular networks; By tightly coupling the satellite navigation system and the means for measuring signals from base stations of the wireless cellular network to form a satellite navigator and / or a base station navigator, an estimator with simultaneous positioning and mapping is realized. The satellite navigation system and the means for measuring signals from the base station of the wireless cellular network are loosely coupled to the inertial navigation system to adjust the inertial navigation system; The navigation method includes the following steps: determining the operating mode of the satellite navigator and / or the base station navigator in three operating modes, wherein the first mode is a mode in which pseudo-range, pseudo-velocity, and carrier-to-noise ratio of signals from the base station of the wireless cellular network and pseudo-range, pseudo-velocity, and carrier-to-noise ratio of signals from the satellite are available; the second mode is a mode in which pseudo-range, pseudo-velocity, and carrier-to-noise ratio of signals from the base station of the wireless cellular network are available; and the third mode is a mode in which pseudo-range, pseudo-velocity, and carrier-to-noise ratio of signals from the satellite are available.

6. The navigation method according to claim 5, characterized in that, The navigation method further includes a matrix generation step of forming multiple matrices, including a transition matrix, a Jacobian matrix, a measurement noise covariance matrix, and a process noise covariance matrix.

Images