Integrated navigation system combining GNSS / SINS with vehicle CAN bus speed information

Abstract

The invention relates to the technical field of navigation, in particular to a GNSS / SINS (Global Navigation Satellite System / Strapdown Inertial Navigation System) combined vehicle CAN (Controller Area Network) bus speed information combined positioning system, which is based on an MEMS-IMU (Micro Electro Mechanical System-Inertial Measurement Unit) and a GNSS receiving module, acquires required inertial navigation resolving information and GNSS positioning information in real time, acquires vehicle speed information of a vehicle through an OBD (On-Board Diagnostic) port of the vehicle, and finally processes the parameters by a PC (Personal Computer) to obtain a fusion positioning result. The principle of the combined navigation algorithm is based on a fusion positioning algorithm of extended Kalman filtering, and a non-integrity constraint algorithm and a zero-velocity correction technology are introduced. According to the invention, the robustness of the integrated navigation system is enhanced; and the requirements of high precision and high stability of vehicle-mounted positioning can be better met.

Description

Technical Field

[0001] The present invention relates to the technical field of navigation, and particularly relates to a combined positioning system that combines GNSS / SINS with vehicle CAN bus speed information. Background Art

[0002] Combined navigation and positioning systems are widely used in multiple fields and are the basis for the development of in-vehicle auxiliary industries such as autonomous driving. The satellite navigation and positioning system (GNSS) can obtain real-time and all-weather positioning and orientation information through the communication between satellites and signal receivers, but the positioning accuracy is restricted by a series of environmental conditions. The strapdown inertial navigation system (SINS) is a device that can achieve the positioning and orientation of a carrier through its own calculation, and can independently obtain relevant required content without interacting with the outside world. However, without the correction of external information, especially for low-precision hardware, there will be obvious cumulative error problems. These two systems have their own advantages and disadvantages in performance, and have complementary characteristics, so they can be designed to work in combination.

[0003] In the urban vehicle driving environment, GNSS signals are extremely vulnerable to interference or occlusion, resulting in the combined navigation result being unable to meet the positioning requirements due to large errors. Therefore, other components are usually added to the combined navigation system to enhance the stability of the system. The most common is to add an odometer, and use the odometer observation information to constrain the inertial navigation unit calculation when the GNSS signal is unavailable.

[0004] The current common speed information obtained by an external odometer is calculated from the difference between the pulse count values output at two sampling intervals, the resolution of the odometer encoder itself, and information such as the wheel diameter. The odometer speed obtained by this method has a large error, and generally, it is necessary to accumulate data information for a period of time and then calculate the average speed. This will result in a delay in the calculated speed and a low output frequency, reducing the constraint effect of the odometer information in the real-time combined positioning and navigation system.

[0005] The conventional Kalman filter scheme in information fusion cannot match the non-linear environment. In the urban vehicle driving environment, the measurement noise of the system is not all approximately Gaussian white noise. In some noisy areas, the non-linearity of the noise is very strong and there are changes, which will cause large errors in the filtering results. Summary of the Invention

[0006] Object of the Invention

[0007] Aiming at the problems of low accuracy and easy susceptibility to environmental occlusion and interference in the current consumer-grade GNSS / SINS combined navigation system, the present invention aims to provide a combined positioning system that combines GNSS / SINS with vehicle CAN bus speed information, which can effectively ensure the use effect in various environments.

[0008] Technical solution

[0009] A combined positioning system that combines GNSS / SINS with vehicle CAN bus speed information, based on a MEMS-IMU and a GNSS receiving module, obtains the required inertial navigation solution information and GNSS positioning information in real time, obtains the vehicle speed information through the vehicle's OBD port, and finally processes these parameters by a PC to obtain a fused positioning result;

[0010] The combined positioning system includes: a SINS module, a GPS module, a CAN data transceiver module, and a combined navigation calculation unit;

[0011] The SINS module includes an inertial measurement unit and a pose solution module;

[0012] The inertial measurement unit includes a gyroscope and an accelerometer, where

[0013] The gyroscope measures and obtains the carrier angular velocity information. Through the initial angle determined by initial alignment, the coordinate transformation matrix from the carrier coordinate system to the navigation coordinate system at subsequent moments can be deduced;

[0014] The accelerometer measures and obtains the acceleration information, and obtains the acceleration components mapped on the three axes of the navigation coordinate system through the coordinate transformation matrix.

[0015] The pose solution module includes a strapdown inertial navigation algorithm module. Through the angular velocity information and acceleration information obtained in the inertial measurement unit, the attitude angle, velocity, and position information of the carrier are calculated. The SINS module can only measure the changes in the attitude angle, velocity, and position of the carrier relative to the initial position. Through the initial alignment of the strapdown inertial navigation, knowing the initial longitude, latitude, and elevation information of the carrier and the angles between the three axes of the carrier coordinate system and the navigation coordinate system, the navigation and positioning information of the carrier during subsequent movements can be obtained in real time;

[0016] The GPS module includes a GNSS receiver, which integrates an RF radio frequency chip, a baseband chip, and a processing chip, and externally connects an antenna to receive and decode the signals sent by the satellite to calculate the three-dimensional coordinates of the observation point; for a moving carrier, its speed information can be deduced through the changes in the three-dimensional coordinate information at some consecutive moments; this module is a prior art;

[0017] The CAN data transceiver module includes a CAN transceiver chip. The CAN data transceiver module is connected to the integrated navigation calculation unit and the vehicle OBD port (an external module, not part of the present invention), and is connected to the vehicle CAN bus (an external module, not part of the present invention) through the vehicle OBD port; the CAN transceiver chip realizes the communication between the integrated navigation calculation unit and the vehicle CAN bus, and the integrated navigation calculation unit sends instructions to the CAN bus via the CAN transceiver chip and receives data packets containing vehicle speed information.

[0018] The integrated navigation calculation unit includes an MCU single-chip microcomputer, which runs the integrated navigation algorithm and completes the calculation content of integrated navigation.

[0019] The principle of the integrated navigation algorithm is a fusion positioning algorithm based on the extended Kalman filter:

[0020] When the number of satellites received by the satellite signal receiver and the GNSS accuracy coefficient are within the set range, the satellite signal is regarded as valid; when the satellite signal is valid, the difference between the position and speed output by the receiver and the position and speed output by the INS forms the observed measurements of the carrier position and speed in the fusion filtering algorithm; when the satellite signal is invalid, the difference between the speed in the CAN and the output speed of the INS forms the observed measurement of the carrier speed and calculates the filtering result; the adaptive extended Kalman filter algorithm is used to adaptively adjust the environmental noise covariance matrix R during state update and the environmental noise covariance matrix Q during measurement update, so that these two important parameters of the extended Kalman filter update can be autonomously adjusted according to different environmental conditions to reduce errors.

[0021] Furthermore, generally when a vehicle is moving, it will not jump from the ground or skid on the ground, and the speed component in the plane perpendicular to the vehicle's forward direction is equal to zero. A non-holonomic constraint algorithm is designed to limit the speed update of the inertial navigation; in actual driving, there are relatively frequent parking situations. When the detected GNSS speed or the inertial navigation calculated speed is less than this threshold, it can be considered that the vehicle has not moved.

[0022] Furthermore, a zero-speed detection algorithm is designed. Only the difference between the speed information calculated by the inertial navigation and the speed information obtained by the GNSS is used as the observed quantity, and the position and attitude angle information are averaged over the previous time period to reduce the error accumulation of the inertial navigation.

[0023] The system working process includes the following steps:

[0024] Step 1: Based on the development board integrating the MEMS-IMU and the GNSS receiving module, obtain the required inertial navigation calculation information and GNSS positioning information in real time.

[0025] Step 2: The development board is connected to the CAN bus through the vehicle OBD port for communication, sending instructions to it and receiving the returned data frames. The development board is connected to an external PC wirelessly or wiredly to transmit and display the positioning results in real time;

[0026] Step 3: The integrated navigation calculation unit of the development board evaluates the effectiveness of the acquired satellite signals based on the GNSS signal status information such as the number of received satellites and dilution of precision obtained by the GNSS signal receiving module. When the satellite signals are valid, an adaptive extended Kalman filter fusion algorithm is adopted for the GNSS positioning results and the inertial navigation solution results; while when the satellite signals are determined to be invalid, a dead reckoning algorithm is used to fuse the inertial navigation solution results and the OBD speed information to obtain the positioning results;

[0027] Step 4: According to the actual motion of the vehicle, non-holonomic constraints and vehicle kinematic constraints are used to constrain the parameters in the fusion filtering algorithm to improve the positioning accuracy.

[0028] Technical Effects

[0029] The advantages of the present invention are as follows:

[0030] (1) The present invention proposes a fusion positioning algorithm based on extended Kalman filter. When the satellite signals are received well, a positioning system is composed of a satellite signal receiver and an inertial navigation element; while when the satellite signals are distorted or cannot be received, the vehicle speed and inertial navigation obtained by other sensors are combined; the two systems are independent of each other and do not interfere with each other, and are applied to different scenarios, enhancing the robustness of the integrated navigation system while maintaining the flexibility of GNSS / SINS loose coupling;

[0031] (2) By obtaining the vehicle speed information in the vehicle CAN bus through the vehicle OBD port and using it as auxiliary information when GNSS signals are missing, it effectively reduces the output delay and increases the output frequency compared with a conventional externally connected odometer, realizes real-time communication and online filtering, and meets the requirements of high-precision and high-stability vehicle positioning;

[0032] (3) In the fusion positioning algorithm, a simplified adaptive filtering algorithm is proposed to meet the situation where environmental noise changes; in addition, vehicle kinematic constraints are added to the algorithm, and a non-holonomic constraint algorithm and a zero-speed detection algorithm are proposed to correct the pose information of the vehicle to make it more conform to the actual vehicle motion model and improve the overall positioning accuracy. Brief Description of the Drawings

[0033] Figure 1 It is a functional module diagram of a combined positioning system of GNSS / SINS / vehicle CAN bus speed of the present invention;

[0034] Figure 2It is the working flowchart of a combined positioning system for GNSS / SINS / vehicle CAN bus speed according to the present invention;

[0035] Figure 3 It is the positioning trajectory (solid line) and reference trajectory (dashed line) of a general GPS / SINS navigation system in the actual measurement of an embodiment of the present invention;

[0036] Figure 4 It is the positioning trajectory (solid line) and reference trajectory (dashed line) of a general GNSS / SINS navigation system when the GPS signal is lost in some sections during the actual measurement of an embodiment of the present invention;

[0037] Figure 5 It is the positioning trajectory (solid line) and reference trajectory (dashed line) of the system of the present invention when the GPS signal is lost in some sections during the actual measurement of an embodiment of the present invention. Specific embodiments

[0038] The present invention will be described in detail below with reference to the accompanying drawings.

[0039] A combined positioning system of the present invention that combines GNSS / SINS with vehicle CAN bus speed information, as Figure 1 shown, includes a SINS module 1, a GPS module 2, a CAN data transceiver module 3, and a combined navigation calculation unit 4;

[0040] In the SINS module 1, the inertial measurement elements gyroscope and accelerometer respectively measure the angular rate and specific force of the carrier. Through the common inertial navigation quaternion solution method, the real-time updated Euler angles and coordinate transformation matrix can be obtained; the specific force obtained by the accelerometer is integrated to obtain speed update, and after double integration, position update is obtained. The calculation formulas are respectively:

[0041]

[0042] Among them, are the three-axis accelerations of the carrier in the carrier coordinate system, f b is the acceleration calculated by the accelerometer, is the transformation matrix from the navigation coordinate system to the carrier coordinate system;

[0043]

[0044] Among them, λ, L, and h are respectively the latitude, longitude, and altitude of the carrier, is the three-axis speed of the carrier in the navigation coordinate system, R N is the radius of curvature of the prime vertical, R M is the radius of curvature of the meridian;

[0045] In the GPS module 2, the GNSS antenna transmits the received satellite signals to the RF front end through the data channel, and after processing by the RF front end, interference suppression module, baseband processing unit, etc., the decoded GNSS positioning information required is transmitted outward;

[0046] The CAN data transceiver module 3 can automatically encode and decode CAN data, and only needs to query parameters such as vehicle speed according to the CAN protocol to receive feedback;

[0047] The integrated navigation calculation unit 4 includes an MCU single-chip microcomputer that runs the integrated navigation algorithm; the integrated navigation algorithm includes: an adaptive extended Kalman filtering algorithm when the GNSS signal is valid and a dead reckoning algorithm when the GNSS signal is invalid;

[0048] Adaptive extended Kalman filtering algorithm:

[0049] The adaptive extended Kalman filtering algorithm is divided into two parts: state update and measurement update; both GNSS and inertial navigation systems can obtain the carrier speed and position information. The inertial navigation solution frequency is high while the GNSS signal output frequency is low; taking the inertial navigation solution part as the basic system and using the satellite signals obtained by the satellite receiver as a reference to correct the parameters of the inertial navigation to obtain the fusion result. The system state equation and measurement equation are respectively expressed as:

[0050] X k =F k|k-1 X k-1 +Γ k|k-1 W k-1

[0051]

[0052] Among them, X k is the system state variable at time k, the system observation vector is Z k , Γ k|k-1 and W k-1 are respectively the system noise coefficient and system noise matrix, H k is the system measurement matrix, V k is the measurement noise matrix, W k-1 and V k are respectively the environmental noise during system state update and observation, and F k|k-1 is the system state transition matrix determined by this algorithm, v INS 、v GNSS 、p INS 、p GNSS are respectively the three-axis velocity vectors and three-axis position vectors obtained by INS and GNSS solutions in the navigation coordinate system. And the system state transition matrix F is expressed as:

[0053]

[0054] Among them, M aa 、M av 、M ap are the coefficient matrices of the angle term, velocity term, and position term in the attitude error equation during inertial navigation solution. M va 、M vv 、M vp are the coefficient matrices of the angle term, velocity term, and position term in the velocity error equation. M pv 、M pp are the coefficient matrices of the velocity term and position term in the position error equation. The system measurement matrix H, system noise coefficient Γ, system noise matrix W, and measurement noise matrix V can be expressed as follows:

[0055]

[0056] Among them are the random walk error vectors of the gyroscope and accelerometer respectively, and v v 、v p are the noise vectors of the velocity measurement and position measurement at the GNSS receiver end respectively.

[0057] Using the extended Kalman filter model, some problems with not particularly obvious non-linear situations are transformed into relatively easy-to-handle linear problems; the extended Kalman filter is expanded using Taylor's formula, and the first-order term is taken. The filtering formula of the EKF under non-linear system states can be expressed as:

[0058]

[0059] Among them, P k is the covariance matrix of the state variables, Q k and R k are the environmental noise covariances during system state update and observation respectively, K k is the Kalman filter gain, and other variables are the same as above; f(X k ) = [f 1 (X k ) f 2 (X k )…f n (X k )] T h(X k ) = [h 1 (X k ) h 2 (X k )…h m (X k )] T They are the nth and mth Taylor expansions of the k-dimensional state variables respectively. The expansion of Taylor's formula can be written as:

[0060]

[0061] where f (n) (x) represents the nth derivative of f(x), and R n (x) is the remainder term of the Taylor formula.

[0062] The state variables of the system are 15-dimensional, which are the errors of the carrier's velocity, position, and attitude, as well as the biases of the three axes of the accelerometer and gyroscope:

[0063]

[0064] where are the attitude angle errors mapped to the three coordinate axes calculated by the inertial measurement unit, and δV E , δV N , δV U are the velocity errors of the three axes, δL, δλ, and δh are the position errors of the three axes, and the latter six dimensions are the random walk errors of the gyroscope and accelerometer respectively.

[0065] The adaptive Kalman filtering algorithm is adopted to adaptively adjust the two parameters that most significantly affect the fusion filtering effect: the environmental noise covariance matrix R during state update and the environmental noise covariance matrix Q during measurement update; the weighted method is used to increase the weighting coefficient of the recent data items and weaken the role of the old measurement data in the estimation, so that the recent measurement data plays a major role in the estimation. In this way, in the case of relatively complex noise conditions, the initial parameters are determined by tuning, and the subsequent system parameters mainly depend on the values at the adjacent moments before. The expression can be written as:

[0066]

[0067] b k =(1 - b) / (1 - b k )

[0068] where, R k is the measurement noise matrix, is the initially determined measurement matrix noise value, and b is the iteration base quantity, and it is more appropriate to take values between 0.9 and 0.99.

[0069] Nonholonomic constraint conditions:

[0070] When the vehicle moves conventionally, it will not jump off the ground or skid on the ground, that is, the velocity components in the plane perpendicular to the vehicle's forward direction are all equal to zero, and this direction is also the y-axis in the carrier coordinate system. This is the nonholonomic constraint condition of the vehicle; when the angular velocity of the vehicle during turning is greater than 0.05 rad / s, side slip may occur, and at this time, the nonholonomic constraint condition needs to be disabled. The expression can be written as:

[0071]

[0072] Among them, are the velocity components of the vehicle body in the y and z axes in the vehicle body coordinate system, and ω yaw is the angular velocity of the vehicle on the horizontal plane.

[0073] Using the two-axis velocities of the vehicle body coordinate system calculated by the inertial navigation system as the interference velocities, feedback and compensation are performed on the inertial navigation parameters to improve the observability of the system, and the measurement matrix of the system is rewritten as:

[0074]

[0075] Among them, Z PI -Z PG is the position difference between the GNSS and INS calculation results, and Z VI -Z VG is the velocity difference between the GNSS and INS calculation results. These two disturbance velocities are the solutions of the inertial navigation on two axes perpendicular to the forward direction; dividing the difference between the disturbance velocities of adjacent time intervals by the time interval can obtain the dynamic bias of the accelerometer. Modifying the parameters of the state variables in the integrated positioning system in real time can be expressed as:

[0076]

[0077] Among them are the dynamic biases in the z and y axis directions of the accelerometer respectively, are the disturbance velocity values in the z and y axis directions at adjacent moments respectively, and t is the time interval.

[0078] Zero velocity correction technology:

[0079] Adopting the zero velocity correction technology, when it is determined that the vehicle body is in a stationary state, the velocity in the observation variables is set to zero, and the predicted value output by the INS is corrected; during actual driving, the vehicle often stops. Using the zero velocity correction technology can theoretically effectively suppress the error accumulation caused by the long-term operation of the INS system and improve the positioning accuracy of the entire integrated navigation system; by setting a threshold, when the detected GNSS velocity or the inertial navigation calculation velocity is less than this threshold, it is considered that the vehicle is not moving, and only the difference between the velocity information calculated by the inertial navigation and the velocity information calculated by the GNSS is used as the observation quantity, and the position and attitude information are averaged over the previous time period to reduce the error accumulation of the inertial navigation. It is more appropriate to set the threshold to 0.03 - 0.07 m / s, and the expression can be written as:

[0080]

[0081] Among them, Euler_angle represents the attitude angle of the vehicle, and position represents the position of the vehicle. The threshold is set to 0.05 m / s, and the time interval t 1 is set to 20 s.

[0082] Dead reckoning algorithm:

[0083] The dead reckoning algorithm combines CAN speed information and IMU, and calculates the position and motion state at the next moment according to the current position and motion state of the vehicle. The CAN speed information can be expressed as:

[0084]

[0085] where v D is the speed indication returned by the CAN bus, and it can be converted to the navigation coordinate system through the coordinate system transformation matrix:

[0086]

[0087] where is the CAN speed information, which is used as the observable of the combined system. The coordinate system transformation matrix is obtained from the inertial navigation solution.

[0088] The position information output by the odometer can be obtained through the CAN speed information:

[0089]

[0090] where L O , λ O , h O represent the longitude, latitude and elevation information obtained through the CAN speed, is the speed of the CAN return speed indication mapped to the three axes of north, east and up in the navigation coordinate system.

[0091] The fusion filtering state variables adopted by the system can be written as:

[0092]

[0093] Among them, the first 15 dimensions of the state variables are the three-axis attitude angle error, three-axis speed error, three-axis position error, random walk of the three axes of the gyroscope, and random walk of the three axes of the accelerometer solved by the IMU. The last three-dimensional variable is the scale error of the three axes of the odometer.

[0094] The state equation of the system is expressed as:

[0095]

[0096] Among them, the state transition matrix F is expressed as:

[0097]

[0098] Among them, M aa 、M av 、M ap are the coefficient matrices of the angle term, velocity term, and position term in the attitude error equation during inertial navigation solution. M va 、M vv 、M vp are the coefficient matrices of the angle term, velocity term, and position term in the velocity error equation. M pv 、M pp are the coefficient matrices of the velocity term and position term in the position error equation.

[0099] The system noise coefficient G and the system noise matrix W are expressed as:

[0100]

[0101] Among them are respectively the random walk error vectors of the gyroscope and the accelerometer.

[0102] The system measurement equation is expressed as:

[0103]

[0104] Among them, the measurement noise matrix V D is the noise of the odometer measurement environment, and the measurement matrix H D can be expressed as:

[0105] H D =[0 3×6 I 3×3 -I 3×3 0 3×3 T

[0106] The flow of a method for realizing combined positioning of a GNSS / SINS combined with vehicle CAN bus speed information according to the present invention is as Figure 2 shown, and specifically includes the following steps:

[0107] Step 1: Power on and initialize the module port configuration;

[0108] ​Initialize and calibrate the MCU single-chip microcomputer, SINS module, GPS module, and CAN data transceiver module, and set the interrupt controller and UART serial port; the priority of the MCU single-chip microcomputer's data transceiver serial port needs to be set to enable each sensor to receive data orderly and effectively, without data conflicts and omission of key data; the priority of the MCU single-chip microcomputer to receive GNSS data through the serial port should be set relatively high, which enables each incoming GNSS signal to be fused and filtered with the SINS solution data following it immediately.

[0109] Step 2: SINS initial alignment and solution, and obtain GNSS positioning information;

[0110] After the code is burned in and the product is powered on, the MCU single-chip microcomputer will automatically configure the ports; in the normal working mode, read the inertial element data output by the SINS solution and receive the navigation positioning information sent by the satellite at a frequency of 200 Hz; complete the initial alignment of the SINS by obtaining continuous GNSS positioning information;

[0111] Step 3: Fusion filtering algorithm;

[0112] When the GNSS information is valid, fuse the inertial element data output by the IMU solution with the GNSS positioning information; when the GNSS information is invalid, fuse the speed information output by the on-vehicle diagnostic system (OBD) interface with the IMU solution information to obtain the attitude, position, and speed information of the current vehicle, and transmit the obtained fusion result outward through the serial communication protocol.

[0113] The combined positioning system designed in the present invention, which combines GNSS / SINS with vehicle CAN bus speed information, can improve the positioning accuracy to a certain extent compared with a single GNSS system. It still has good positioning performance within a certain period of time in the case of GNSS signal loss, with low cost and low latency, and can communicate with the outside world conveniently; adding an adaptive extended Kalman filtering algorithm and using a vehicle kinematic model for constraint can improve the overall positioning accuracy of the system to a certain extent.

[0114] Table 1 shows the comparison of the improvement amplitudes of different algorithms in improving the system accuracy in a test result. The experimental vehicle is a Volkswagen Tiguan SUV. The test system includes a GNSS antenna placed in front of the roof, a high-precision Wuhan Kuibu positioning (decimeter position accuracy) receiver used as a reference, and a development board fixed on the vehicle and connected to the vehicle OBD port. The experimental environment is the Jiading Campus of Tongji University.

[0115] Table 1 Comparison of the improvement amplitudes of different algorithms in improving the system accuracy in a test result

[0116]

[0117] As can be seen from Table 1, for the test data this time, the zero-speed detection algorithm has the most obvious improvement in the system positioning accuracy. The adaptive Kalman filtering algorithm and the non-integrity constraint have a relatively small improvement, while adding the three algorithms simultaneously can greatly improve the positioning performance of the system.

[0118] Figures 3 - 5 It is the experimental satellite map of the normal situation in this test, the ordinary GPS / SINS integrated navigation system when the GPS signal is missing, and the designed system of the present invention when the GPS signal is missing. It can be seen that in the case of the lack of GPS signal, the positioning result of the ordinary GPS / SINS will diverge rapidly, while the designed system of the present invention can achieve non-divergence within a certain time (<30 s, 40 km / h), and the curve is smoother.

[0119] It should be noted that: the combination navigation method disclosed in the embodiments of the present invention only describes the preferred embodiments in this aspect of the art, and is only used to illustrate the technical solutions of the present invention rather than to limit them; the selection of the terms used herein and the explanation of the principles of each embodiment can be understood by other ordinary technicians in this technical field.

Claims

1. A combined positioning system of GNSS / SINS combined with vehicle CAN bus speed information, characterized in that: Based on MEMS-IMU and GNSS receiving modules, the required inertial navigation solution information and GNSS positioning information are obtained in real time, and the vehicle speed information is obtained through the OBD port of the vehicle. Finally, these parameters are processed by the PC to obtain the fusion positioning result; The combined positioning system comprises: a SINS module, a GPS module, a CAN data transceiver module and a combined navigation calculation unit; The combined navigation calculation unit includes an MCU single chip computer, which runs a combined navigation algorithm to complete the calculation content of the combined navigation; The principle of the combined navigation algorithm is based on the fusion positioning algorithm of the extended Kalman filter: When the number of satellites received by the satellite signal receiver and the GNSS precision coefficient are within the set range, the satellite signal is considered valid; when the satellite signal is valid, the difference between the position and velocity output by the receiver and the position and velocity output by the INS forms the observation measurement of the carrier position and velocity in the fusion filtering algorithm; when the satellite signal is invalid, the difference between the speed in the CAN and the output speed of the INS forms the observation measurement of the carrier speed and calculates the filtering result; an adaptive extended Kalman filter algorithm is used to adaptively update the environmental noise covariance matrix R during state update and the environmental noise covariance matrix Q during measurement update, so that these two important parameters updated by the extended Kalman filter can be adjusted autonomously according to different environmental conditions to reduce errors.

2. A combined positioning system of GNSS / SINS combined with vehicle CAN bus speed information as claimed in claim 1, characterized in that: The SINS module includes an inertial measurement unit and a posture solution module; The inertial measurement unit includes a gyroscope and an accelerometer, wherein The gyroscope measures and obtains the carrier angular velocity information, and through the initial angle determined by the initial alignment, the coordinate conversion matrix from the carrier coordinate system to the navigation coordinate system at subsequent moments can be calculated; The accelerometer measures and obtains acceleration information, and obtains acceleration components mapped on three axes of the navigation coordinate system through a coordinate conversion matrix; The posture calculation module includes a strapdown inertial navigation algorithm module, which calculates the attitude angle, velocity and position information of the carrier through the angular velocity information and acceleration information obtained in the inertial measurement unit; the SINS module can only measure the change of the attitude angle, velocity and position of the carrier relative to the initial position, and through the initial alignment of the strapdown inertial navigation, the initial latitude and longitude elevation information of the carrier and the angle between the three axes of the carrier coordinate system and the navigation coordinate system are known, so that the navigation positioning information of the carrier in subsequent movement can be obtained in real time; The GPS module includes a GNSS receiver, which integrates an RF chip, a baseband chip and a processing chip, and an external antenna to receive and decode the signal sent by the satellite to calculate the three-dimensional coordinates of the observation point; for a moving carrier, its speed information can be inferred through the changes in the three-dimensional coordinate information at some consecutive moments; This module is prior art; The CAN data transceiver module includes a CAN transceiver chip, which connects the combined navigation calculation unit and the vehicle OBD port (an external module, not part of the present invention), and is connected to the vehicle CAN bus (an external module, not part of the present invention) through the vehicle OBD port; the CAN transceiver chip realizes communication between the combined navigation calculation unit and the vehicle CAN bus, and the combined navigation calculation unit sends instructions to the CAN bus via the CAN transceiver chip and receives data messages containing vehicle speed information.

3. A combined positioning system of GNSS / SINS combined with vehicle CAN bus speed information as claimed in claim 1, characterized in that: The combined navigation algorithm also includes a non-integrity constraint algorithm; Generally, when a vehicle is moving, it will not jump off the ground or slip on the ground. The velocity components in the plane perpendicular to the vehicle's forward direction are all equal to zero. A non-holonomic constraint algorithm is designed to limit the velocity update of the inertial navigation system. In actual driving, there are relatively frequent parking situations. When the detected GNSS speed or inertial navigation solution speed is less than this threshold, it can be considered that the vehicle has not moved.

4. A combined positioning system of GNSS / SINS combined with vehicle CAN bus speed information as claimed in claim 3, characterized in that: The combined navigation algorithm also includes designing a zero-speed detection algorithm: only the difference between the speed information calculated by inertial navigation and the speed information obtained by GNSS is used as the observation quantity, and the position and attitude angle information is taken as the average processed value in the previous period to reduce the error accumulation of inertial navigation.

5. A combined positioning system of GNSS / SINS combined with vehicle CAN bus speed information as claimed in claim 1, characterized in that: The system workflow includes the following steps: Step 1: Based on the development board with integrated MEMS-IMU and GNSS receiver module, obtain the required inertial navigation solution information and GNSS positioning information in real time; Step 2: The development board communicates with the CAN bus through the vehicle's OBD port, sends commands to it and receives returned data frames; the development board connects to an external PC wirelessly or wired to transmit and display positioning results in real time; Step 3: The integrated navigation calculation unit of the development board evaluates the validity of the acquired satellite signal based on the GNSS signal status information such as the number of received satellites and precision factor obtained by the GNSS signal receiving module; when the satellite signal is valid, the GNSS positioning result and the inertial navigation solution result are used to perform an adaptive extended Kalman filter fusion algorithm; when the satellite signal is judged to be invalid, the dead reckoning algorithm is used to fuse the inertial navigation solution result and the OBD speed information to obtain the positioning result; Step 4: According to the actual movement of the vehicle, the non-holonomic constraints and vehicle kinematic constraints are used to constrain the parameters in the fusion filtering algorithm to improve the positioning accuracy.

6. A combined positioning system of GNSS / SINS combined with vehicle CAN bus speed information as claimed in claim 4, characterized in that: The adaptive extended Kalman filter algorithm is specifically as follows: The adaptive extended Kalman filter algorithm is divided into two parts: state update and measurement update. Both GNSS and inertial navigation systems can obtain carrier velocity and position information. The inertial navigation solution frequency is high while the GNSS signal output frequency is low. The inertial navigation solution part is used as the basic system, and the satellite signal obtained by the satellite receiver is used as a reference correction to correct the inertial navigation parameters to obtain the fusion result. The system state equation and measurement equation are expressed as follows: X k =F k|k-1 X k-1 +Γ k|k-1 W k-1 Among them, X k is the system state variable at time k, and the system observation vector is Z k , Γ k|k-1 and W k-1 are the system noise coefficient and system noise matrix, H k is the system measurement matrix, V k is the measurement noise matrix, W k-1 and V k are the system state update and the environmental noise during the observation period, and F k|k-1 is the system state transfer matrix determined by this algorithm, v INS 、v GNSS 、p INS 、p GNSS are the three-axis velocity vector and three-axis position vector solved by INS and GNSS in the navigation coordinate system respectively; and the system state transfer matrix F is expressed as: Among them, M aa 、M av 、M ap It is the coefficient matrix of angle term, velocity term and position term in the attitude error equation during inertial navigation solution. va 、M vv 、M vp is the coefficient matrix of angle term, velocity term and position term in the velocity error equation, M pv 、M pp is the coefficient matrix of the velocity term and the position term in the position error equation; the system measurement matrix H, the system noise coefficient Γ, the system noise matrix W and the measurement noise matrix V can be expressed as: in are the random walk error vectors of the gyroscope and accelerometer, respectively, v v 、v p They are the noise vector of velocity measurement and the noise vector of position measurement at the GNSS receiver, respectively; The extended Kalman filter model is used to transform some nonlinear problems that are not particularly obvious into linear problems that are easier to handle; the extended Kalman filter is expanded using the Taylor formula, and the first-order term is taken. The filter formula of the EKF in the nonlinear system state can be expressed as: Among them, P k is the covariance matrix of the state variables, Q k and R k are the environmental noise covariance during system state update and observation, K k is the Kalman filter gain, and other variables are the same as above; f(X k )=[f1(X k )f2(X k )…f n (X k )] T h(X k )=[h1(X k )h2(X k )…h m (X k )] T The nth and mth Taylor expansions of the k-dimensional state variable are respectively; the expansion of Taylor's formula can be written as: where f (n) (x) represents the nth-order derivative of f(x), R n (x) is the remainder of Taylor's formula; The state variables of the system are 15 dimensions, which are the errors of the carrier's velocity, position and attitude, and the deviations of the three axes of the accelerometer and gyroscope: in are the attitude angle errors mapped to the three coordinate axes calculated by the inertial measurement unit, δV E , δV N , δV U is the velocity error of the three axes, δL, δλ, δh are the position errors of the three axes, and the last six dimensions are the random walk errors of the gyroscope and accelerometer respectively; Adaptive Kalman filter algorithm is used to adapt the two parameters that most obviously affect the fusion filter effect: the environmental noise covariance matrix R during state update and the environmental noise covariance matrix Q during measurement update; the weighted method is used to increase the weight coefficient of the recent data item, weaken the role of the old measurement data in the estimation, and make the recent measurement data play a major role in the estimation. In this way, in the case of complex noise conditions, the initial parameters are determined by the adjustment parameters, and the subsequent system parameters mainly depend on the values ​​of the previous adjacent moments. The expression can be written as: b k =(1-b) / (1-b k ) Among them, R k is the measurement noise matrix, is the initially determined measurement matrix noise value, b is the iteration basis, and 0.9-0.99 is more appropriate.

7. A combined positioning system of GNSS / SINS combined with vehicle CAN bus speed information as claimed in claim 4, characterized in that: The non-integrity constraints are specifically: When the vehicle moves normally, it will not jump off the ground or slip on the ground, that is, the velocity components in the plane perpendicular to the vehicle's forward direction are all equal to zero. This direction is also the y-axis in the carrier coordinate system. This is the vehicle's non-complete constraint condition. When the angular velocity of the vehicle when turning is greater than 0.05rad / s, side slip may occur. At this time, the non-complete constraint condition needs to be disabled. The expression can be written as: in, is the velocity component of the carrier on the y and z axes in the carrier coordinate system, ω yaw is the angular velocity of the vehicle on the horizontal plane; The two-axis velocities of the carrier coordinate system calculated by the inertial navigation system are used as the interference velocity to feedback and compensate the inertial navigation parameters to improve the observability of the system. The measurement matrix of the system is rewritten as: Where Z PI -Z PG is the position difference between the GNSS and INS solutions, and Z VI -Z VG It is the velocity difference between the GNSS and INS solution results. These two perturbation velocities are the solutions of the inertial navigation on two axes perpendicular to the forward direction. The dynamic bias of the accelerometer can be obtained by dividing the perturbation velocity difference of adjacent time intervals by the time interval. The parameters of the state variables in the combined positioning system are modified in real time, which can be expressed as: in are the dynamic bias of the accelerometer in the z and y axis directions, are the disturbance velocity values ​​in the z and y axis directions at adjacent moments respectively, and t is the time interval.

8. A combined positioning system of GNSS / SINS combined with vehicle CAN bus speed information as claimed in claim 4, characterized in that: The zero speed correction technology is specifically as follows: The zero-speed correction technology is used to determine that the carrier is in a stationary state, and the speed in the observed variable is set to zero to correct the predicted value output by the INS. In actual driving, the vehicle often stops. The use of zero-speed correction technology can theoretically effectively suppress the error accumulation caused by long-term operation of the INS system and improve the positioning accuracy of the entire integrated navigation system. By setting a threshold, when the detected GNSS speed or inertial navigation solution speed is less than this threshold, it is considered that the vehicle has not moved, and only the difference between the speed information calculated by the inertial navigation and the speed information calculated by the GNSS is used as the observation quantity. The position and attitude information are taken as the average processing value in the previous period to reduce the error accumulation of inertial navigation. It is more appropriate to set the threshold to 0.03~0.07m / s. The expression can be written as: Wherein, Euler_angle represents the attitude angle of the carrier, position represents the position of the carrier, the threshold is set to 0.05m / s, and the time interval t1 is set to 20s.

9. A combined positioning system of GNSS / SINS combined with vehicle CAN bus speed information as claimed in claim 1, characterized in that: The dead reckoning algorithm is specifically: The dead reckoning algorithm combines CAN speed information with IMU to calculate the next position and motion state based on the current position and motion state of the carrier. The CAN speed information can be expressed as: where v D It is the speed indication returned by the CAN bus, which can be converted to the navigation coordinate system through the coordinate system conversion matrix: in is the CAN speed information, as the observation quantity of the combined system, the coordinate system transformation matrix Derived from inertial navigation solution; The position information output by the odometer can be obtained through the CAN speed information: Where L O , O 、h O Indicates the latitude, longitude and elevation information obtained through CAN speed. The speed indication returned by CAN in the navigation coordinate system is mapped to the speed of the three axes of the northeast and celestial sky; The fusion filter state variable used by the system can be written as: The first 15 dimensions of the state variables are the three-axis attitude angle error, three-axis velocity error, three-axis position error calculated by the IMU, the random walk of the three axes of the gyroscope, and the random walk of the three axes of the accelerometer. The last three-dimensional variable is the scale error of the three axes of the odometer. The state equation of the system is expressed as: The state transfer matrix F is expressed as: Among them, M aa 、M av 、M ap It is the coefficient matrix of angle term, velocity term and position term in the attitude error equation during inertial navigation solution. va 、M vv 、M vp is the coefficient matrix of angle term, velocity term and position term in the velocity error equation, M pv 、M pp It is the coefficient matrix of velocity term and position term in the position error equation; The system noise factor G and system noise matrix W are expressed as: in are the random walk error vectors of the gyroscope and accelerometer, respectively; The system measurement equation is expressed as: The measurement noise matrix V D The noise of the odometer measurement environment, the measurement matrix H D It can be expressed as: H D =[0 3×6 I 3×3 -I 3×3 0 3×3 ] T 。

Images