A method for adaptive attitude estimation of vehicle-mounted integrated navigation equipment

Abstract

The invention discloses a method for adaptively estimating the attitude of an on-vehicle integrated navigation device, which is applied to an on-vehicle integrated navigation system and includes an inertial sensor IMU and a GNSS module. The method comprises the following steps: S1, collecting and correcting sensor data; S2, calculating the pitch angle and roll angle of the attitude angle of the navigation device; S3, setting and initializing the navigation system state; S4, initializing the integrated navigation algorithm; S5, performing inertial navigation calculation on the current navigation system state; S6, identifying the vehicle state; S7, selecting an error observation matrix of a Kalman filter based on the vehicle state; S8, estimating the navigation system state error by using the Kalman filter, and performing error correction on the current navigation system state; and S9, evaluating the estimation result of the installation attitude of the navigation device.

Description

Technical Field

[0001] The invention belongs to the field of multi-sensor information fusion, and in particular relates to a method for adaptively estimating the attitude of a vehicle-mounted integrated navigation device. Background Art

[0002] Before the vehicle-mounted integrated navigation device (vehicle-mounted integrated navigation system) can work properly, it needs a good initial state alignment. Due to the uncertainty of the installation method of the vehicle-mounted integrated navigation device on the vehicle body, the relative relationship between the navigation device carrier coordinate system and the vehicle body coordinate system requires an adaptive method to estimate.

[0003] Currently, the navigation method used by major commercial vehicles both domestically and internationally is a combination of GNSS and digital maps. One drawback of this method is that the vehicle navigates on a flat surface and does not output attitude angles. While this is sufficient for positioning purposes, it cannot provide a reference for the vehicle's attitude when monitoring its external environment.

[0004] A considerable amount of research has been conducted both domestically and internationally on integrated navigation, and it has been put into practical use in vehicle-mounted integrated navigation systems. Inertial navigation systems (INS) are typically initialized using GNSS position information. INS can also utilize GNSS position and velocity to perform navigation attitude alignment while driving, or utilize GNSS dual-antenna direction finding technology to align the navigation device attitude group. However, these aforementioned attitude estimation methods suffer from shortcomings such as low accuracy and complex processes. Alternatively, some technical solutions utilize the position difference between inertial sensors and global satellite navigation system outputs, employing Kalman filtering to estimate the navigation device's attitude angle relative to the vehicle, or employing geomagnetic sensors and inertial sensors to estimate the navigation device's attitude angle relative to the vehicle. However, these solutions, when used in complex environments, suffer from slow convergence, suboptimal estimation accuracy, and limited environmental adaptability. Summary of the Invention

[0005] Aiming at the problem that a vehicle-mounted integrated navigation device cannot be initialized due to an uncertain installation mode, the present invention provides a vehicle-mounted integrated navigation device attitude adaptive estimation method.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A method for adaptively estimating the attitude of a vehicle-mounted integrated navigation device is applied to a vehicle-mounted integrated navigation system, including an inertial sensor IMU and a GNSS module. The method comprises:

[0008] S1, sensor data acquisition and correction;

[0009] S2, calculates the pitch angle and roll angle of the navigation device attitude angle;

[0010] S3, set the navigation system status and initialize;

[0011] S4, integrated navigation algorithm initialization;

[0012] S5, performing inertial navigation calculation on the current navigation system status;

[0013] S6, vehicle status recognition;

[0014] S7, selects the error observation matrix of the Kalman filter based on the vehicle state;

[0015] S8, estimating the navigation system state error by the Kalman filter and performing error correction on the current navigation system state;

[0016] S9, evaluating the estimated result of the installation attitude of the navigation equipment.

[0017] Preferably, the S1 comprises the following steps:

[0018] S1.1. Collect inertial sensor information from the IMU, vehicle speed information from the CAN bus, and GNSS output information from the GNSS module at a certain sampling frequency, and mark the sampling timestamps. The IMU includes a three-axis accelerometer and a three-axis angular velocity meter, and the inertial sensor information includes three-axis acceleration data and three-axis angular velocity data.

[0019] S1.2, uses a second-order Butterworth low-pass filter to filter the three-axis acceleration data, and uses a sliding window to filter the three-axis angular velocity data.

[0020] Preferably, said S2 comprises the following steps:

[0021] On a flat road and with the vehicle stationary at t0, calculate the pitch angle θ of the navigation device attitude angle p (t0) and roll angle θ r (t0); the pitch angle calculation formula is:

[0022] The roll angle calculation formula is:

[0023] Among them, x, y, and z are the acceleration values ​​of the three-axis acceleration x-axis, y-axis, and z-axis respectively.

[0024] Preferably, the step S3 includes the following steps:

[0025] S3.1, set the attitude angle of the navigation device relative to the vehicle body is the pitch angle of the navigation device relative to the vehicle body, is the roll angle of the navigation device relative to the vehicle body, The heading angle of the navigation device relative to the vehicle body, the initial value is 0;

[0026] S3.2, set the attitude angle of the navigation system is the pitch angle of the navigation system, and its initial value is the pitch angle θ obtained in step S2 p (t0), is the roll angle of the navigation system, and its initial value is the pitch angle θ obtained in step S2 r (t0), is the heading angle of the navigation system, Among them, θ gnss For the vehicle speed to be greater than V h When , the vehicle heading angle is obtained based on the GNSS module;

[0027] S3.3, the global position and three-dimensional velocity information obtained by the GNSS module initialize the velocity and position of the navigation system respectively.

[0028] Preferably, the S4 comprises the following steps:

[0029] Set the navigation transformation matrix from the navigation coordinate system to the carrier coordinate system, and initialize the navigation transformation matrix based on the initial value of the navigation system attitude angle;

[0030] Set the vehicle body transformation matrix from the vehicle body coordinate system to the carrier coordinate system, and initialize the vehicle body transformation matrix based on the initial value of the navigation device's relative attitude angle to the vehicle body;

[0031] Obtain the corresponding four-element initial values ​​based on the navigation transformation matrix;

[0032] Set the state vector of the Kalman filter, including attitude error, velocity error, position error, angular velocity meter dynamic zero drift, and accelerometer dynamic zero drift;

[0033] Set the state transition matrix of the Kalman filter.

[0034] Preferably, based on the navigation transformation matrix, the vehicle body transformation matrix and the initial values ​​of the four elements, the strapdown inertial navigation update equation is used to perform inertial navigation solution to update the navigation system state.

[0035] Preferably, the vehicle status identification includes:

[0036] Vehicle speed is less than v stop , the acceleration value variance is less than the threshold and the duration is greater than t stop When , it is in stop state;

[0037] The absolute value of the angular velocity value of the angular velocity meter Z axis projected onto the vehicle coordinate system is less than wb straight And the vehicle speed is greater than v straightWhen the vehicle is in a high-speed straight-ahead state;

[0038] The absolute value of the angular velocity value of the angular velocity meter Z axis projected onto the vehicle coordinate system is greater than wb turn When , the vehicle is in a turning state;

[0039] Otherwise, it is other status;

[0040] Among them, v stop is the preset speed threshold when the vehicle stops, t stop is the preset vehicle stop duration threshold, wb straight is the preset angular velocity threshold of the vehicle when it is moving straight, v straight is the preset speed threshold of the vehicle when going straight, wb turn It is the preset angular velocity threshold of the vehicle when turning.

[0041] Preferably, the S7 includes:

[0042] When the vehicle is stationary, the observation matrix H of the Kalman filter is set to the observed three-dimensional velocity error, the noise matrix R is set to a constant, and the observation vector Z is the three-dimensional velocity of the vehicle in the navigation coordinate system;

[0043] When the vehicle is turning or traveling straight at high speed, the Kalman filter's observation matrix H is set to observe the three-dimensional velocity and position errors. The noise matrix R is adaptively adjusted according to a preset adaptive calculation formula based on the angular velocity meter and satellite signal-to-noise ratio. The observation vector Z is the three-dimensional velocity error and position error of the vehicle body in the navigation coordinate system.

[0044] Preferably, the adaptive calculation formula is R=N1×P 900 / CN0-0.7 +N2×Q abs(wbz)*10 , where N1, N2, Q and P are constants, CNO is the GNSS signal-to-noise ratio, and wbz is the vehicle steering angular velocity.

[0045] Preferably, step S9 includes the following steps:

[0046] S9.1, basic inch Get the current heading angle of the navigation device relative to the vehicle;

[0047] S9.2: When it is detected that the current heading angle of the navigation device relative to the vehicle body converges to a variance less than a certain threshold, the estimation is considered complete, otherwise return to step 5.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] The present invention utilizes low-cost inertial sensors and a global navigation satellite system, supplemented by vehicle body speed information, to estimate the attitude angle of the navigation device relative to the vehicle body. The invention involves inertial navigation, Kalman filtering, and vehicle model constraints, among other aspects. It is highly adaptable to the installation method of the navigation and positioning device and has better accuracy. It also solves the low efficiency and cumbersome process of manual calibration in the past. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0051] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0052] This invention is developed and designed based on the Rockchip RV1126 platform and the NXP S32K144 32-bit microcontroller. Considering the relatively good real-time performance of the microcontroller, sensor data acquisition and analysis are performed on the S32K144 throughout the system. The FreeRTOS real-time operating system is used for task scheduling, ensuring good time alignment between sensors. The algorithm involves extensive matrix calculations, so the EIGEN library was ported during implementation. The algorithm runs on the RV1126 running the Linux operating system.

[0053] Reference Figure 1 As shown, a method for adaptively estimating the attitude of an on-board integrated navigation device is applied to the on-board integrated navigation system, including an inertial sensor IMU and a GNSS module. The inertial sensor IMU is an inertial measurement unit, which is a sensor mainly used to detect and measure acceleration and rotational motion. The inertial sensor IMU includes a three-axis accelerometer and a three-axis angular velocity meter; the method for adaptively estimating the attitude of an on-board integrated navigation device includes 9 steps.

[0054] Step 1: Sensor data collection and correction, specifically:

[0055] Step 1.1: Collect inertial sensor information from the IMU, vehicle speed information from the CAN bus, and GNSS output information from the GNSS module at sampling frequencies of 100 Hz, 10 Hz, and 2 Hz, respectively, and set the system timestamp; the inertial sensor information includes three-axis acceleration data and three-axis angular velocity data;

[0056] In step 1.2, a second-order Butterworth low-pass filter is used to filter the three-axis acceleration data, and a sliding window is used to filter the three-axis angular velocity data.

[0057] In step 1.2 of the present invention, considering that the vibration of the vehicle engine has a greater impact on the triaxial accelerometer, a second-order Butterworth low-pass filter is used to filter the accelerometer measurement value. The low-pass filter is designed using the Filter Designer in the Matlab toolbox, and the sampling frequency F of the low-pass filter is set. s is the IMU sampling frequency F s =100HZ, set the cutoff frequency of the low-pass filter to F c Set it to be less than the vibration frequency of the vehicle body. Here the vibration frequency of the vehicle body is 30HZ, then F c = 20 Hz, and then automatically generates a parameter table for the filter difference equation. The angular velocity output is smoothed using sliding window filtering. The sliding window size is selected based on the sampling frequency; in this embodiment, the window size is set to 50. Finally, the accelerometer and angular velocity biases are subtracted from each other. The accelerometer bias is obtained using the six-plane method, while the angular velocity bias is calculated from the mean value after rest. The "mean value after rest" here refers to the average value of the angular velocity output over a period of time when the IMU is stationary.

[0058] Step 2: Calculate the pitch and roll angles of the navigation device attitude angles, specifically:

[0059] On a straight road and in a stationary state t0, the pitch angle θ of the navigation device attitude angle is calculated by equations (1) and (2) respectively: p (t0) and roll angle θ r (t0);

[0060]

[0061]

[0062] Among them, x, y, and z are the acceleration values ​​of the three-axis acceleration x-axis, y-axis, and z-axis respectively.

[0063] The attitude angle refers to the angle between the body coordinate system and the ground inertial coordinate system, which can be expressed by three angles: roll angle, pitch angle, and yaw angle.

[0064] Step 3: Set the navigation system status and initialize;

[0065] S3.1, set the attitude angle of the navigation device relative to the vehicle body is the pitch angle of the navigation device relative to the vehicle body, is the roll angle of the navigation device relative to the vehicle body, The heading angle of the navigation device relative to the vehicle body, the initial value is 0;

[0066] Step 3.2: The pitch angle θ of the navigation device attitude angle obtained in step 2 is p (t0) and roll angle θ r (t0) is the attitude angle θ of the navigation system nb Pitch angle and roll angle The initial value of the vehicle speed is calculated by the CAN bus information, and when the vehicle speed is greater than the preset speed threshold V h When the GNSS module is used to obtain the vehicle system x-axis (here the front of the vehicle is set as the vehicle system x-axis, the right side of the vehicle is set as the vehicle system y-axis, and the bottom of the vehicle is set as the vehicle system z-axis) in the navigation coordinate system, that is, the vehicle body heading angle θ gnss , clockwise is positive; the heading angle of the navigation system attitude angle is calculated by formula (4)

[0067]

[0068] Then the attitude angle θ of the carrier system relative to the navigation coordinate system is nb , that is, the attitude angle information of the navigation system can be expressed as:

[0069]

[0070] In step 3.3, the global position and three-dimensional velocity information obtained by the GNSS module are used to initialize the velocity and position of the navigation system respectively.

[0071] Step 4: Initialize the integrated navigation algorithm, including:

[0072] Set the navigation transformation matrix from the navigation coordinate system to the carrier coordinate system And the initial value of the navigation system attitude angle [θ p (t0) θ r (t0) θ gnss ]Initialize the navigation transformation matrix;

[0073] Set the vehicle body transformation matrix from the vehicle body coordinate system to the carrier coordinate system The attitude angle θ of the navigation device relative to the vehicle body obtained in step 3 vb The initial value [θ p (t0) θ r (t0) 0] Initialize the vehicle body transformation matrix

[0074] Based on the navigation transformation matrix Get the corresponding four-element initial value q0;

[0075] Set the state vector X of the Kalman filter p, expressed as:

[0076] X p =[δθ n δV n δP n δb w δb a ] T (6)

[0077] where δθ n is the attitude error, δV n is the speed error, δP n is the position error, δb w is the dynamic zero drift of the angular velocity meter, δb a Accelerometer dynamic zero drift;

[0078] Set the state transfer matrix F of the Kalman filter. The state transfer matrix F is determined by the attitude error equation, velocity error equation, position error equation of the inertial navigation algorithm and the IMU dynamic zero bias model.

[0079] The attitude error equation is expressed as:

[0080]

[0081] Where, is the angle meter measurement error, Calculate the error for the navigation system, & represents the differential.

[0082] The velocity error equation is expressed as:

[0083]

[0084] Where, is the accelerometer measurement error, δg n They are the calculation error of the earth's rotation angular velocity, the calculation error of the navigation system rotation and the gravity error.

[0085] The position error equation is expressed as:

[0086]

[0087]

[0088] δh & =δv U (11)

[0089] Among them, δL, δλ, and δh represent the latitude error, longitude error, and height error respectively, and R M is the meridian curvature radius, R Nis the lateral curvature radius, L is the latitude, and h is the altitude.

[0090] The state transfer matrix of the IMU dynamic bias model is expressed as:

[0091]

[0092] Among them, F aa 、F gg represents the dynamic bias state transfer matrix of the accelerometer and angular velocity meter, α and β are constants, and I is the three-dimensional unit matrix. In this embodiment, α=150 and β=1.

[0093] Here, the attitude error equation, velocity error equation, position error equation and IMU dynamic zero bias model are common knowledge in the field, and those skilled in the art can set them according to actual conditions.

[0094] Step 5: Perform inertial navigation calculation on the current navigation system status, specifically:

[0095] The attitude update equation, velocity update equation and position update equation are used to perform inertial navigation to update the attitude, velocity and position of the system. Among them, the attitude update equation can obtain the updated heading of the carrier in the navigation coordinate system. The GNSS module can obtain the vehicle's heading θ under the navigation system gntss , then the updated heading angle of the navigation device relative to the vehicle attitude angle is The calculation formula is:

[0096]

[0097] Here, the attitude update equation, velocity update equation and position update equation are common knowledge in the art and will not be described in detail here.

[0098] Step 6: Vehicle state identification. The vehicle state during operation is divided into stop state, high-speed straight state, turning state and other states based on the vehicle body speed and IMU measurement value. The determination status of each state is as follows:

[0099] Vehicle speed is less than v stop , the acceleration value variance is less than the threshold and the duration is greater than t stop When , it is in stop state; v stop is the preset speed threshold when the vehicle stops, t stop is the preset duration threshold when the vehicle stops. In this embodiment, t stop =0.8s;

[0100] The absolute value of the angular velocity value of the angular velocity meter Z axis projected onto the vehicle coordinate system is less than wb straight And the vehicle speed is greater than vstraight When the vehicle is in a high-speed straight-ahead state, wb straight is the preset angular velocity threshold of the vehicle when it is moving straight, v straight is the preset speed threshold of the vehicle when traveling straight ahead. In this embodiment,

[0101] The absolute value of the angular velocity value of the angular velocity meter Z axis projected onto the vehicle coordinate system is greater than wb turn When the vehicle is in a turning state, wb turn is the preset angular velocity threshold value when the vehicle turns; in this embodiment,

[0102] Otherwise, it is other status.

[0103] It should be noted here that Kalman observation fusion is not performed in other states, and only states with obvious characteristics, such as stationary state, turning state or high-speed straight state, are selected for fusion.

[0104] Step 7: Select the error observation matrix of the Kalman filter based on the vehicle state.

[0105] In the present invention, different observation matrices H are first selected based on different vehicle states, and the observation equation V = ZH*X p , thereby obtaining different Kalman observation information (new information).

[0106] The error observation matrix of the Kalman filter selected based on the vehicle state is:

[0107] (1) When the vehicle is stationary, the observation matrix H of the Kalman filter is set to the three-dimensional velocity error observation, that is, the three-dimensional velocity element corresponding to the observation matrix H is set to 1, and the noise matrix R is 0.0625. The observation vector Z obtained by the GNSS module is the three-dimensional velocity of the vehicle in the navigation coordinate system. When the vehicle is stationary, the three-dimensional velocity observed by the GNSS module is the measurement error.

[0108] (2) When the vehicle is in a turning state or in a high-speed straight state, the observation matrix H of the Kalman filter is set to three-dimensional velocity error and position error observation, the three-dimensional velocity observation matrix is ​​set to 1, and the position observation matrix Tpr is calculated by the following formula:

[0109] Tpr=diag([Rm+h, (Rn+h)cosL, -1]) (14)

[0110] Where Rm is the principal radius of curvature of the meridian circle, Rn is the principal radius of curvature of the meridian circle, and h and L are the altitude and latitude output by the GNSS module, respectively.

[0111] The noise matrix R is adaptively adjusted by the angular velocity meter and the satellite signal-to-noise ratio according to the preset adaptive calculation formula:

[0112] R=N1×P 900 / CN0-0.7 +N2×Q abs(wbz)*10 (15) Where N1, N2, P, and Q are constants, Q and P are the amplification coefficients of the contribution of the signal-to-noise ratio and the vehicle steering angular velocity value to the observation noise at this time, respectively, CN0 is the GNSS signal-to-noise ratio, and wbz is the vehicle steering angular velocity value. In this embodiment, N1 = 0.2, N2 = 0.1, P = 3.0, and Q = 5.0.

[0113] The observation vector Z obtained by the GNSS module is the three-dimensional velocity error and position error of the vehicle body in the navigation coordinate system, where the three-dimensional velocity error is set to:

[0114] v error =v nav -v gnss (16)

[0115] v nav is the three-dimensional velocity calculated by inertial navigation, v gnss is the three-dimensional velocity measured by GNSS.

[0116] The three-dimensional position error is set as:

[0117] p error =p nav -p gnss (17)

[0118] p nav is the three-dimensional position calculated by inertial navigation, p gnss is the three-dimensional position measured by GNSS.

[0119] Step 8: The Kalman filter estimates the navigation system state error and performs error correction on the current navigation system state.

[0120] In step 8 of the present invention, a Kalman filter calculation is performed based on the state transfer matrix of the Kalman filter in step 4 and the error observation matrix of the Kalman filter in step 7 to obtain the optimal estimate of the inertial navigation system error. The navigation system attitude, velocity and position updated by the inertial navigation algorithm in step 5 are subtracted from the error output by the Kalman filter to obtain the corrected navigation system attitude, velocity and position.

[0121] Step 9: Evaluate the estimated result of the navigation equipment installation attitude, specifically:

[0122] S9.1, basic inch Get the current heading angle of the navigation device relative to the vehicle

[0123] S9.2, when the current heading angle of the navigation device relative to the vehicle is detected When the variance converges to a value less than a certain threshold, the estimation is considered complete, otherwise return to step 5.

[0124] In step 9.1 of the present invention, the posture corrected in step 8 is obtained Based on formula (13), the heading angle of the navigation device relative to the vehicle body at the current moment is obtained; in step 9.2 of the present invention, the process from step 5 to step 9 is continuously iterated, and under the constraints of multiple information observations, the attitude angle θ of the navigation device relative to the vehicle body is obtained. vb Heading angle Converge quickly to the true value and finally pass the variance test when The variance of is less than 0.001 threshold and lasts for 50 calculation cycles. Convergence complete.

[0125] The heading angle at which convergence is completed The heading angle of the vehicle-mounted integrated navigation device attitude angle obtained by adaptive estimation, the pitch angle θ of the navigation device attitude angle calculated in step 2 p (t0) and roll angle θ r (t0) is the pitch angle and roll angle of the vehicle-mounted integrated navigation device attitude angle obtained by adaptive estimation.

Claims

1. A method for adaptively estimating the attitude of a vehicle-mounted integrated navigation device, characterized in that: Applied to a vehicle-mounted integrated navigation system, including an inertial sensor IMU and a GNSS module, the method includes: S1, sensor data acquisition and correction; S2, calculate the pitch angle and roll angle of the navigation device attitude angle; on a flat road and in a stationary state t0, calculate the pitch angle θ of the navigation device attitude angle p (t0) and roll angle θ r (t0); the pitch angle calculation formula is: , the roll angle calculation formula is: , Among them, x, y, and z are the acceleration values ​​of the three-axis acceleration x-axis, y-axis, and z-axis respectively; S3, setting the navigation system state and initializing it, includes the following steps: S3.1, set the attitude angle of the navigation device relative to the vehicle body , is the pitch angle of the navigation device relative to the vehicle body, , is the roll angle of the navigation device relative to the vehicle body, , The heading angle of the navigation device relative to the vehicle body, the initial value is 0; S3.2, set the attitude angle of the navigation system , is the pitch angle of the navigation system, and its initial value is the pitch angle θ obtained in step S2 p (t0), is the roll angle of the navigation system, and its initial value is the roll angle θ obtained in step S2 r (t0), is the heading angle of the navigation system, , where θ gnss When the vehicle speed is greater than the speed threshold V h When , the vehicle heading angle is obtained based on the GNSS module; S3.3, the global position and three-dimensional velocity information obtained by the GNSS module are used to initialize the velocity and position of the navigation system respectively; S4, integrated navigation algorithm initialization; S5, performing inertial navigation calculation on the current navigation system status; S6, vehicle status recognition; S7, selects the error observation matrix of the Kalman filter based on the vehicle state; S8, estimating the navigation system state error by the Kalman filter and performing error correction on the current navigation system state; S9, evaluating the estimated result of the navigation equipment installation attitude, including the following steps: S9.1, based on Get the current heading angle of the navigation device relative to the vehicle; S9.2, when it is detected that the current heading angle of the navigation device relative to the vehicle body converges to a variance less than a certain threshold, it is considered that the estimation is completed, otherwise return to step S5.

2. The method for adaptively estimating the attitude of a vehicle-mounted integrated navigation device according to claim 1, wherein: Said S1 comprises the following steps: S1.

1. Collect inertial sensor information from the IMU, vehicle speed information from the CAN bus, and GNSS output information from the GNSS module at a certain sampling frequency, and mark the sampling timestamps. The IMU includes a three-axis accelerometer and a three-axis angular velocity meter, and the inertial sensor information includes three-axis acceleration data and three-axis angular velocity data. S1.2, uses a second-order Butterworth low-pass filter to filter the three-axis acceleration data, and uses a sliding window to filter the three-axis angular velocity data.

3. The method for adaptively estimating the attitude of a vehicle-mounted integrated navigation device according to claim 1, wherein: The S4 comprises the following steps: Set the navigation transformation matrix from the navigation coordinate system to the carrier coordinate system, and initialize the navigation transformation matrix based on the initial value of the navigation system attitude angle; Set the vehicle body transformation matrix from the vehicle body coordinate system to the carrier coordinate system, and initialize the vehicle body transformation matrix based on the initial value of the navigation device's relative attitude angle to the vehicle body; Obtain the corresponding four-element initial values ​​based on the navigation transformation matrix; Set the state vector of the Kalman filter, including attitude error, velocity error, position error, angular velocity meter dynamic zero drift, and accelerometer dynamic zero drift; Set the state transition matrix of the Kalman filter.

4. The method for adaptively estimating the attitude of a vehicle-mounted integrated navigation device according to claim 1, wherein: Based on the navigation transformation matrix, vehicle transformation matrix and the initial values ​​of the four elements, the strapdown inertial navigation update equation is used to perform inertial navigation solution and update the navigation system status.

5. The method for adaptively estimating the attitude of a vehicle-mounted integrated navigation device according to claim 1, wherein: The vehicle state identification includes: Vehicle speed is less than v stop , the acceleration value variance is less than the threshold and the duration is greater than t stop When , it is in stop state; The absolute value of the angular velocity value of the angular velocity meter Z axis projected onto the vehicle coordinate system is less than wb straight And the vehicle speed is greater than v straight When the vehicle is in a high-speed straight-ahead state; The absolute value of the angular velocity value of the angular velocity meter Z axis projected onto the vehicle coordinate system is greater than wb turn When , the vehicle is in a turning state; Otherwise, it is other status; Among them, v stop is the preset speed threshold when the vehicle stops, t stop is the preset vehicle stop duration threshold, wb straight is the preset angular velocity threshold of the vehicle when it is moving straight, v straight is the preset speed threshold of the vehicle when going straight, wb turn It is the preset angular velocity threshold of the vehicle when turning.

6. The method for adaptively estimating the attitude of a vehicle-mounted integrated navigation device according to claim 5, wherein: The S7 includes: When the vehicle is stationary, the observation matrix H of the Kalman filter is set to the observed three-dimensional velocity error, the noise matrix R is set to a constant, and the observation vector Z is the three-dimensional velocity of the vehicle in the navigation coordinate system; When the vehicle is turning or traveling straight at high speed, the Kalman filter's observation matrix H is set to observe the three-dimensional velocity and position errors. The noise matrix R is adaptively adjusted according to a preset adaptive calculation formula based on the angular velocity meter and satellite signal-to-noise ratio. The observation vector Z is the three-dimensional velocity error and position error of the vehicle body in the navigation coordinate system.

7. The method for adaptively estimating the attitude of a vehicle-mounted integrated navigation device according to claim 6, wherein: The adaptive calculation formula is R=N1×P 900 / CN0-0.7 +N2×Q abs(wbz)*10 , where N1, N2, Q and P are constants, CN0 is the GNSS signal-to-noise ratio, and wbz is the vehicle steering angular velocity.

Images