A vehicle positioning method based on an inertial measurement unit mounted on a wheel

Abstract

The present invention discloses a vehicle positioning method based on an inertial measurement unit (IMU) mounted on a vehicle wheel. The IMU is mounted on the vehicle wheel. Based on the known IMU mounting parameters, a state equation and an observation equation are constructed through the kinematic constraints of the vehicle and the wheel. Using a Kalman filter, the wheel rotation angle, the wheel rotation angular velocity, and the vehicle body attitude are estimated as state variables, and then the vehicle pose and speed are obtained. At the same time, a method for estimating the IMU mounting parameters of the wheel is provided. First, in the wheel-IMU alignment stage, the preliminary mounting attitude of the IMU relative to the wheel is obtained. Then, the mounting parameters are included in the state variables of the Kalman filter for estimation to obtain the wheel IMU mounting parameters, the vehicle pose, and the speed. The present invention improves the estimation accuracy of the mounting parameters of the IMU relative to the wheel and further improves the vehicle positioning accuracy based on the IMU mounted on the wheel.

Description

Technical Field

[0001] The present invention belongs to the technical field of inertial navigation, and particularly relates to a vehicle positioning method based on a wheel-mounted inertial measurement unit. Background Art

[0002] For the calculation of vehicle pose, conventional odometers often need to be installed before the vehicle leaves the factory, and conventional rotationally modulated strapdown inertial navigation systems require complex and precise electromechanical control systems to effectively suppress the systematic errors of inertial devices. However, an inertial measurement unit (Initial Measurement Unit, IMU) can be equivalent to the combination of an odometer and a single-axis rotationally modulated strapdown inertial navigation system only by being installed on the wheel, greatly improving the inertial navigation positioning accuracy and reducing costs.

[0003] Micro-Electro-Mechanical System (MEMS) IMUs are widely used in vehicle positioning due to their advantages such as low cost, small size, and low power consumption. The main problem faced by MEMS inertial navigation is that the positioning error diverges severely over time. To address this problem, in reference [1] ([1] Collin J. MEMS IMU Carouseling for Ground Vehicles. 《IEEE Transactions on Vehicular Technology》. 2015, Vol. 64, No. 6), Collin used a MEMS IMU installed on the wheel for vehicle positioning. By sensing the gravitational acceleration with a three-axis accelerometer to obtain the number of wheel rotations, the "odometer" measurement effect was achieved. The zero drift of the gyro was suppressed by the wheel rotation modulation effect, and the vehicle attitude was sensed by the gyro, ultimately achieving the positioning purpose. In a positioning experiment of nearly 1 km, the maximum error did not exceed 8 m. The positioning model in reference [1] only holds under extremely ideal conditions. In reference [1], a special fixture was used to install the MEMS IMU at the wheel axis center, but the center of the IMU housing is not the same as the sensitive center of the accelerometer. It is difficult to ensure that the sensitive center of the accelerometer is exactly on the wheel rotation axis during installation, and there is an eccentricity. As a result, centripetal acceleration and tangential acceleration will be generated due to wheel rotation and act on the accelerometer, affecting the accurate estimation of wheel speed. The axis alignment error of the IMU relative to the wheel is also inevitable during installation, and it will be equivalent to the systematic error of the IMU, affecting the vehicle azimuth estimation accuracy. Especially when the vehicle is moving at a high speed, the installation error will make the error of the vehicle positioning model based on the wheel-mounted IMU prominent, further reducing the vehicle positioning accuracy. Using a special fixture can reduce the installation error to a certain extent, but it increases the installation difficulty and usage cost.

[0004] In [2] (Method for on-site calibration of installation parameters of a wheel-mounted inertial measurement unit, CN 114152269 A, March 8, 2022), a method for calibrating the installation parameters of a wheel IMU and a vehicle positioning method are proposed. However, the following problems mainly exist: (1) The accuracy of the established kinematic model can still be improved. For example, the Coriolis acceleration term is ignored, and the pitch angular velocity of the vehicle body is directly defaulted to 0, etc.; (2) The method focuses on estimating the installation parameters of the inertial measurement unit. Therefore, in the Kalman filter adopted, the state variables only include the wheel rotation parameters and the installation parameters of the inertial measurement unit, without the vehicle body attitude data, and the zero horizontal angle of the vehicle body is not observed, resulting in a large vehicle body attitude error and thus a large positioning error. (3) The method is mainly a method for calibrating the installation parameters of the IMU relative to the wheel, and no specific scheme for vehicle positioning after obtaining stable calibration parameters is given. Therefore, there is still room for improvement in the accuracy of the vehicle positioning model based on the wheel-mounted IMU established by it. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to solve the problem that the installation parameters of the IMU are unknown, and to solve the problem of the positioning accuracy of the vehicle positioning model based on the wheel-mounted IMU due to the IMU installation error, the movement of the vehicle body and the wheels, and a vehicle positioning scheme after the known IMU installation parameters are given.

[0006] To solve the above technical problems, the present invention adopts the following technical solutions:

[0007] A vehicle positioning method based on a wheel-mounted inertial measurement unit, based on an inertial measurement unit installed at a preset position of a vehicle wheel, when the installation parameters of the inertial measurement unit are unknown, perform steps A - B to obtain the installation parameters of the inertial measurement unit, as well as the real-time position, attitude and speed of the vehicle; when the installation parameters of the inertial measurement unit are known, perform step C to obtain the real-time position, attitude and speed of the vehicle:

[0008] Step A: Based on the output data of the inertial measurement unit, use a preset attitude solution algorithm to obtain a preliminary installation attitude matrix of the inertial measurement unit relative to the wheel;

[0009] Step B: Based on the preliminary installation attitude matrix of the inertial measurement unit relative to the wheel, include preset various types of wheel parameters, preset various types of vehicle body attitude parameters and preset inertial measurement unit installation-related parameters in the state variables, include the output data of the inertial measurement unit and the zero horizontal angle of the vehicle body in the observed variables, and combine with a Kalman filter to obtain the installation parameters of the inertial measurement unit, as well as the real-time position, attitude and speed of the vehicle;

[0010] Step C: Based on the known installation parameters of the inertial measurement unit, include the preset wheel parameters of various types and the preset body attitude parameters of various types in the state variables, include the output data of the inertial measurement unit and the body zero-level angle in the observation variables, and combine with the Kalman filter to obtain the real-time position, attitude, and speed of the vehicle.

[0011] As a preferred technical solution of the present invention, in the step A, the following steps are specifically executed to obtain the preliminary installation attitude matrix of the inertial measurement unit relative to the wheel:

[0012] Step A1: Collect the output data of the three-axis accelerometer in the inertial measurement unit within a preset time period when the vehicle is in a stopped state, and take the average value of the output data of the three-axis accelerometer within the preset time period, denoted as Collect the output data of the three-axis gyroscope in the inertial measurement unit within a preset time period when the vehicle is in a stopped state, and take the average value of the output data of the three-axis gyroscope within the preset time period, denoted as the gyro zero bias;

[0013] Step A2: Collect the output data of the three-axis gyroscope within a preset time period when the vehicle is in a driving state, and subtract the gyro zero bias from the output data of the three-axis gyroscope, and then obtain the average value of the output data of the three-axis gyroscope within the preset time period, denoted as

[0014] Step A3: Based on the average value data obtained in Step A1 and Step A2, obtain the preliminary installation attitude matrix of the inertial measurement unit relative to the wheel through the attitude solution algorithm shown by the following formula

[0015]

[0016] Wherein,

[0017]

[0018]

[0019]

[0020] Wherein, It represents the attitude transformation matrix from the b - system to the h - system, that is, the preliminary installation attitude matrix of the inertial measurement unit relative to the wheel; the h - system is a coordinate system that coincides with the w - system after rotating by β around its z - axis according to the right - hand rule. The origins of the h - system and the w - system coincide. At the initial moment, its y - axis points vertically upward; the w - system represents the wheel coordinate system, whose origin is located at the center of the wheel, its x - axis is perpendicular to the wheel rotation axis and points from the center of the wheel to the center of the inertial measurement unit, the z - axis points to the left side of the vehicle along the wheel rotation axis, and the y - axis is determined by the right - hand coordinate system rule; both the h - system and the w - system are fixed relative to the wheel; the inertial measurement unit coordinate system is the b - system, and its three axes are respectively along the three sensitive directions of the inertial measurement unit, and its origin is located at the center of the inertial measurement unit, that is, the sensitive center of the tri - axial accelerometer; the i - system is the inertial coordinate system; It represents the rotation of the original coordinate system around the j - axis of this system according to the right - hand rule The attitude transformation matrix from the obtained coordinate system to the original coordinate system, where j represents the x, y, or z - axis; θ h represents the pitch angle of the inertial measurement unit in the h - system, γ h represents the roll angle of the inertial measurement unit in the h - system, ψ h represents the azimuth angle of the inertial measurement unit in the h - system.

[0021] As a preferred technical solution of the present invention, in step B, the following steps are specifically executed to obtain the installation parameters of the inertial measurement unit, as well as the real - time position, attitude, and speed of the vehicle;

[0022] Step B1: For the state quantity including preset parameters of various types of wheels, preset parameters of various types of vehicle body attitudes, and preset parameters related to the installation of the inertial measurement unit, that is, the state quantity Construct a state equation; for the observable quantity including the output data of the inertial measurement unit and the zero - level angle of the vehicle body, that is, the observable quantity Construct an observation equation; set the initial value;

[0023] where α is the wheel rotation angle, is the wheel angular velocity, is the wheel angular acceleration; r is the distance between the center of the inertial measurement unit and the wheel rotation axis, that is, the eccentricity; γ, θ, and ψ are respectively the roll angle, pitch angle, and azimuth angle of the vehicle; is the output data of the tri - axial accelerometer in the inertial measurement unit measured; is the output data of the z - axis in the projection of the output angular velocity of the tri - axial gyroscope in the inertial measurement unit measured in the h - system;

[0024] Step B2: Based on the constructed state equation and observation equation, use the Kalman filter to iteratively execute steps B2.1 to B2.4 until the vehicle positioning ends and the iteration ends, thereby obtaining the installation parameters of the inertial measurement unit and the real - time position, attitude, and speed of the vehicle:

[0025] Step B2.1: Obtain the sampling time t based on a preset sampling period k Output data of the inertial measurement unit, where the output data of the three-axis gyroscope of the inertial measurement unit is subtracted by the gyro zero bias;

[0026] Step B2.2: Perform Kalman filter time update, specifically: Based on the state quantity estimated value at sampling time t k-1 , obtain the predicted value of the state quantity at sampling time t through the state equation; And based on the estimated value of the state error covariance matrix at sampling time t k and the state equation, obtain the predicted value of the state error covariance matrix at sampling time t k-1 ; k

[0027] Step B2.3: Perform Kalman filter measurement update, specifically: Based on the predicted value of the state error covariance matrix at sampling time t k , the observation noise variance matrix, and the observation equation, obtain the gain matrix at sampling time t k ; Based on the gain matrix, the predicted value of the state quantity at sampling time t k , the observation data at sampling time t k and the observation equation, obtain the estimated value of the state quantity at sampling time t k ; And based on the gain matrix, the predicted value of the state error covariance matrix at sampling time t k , the observation noise variance matrix, and the observation equation, obtain the estimated value of the state error covariance matrix at sampling time t k ;

[0028] Step B2.4: Based on the estimated value of the state quantity at sampling time t k , obtain the vehicle attitude k at sampling time t Vehicle position Eccentricity r and the installation attitude matrix of the inertial measurement unit relative to the wheel where R is the wheel radius; Return to step B2.1 and enter the iteration of the next sampling time.

[0029] As a preferred technical solution of the present invention, the constructed state equation and observation equation are as follows:

[0030] State equation:

[0031] Observation equation:

[0032]

[0033] Among them, ​

[0034]

[0035]

[0036]

[0037] g v = [0 0 -g] T ;

[0038] Wherein, R is the wheel radius; g represents the magnitude of the gravitational acceleration; represents the attitude transformation matrix from the p system to the q system. Both the p system and the q system refer to coordinate systems and satisfy The superscript 'T' represents taking the transpose of the matrix; is the projection of the motion quantity s of the q system relative to the p system in the o system. The motion quantity s is the angular velocity w, the position p, the velocity v or the acceleration a. is the output angular velocity of the three-axis gyroscope in the inertial measurement unit measured actually; the vehicle body coordinate system is the v system, and its three axes point to the right of the vehicle, in front of the vehicle, and above the vehicle in sequence, and its origin is located at the contact point of the wheel where the IMU is installed and the ground; the navigation coordinate system is the n system, which is fixed relative to the earth, and the n system coincides with the v system at the initial moment; η and ε are preset process noises; ∈, ζ γ and ζ θ are both preset observation noises.

[0039] As a preferred technical solution of the present invention, in step C, based on the known installation parameters of the inertial measurement unit, the following steps are specifically executed to obtain the real-time position, attitude, and velocity of the vehicle;

[0040] Step C1: Collect the output data of the three-axis gyroscope in the inertial measurement unit within a preset time period when the vehicle is in a stopped state, and take the average value of the output data of the three-axis gyroscope within the preset time period, which is recorded as the gyro zero bias;

[0041] Step C2: For the state quantity including preset various types of wheel parameters and preset various types of vehicle body attitude parameters, that is, the state quantity construct a state equation; for the observation quantity including the output data of the inertial measurement unit and the vehicle body zero horizontal angle, that is, the observation quantity construct an observation equation; set the initial value;

[0042] Wherein, α is the wheel rotation angle, is the wheel angular velocity, is the wheel angular acceleration; γ, θ, and ψ are the vehicle roll angle, pitch angle, and azimuth angle respectively; The output data of the triaxial accelerometer in the inertial measurement unit measured in actuality; The output data of the z-axis in the projection of the output angular velocity of the triaxial gyroscope in the inertial measurement unit in the w coordinate system; the w coordinate system represents the wheel coordinate system, whose origin is located at the wheel center, its x-axis is perpendicular to the wheel rotation axis and points from the wheel center to the inertial measurement unit center, the z-axis points to the left side of the vehicle body along the wheel rotation axis direction, the y-axis is determined by the right-hand coordinate system rule, and the w coordinate system is fixed relative to the wheel; the inertial measurement unit coordinate system is the b coordinate system, and the three axes are respectively along the three sensitive directions of the inertial measurement unit, and its origin is located at the inertial measurement unit center, that is, the sensitive center of the triaxial accelerometer; the i coordinate system is the inertial coordinate system;

[0043] Step C3: Based on the constructed state equation and observation equation, use the Kalman filter to iteratively execute Step C3.1 to Step C3.4 until the vehicle positioning ends and the iteration ends, and then obtain the real-time position, attitude, and speed of the vehicle:

[0044] Step C3.1: Based on the preset sampling period, obtain the sampling time t k The output data of the inertial measurement unit, where the output data of the triaxial gyroscope of the inertial measurement unit is subtracted by the gyro zero bias;

[0045] Step C3.2: Perform the Kalman filter time update, specifically: Based on the state quantity estimated value at the sampling time t k-1 , obtain the state quantity predicted value at the sampling time t k through the state equation; and based on the state error covariance matrix estimated value at the sampling time t k-1 and the state equation, obtain the state error covariance matrix predicted value at the sampling time t k ;

[0046] Step C3.3: Perform the Kalman filter measurement update, specifically: Based on the state error covariance matrix predicted value at the sampling time t k , the observation noise variance matrix, and the observation equation, obtain the gain matrix at the sampling time t k ; Based on the gain matrix, the state quantity predicted value at the sampling time t k , the observation data at the sampling time t k and the observation equation, obtain the state quantity estimated value at the sampling time t k ; And based on the gain matrix, the state error covariance matrix predicted value at the sampling time t k , the observation noise variance matrix, and the observation equation, obtain the state error covariance matrix estimated value at the sampling time t k ;

[0047] Step C3.4: Based on the state quantity estimated value at the sampling time t k , obtain the state quantity at the sampling time t kVehicle posture Vehicle location Where R is the wheel radius; return to step C3.1 and enter the next sampling moment iteration.

[0048] As a preferred technical solution of the present invention, the constructed state equation and observation equation are as follows:

[0049] Equation of state:

[0050]

[0051] Observation equation:

[0052]

[0053] in,

[0054]

[0055]

[0056]

[0057]

[0058] In the formula, R is the wheel radius; g represents the magnitude of gravitational acceleration; Represents the attitude transformation matrix from the p system to the q system, where both the p system and the q system refer to coordinate systems; is the projection of the motion s of the q system relative to the p system on the o system. The motion s is the angular velocity w, position p, velocity v or acceleration a. is the measured angular velocity output by the three-axis gyroscope in the inertial measurement unit; the body coordinate system is the v system, whose three axes point to the right, front and top of the vehicle respectively, and its origin is located at the contact point between the wheel where the IMU is installed and the ground; the navigation coordinate system is the n system, which coincides with the v system at the initial moment relative to the solidification of the earth; η and ε are the preset process noises; ∈、ζ γ and θ All are preset observation noises.

[0059] As a preferred technical solution of the present invention, the inertial measurement unit installation parameters include the attitude transformation matrix of the inertial measurement unit coordinate system relative to the wheel coordinate system. And the distance r between the center of the inertial measurement unit and the wheel axis.

[0060] As a preferred technical solution of the present invention, the Kalman filter adopts an extended Kalman filter, an unscented Kalman filter, or a cubature Kalman filter.

[0061] The beneficial effects of the present invention are as follows: The present invention provides a vehicle positioning method based on an inertial measurement unit mounted on a wheel. The present invention fully considers the influence of IMU installation parameters, vehicle body movement, and wheel movement on the output of the wheel IMU, establishes a high-precision vehicle positioning model based on the wheel-mounted IMU, constructs a state equation and an observation equation based on this model, and performs information fusion through the Kalman filter algorithm to achieve efficient estimation of the vehicle position, speed, attitude, and wheel IMU installation parameters, improving the vehicle positioning accuracy based on the wheel-mounted IMU; and gives a vehicle positioning scheme after given wheel installation parameters, without estimating the inertial measurement unit installation parameters each time, which can reduce the process required for vehicle positioning, reduce the model complexity, and improve the calculation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 is a schematic flow chart of a vehicle positioning method based on a wheel-mounted IMU of the present invention;

[0063] Figure 2 is a schematic installation diagram of the IMU mounted on the wheel in an embodiment of the present invention;

[0064] Figure 3 is a schematic diagram of the coordinate system and the definition of some variables in an embodiment of the present invention;

[0065] Figure 4 is a schematic diagram of the estimation result of the wheel IMU installation parameters in the first embodiment of the present invention;

[0066] Figure 5 is a schematic diagram of the vehicle positioning result in the first embodiment of the present invention;

[0067] Figure 6 is a schematic diagram of the vehicle azimuth estimation error in the first embodiment of the present invention;

[0068] Figure 7 is a schematic diagram of the vehicle speed estimation result in the first embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0069] The present invention will be further described below with reference to the accompanying drawings. The following embodiments can enable those skilled in the art to understand the present invention more comprehensively, but do not limit the present invention in any way.

[0070] In the embodiment of this solution, as Figure 2 shown, the MEMS IMU (inertial measurement unit) is fixed to any non-steering wheel of the vehicle. The computer wirelessly collects the IMU data at a sampling rate of 200 Hz. Assume that the vehicle is on a horizontal road surface, as Figure 3As shown in the figure, denote the IMU coordinate system as the b - system, whose origin is located at the center of the IMU (precisely, the sensitive center of the accelerometer), and the three axes are respectively along the three sensitive directions of the inertial measurement unit; denote the wheel coordinate system as the w - system, whose origin is located at the center of the wheel, its x - axis is perpendicular to the wheel rotation axis and points from the wheel center to the IMU center, the z - axis is along the wheel rotation axis and points to the left side of the vehicle body, and the y - axis is determined by the right - hand coordinate system rule; introduce the h - system such that the h - system coincides with the w - system after rotating by β along its z - axis according to the right - hand rule, and the h - system satisfies that its y - axis points upward along the vertical direction at the initial moment, and both the h - system and the w - system are fixed relative to the wheel. Denote the vehicle body coordinate system as the v - system, whose three axes point to the right of the vehicle, in front of the vehicle, and above the vehicle in sequence, and its origin is located at the contact point of the wheel with the IMU installed on the ground; the navigation coordinate system n - system is fixed relative to the earth, and the n - system coincides with the v - system at the initial moment; denote the wheel radius as R and the distance from the IMU center to the wheel rotation axis (i.e., the eccentricity) as r.

[0071] Define as the projection of the motion quantity s of the q - system relative to the p - system in the o - system. The motion quantity s can be angular velocity (denoted as w), position (denoted as p), velocity (denoted as v), and acceleration (denoted as a). Define the Euler angles by the following formula

[0072]

[0073] where represents the attitude transformation matrix from the p - system to the q - system, satisfying The superscript 'T' represents taking the transpose of the matrix; roll, pitch, and heading respectively refer to the roll angle, pitch angle, and azimuth angle. represents the attitude transformation matrix from the obtained coordinate system to the original coordinate system when the original coordinate system rotates by along the j - axis of this coordinate system according to the right - hand rule, and j represents x, y, or z. In this embodiment, is the first - order derivative of is the second - order derivative of is the third - order derivative of

[0074] Embodiment 1

[0075] Embodiment 1 provides a vehicle positioning method based on an inertial measurement unit installed on a wheel, and also provides a method for estimating the installation parameters of the wheel IMU. A vehicle positioning method based on an inertial measurement unit installed on a wheel, based on the inertial measurement unit installed on the vehicle wheel, in the case where the installation parameters of the inertial measurement unit are unknown, performs steps A - B, as Figure 1As shown, it is divided into two stages: wheel-IMU alignment and vehicle positioning, to obtain the installation parameters of the inertial measurement unit, as well as the real-time position, attitude and speed of the vehicle.

[0076] Step A: Perform wheel-IMU alignment. Based on the output data of the inertial measurement unit, use a preset attitude calculation algorithm to obtain the preliminary installation attitude matrix of the inertial measurement unit relative to the wheel; and obtain the gyro zero bias.

[0077] In the said Step A, the following steps are specifically executed to obtain the preliminary installation attitude matrix of the inertial measurement unit relative to the wheel:

[0078] Step A1: The vehicle is on a horizontal road surface and remains stationary. Collect the output data of the three-axis accelerometer in the inertial measurement unit during a preset time period when the vehicle is in a stopped state, and take the average value of the output data of the three-axis accelerometer during the preset time period, denoted as To reduce the influence of noise, take the average of the accelerometer outputs collected within 10 seconds;

[0079] Satisfy:

[0080]

[0081] Here, g is the local gravitational acceleration value;

[0082] Collect the output data of the three-axis gyroscope in the inertial measurement unit during a preset time period when the vehicle is in a stopped state, and take the average value of the output data of the three-axis gyroscope during the preset time period, denoted as the gyro zero bias; to reduce the influence of noise, take the average of the three-axis gyroscope outputs collected within 10 seconds.

[0083] Step A2: The vehicle travels a short distance on a horizontal road surface. Collect the output of the three-axis gyroscope, collect the output data of the three-axis gyroscope during a preset time period when the vehicle is in a driving state, and subtract the gyro zero bias from the output data of the three-axis gyroscope, and then obtain the average value of the output data of the three-axis gyroscope during the preset time period, denoted as To reduce the influence of noise, take the average of the three-axis gyroscope outputs collected within 50 seconds; i is the inertial coordinate system, ignoring small quantities such as the earth's rotation, and satisfy:

[0084]

[0085] Step A3: Based on the average value data obtained in Step A1 and Step A2, obtain the preliminary installation attitude matrix of the inertial measurement unit relative to the wheel through the attitude calculation algorithm shown in the following formula That is, the attitude transformation matrix of the h system relative to the b system

[0086]

[0087] Let be described by Euler angles, and the corresponding roll angle, pitch angle, and azimuth angle are denoted as γ h , θ h and ψ h respectively. Substituting Equation (3) into Equation (2), we get:

[0088]

[0089] Thus, γ h and θ h are calculated from the following formula:

[0090]

[0091] Here denote

[0092]

[0093] Substituting Equation (3) and Equation (4) into Equation (1) successively, we get:

[0094]

[0095] Thus, ψ h is calculated from the following formula:

[0096]

[0097] Furthermore, based on from γ h , θ h and ψ h it can be calculated that

[0098] where represents the attitude transformation matrix from the b system to the h system, that is, the installation attitude matrix of the inertial measurement unit relative to the wheel; the h system is a coordinate system that coincides with the w system after rotating β around its z axis according to the right-hand rule. The origin of the h system coincides with that of the w system. At the initial moment, its y axis points vertically upward; the w system represents the wheel coordinate system, whose origin is located at the center of the wheel, its x axis is perpendicular to the wheel rotation axis and points from the center of the wheel to the center of the inertial measurement unit, the z axis points to the left side of the vehicle body along the wheel rotation axis direction, and the y axis is determined by the right-hand coordinate system rule; both the h system and the w system are fixed relative to the wheel; the inertial measurement unit coordinate system is the b system, and its three axes are respectively along the three sensitive directions of the inertial measurement unit, and its origin is located at the center of the inertial measurement unit, that is, the sensitive center of the three-axis accelerometer; the i system is the inertial coordinate system.

[0099] Step B: Under the constraints of vehicle and wheel kinematics, based on the preliminary installation attitude matrix of the inertial measurement unit relative to the wheel, include the preset parameters of various types of wheels, the preset parameters of various types of vehicle body attitudes, and the preset parameters related to the installation of the inertial measurement unit in the state variables, include the output data of the inertial measurement unit and the body zero-level angle in the observed variables, and combine with the Kalman filter to obtain the installation parameters of the inertial measurement unit, as well as the real-time position, attitude, and speed of the vehicle.

[0100] In the said Step B, specifically perform the following steps to obtain the installation parameters of the inertial measurement unit, as well as the real-time position, attitude, and speed of the vehicle;

[0101] Step B1: For the state variables including the preset parameters of various types of wheels, the preset parameters of various types of vehicle body attitudes, and the preset parameters related to the installation of the inertial measurement unit, that is, the state variables Construct the state equation; for the observed variables including the output data of the inertial measurement unit and the body zero-level angle, that is, the observed variables Construct the observation equation; establish the Kalman filter, discretize the state equation and the observation equation, set the initial values, and enter Step B2; setting the initial values includes setting the initial value of the state variable X, setting the initial value of the state error covariance matrix as P(t0), setting the variance of the state noise as Q, setting the variance matrix of the observation noise V as R, and giving the initial azimuth angle and the initial longitude and latitude.

[0102] Among them, the preset parameters of various types of wheels include the wheel rotation angle α, the wheel angular velocity the wheel angular acceleration The preset parameters related to the installation of the inertial measurement unit include the distance between the center of the inertial measurement unit and the wheel rotation axis, that is, the eccentricity r, and β; the preset parameters of various types of vehicle body attitudes include the vehicle roll angle γ, the pitch angle θ, and the azimuth angle ψ; is the output data of the three-axis accelerometer in the actually measured inertial measurement unit; is the output data of the z-axis in the projection of the output angular velocity of the actually measured three-axis gyroscope in the h system.

[0103] Establishing the Kalman filter includes establishing the state equation and the observation equation of the extended Kalman filter. The constructed state equation and observation equation are as follows:

[0104] State equation:

[0105] Observation equation:

[0106]

[0107] Among them,

[0108]

[0109]

[0110]

[0111] In this embodiment, the following equality is established v =[0 0 -g] T ,

[0112] In the formula, g represents the magnitude of gravitational acceleration; It represents the attitude transformation matrix from the p system to the q system. Both the p system and the q system refer to coordinate systems, satisfying The superscript 'T' means that the matrix is ​​transposed; is the projection of the motion s of the q system relative to the p system on the o system. The motion s is the angular velocity w, position p, velocity v or acceleration a. η and ε are the preset process noises; η is caused by the uncertainty of the moving body, which is approximated as zero-mean Gaussian white noise; small quantities such as the rotation of the earth are ignored, and we have Considering the gyro output noise, the actual gyro output is recorded as ε includes small quantities such as gyro output noise and earth rotation, which are approximated as zero-mean Gaussian white noise; ∈、ζ γ and θ are all observation noises, and are approximately zero-mean Gaussian white noises. is the error and model error generated by the accelerometer, ∈ is the gyro error and model error, ζ is γ and θ Reflect the non-level condition of the vehicle body caused by the uneven road surface, etc., and approximate them as zero-mean Gaussian white noise; is the actual output of the accelerometer, which includes output noise;

[0113] The state equation and observation equation constructed in this embodiment are as follows: the influence of IMU installation parameters, vehicle body motion, and wheel motion on wheel IMU output is fully considered. According to the acceleration synthesis theorem of a point, the absolute acceleration of a moving point at a certain instant is equal to the vector sum of its involved acceleration, relative acceleration, and Coriolis acceleration at that instant. In the observation equation, the rotation of the earth is ignored, and the output of the accelerometer is It is mainly affected by the following accelerations: (1) Gravitational acceleration, which is the term in the observation equation: (2) The involved acceleration is the acceleration of the point (i.e., the involved point) where the moving reference coordinate system (i.e., the V system in this example) coincides with the moving point (i.e., the sensitive center of the accelerometer in this example). The term in the observation equation is: (The smaller terms containing are ignored here); (3) relative acceleration, caused by the rotation of the wheel relative to the vehicle body, and this term in the observation equation is (4) Coriolis acceleration, which is generated by the interaction between the transport motion and the relative motion when the moving reference coordinate system rotates (in this example, the vehicle is turning). The Coriolis acceleration is equal to twice the vector product of the angular velocity vector of the moving reference coordinate system and the relative velocity vector of the point, and this term in the observation equation is Furthermore, in this embodiment, the errors caused by the vehicle body motion, installation parameters, etc. are avoided, and the vehicle positioning accuracy based on the IMU installed on the wheel is improved. In the formula, in the state equation, since the gyro senses the wheel rotation and the vehicle body angular velocity, the gyro output is subtracted by the wheel rotation angular velocity to obtain the vehicle body angular velocity, as shown in the formula

[0114] For the Kalman filter, this filter is a highly efficient recursive filter. Based on the state equation and the measurement equation, the extended Kalman filtering process is as follows: X(t k ) is the system state at time t k , and Z(t k ) is the measured value at time t k ; First, the state of the system at the next moment needs to be predicted, that is, the time update. Assume that the current system state is at time t k . According to the system model, the current state can be predicted based on the previous state of the system:

[0115] X(t k |t k-1 ) = f(X(t k - 1|t k-1 ), U(t k ))

[0116] In the formula, X(t k |t k-1 ) is the result predicted using the previous state, X(t k-1 |t k-1 ) is the optimal result of the previous state, U(t k ) is the control quantity of the current state. If there is no control quantity, it can be 0, and f(·) is the nonlinear process equation; after the system state is updated, the covariance matrix corresponding to X(t k |t k-1 ) also needs to be updated. Let P represent the covariance matrix:

[0117] P(t k |t k-1 ) = AP(t k-1 |t k-1 )A T + BQB T

[0118] In the formula, P(t k |t k-1 ) is the state error covariance matrix corresponding to X(t k |t k-1 ), P(t k-1 |t k-1 ) is the state error covariance matrix corresponding to X(t k-1 |t k-1 ), A is the Jacobian matrix obtained by taking the partial derivative of the state equation with respect to the state quantity, A T represents the transpose matrix of A, Q is the covariance matrix of the process noise, B is the Jacobian matrix obtained by taking the partial derivative of the state equation with respect to the process noise variable, B T represents the transpose matrix of B, then the prediction of the current state of the system is completed, that is, the time update.

[0119] Then, based on the prediction result of the current state, combined with the measured value of the current state. The optimal estimated value X(t k |t k |t k ) of the state at time t is obtained, that is, the measurement update:

[0120] X(t k |t k ) = X(t k |t k-1 ) + Kg(t k )(Z(t k ) - h(X(t k |t k-1 )))

[0121] where Kg is the Kalman Gain:

[0122] Kg(t k ) = P(t k |t k-1 )H T / (HP(t k |t k-1 )H T + R)

[0123] In the formula, h(·) is the nonlinear observation equation, H is the Jacobian matrix of the observation equation, H T is the transpose matrix of H;

[0124] Then the optimal estimated value X(t k |t k |t k ) of the state at time t is obtained. The Kalman filter needs to run continuously until the end of the system process, and also update the X(t k |t k |tk )’s covariance matrix:

[0125] P(t k |t k )=(I-Kg(t k )H)P(t k |t k-1 )

[0126] Among them, I is the unit matrix, so far, t k The measurement update at time. When the system enters t k+1 At time P(t k |t k ) is the next state P(t k-1 |t k-1 ). The algorithm can then perform autoregressive operations until the vehicle positioning is completed.

[0127] Based on the above Kalman filter state update process, combined with the created state equation and observation equation, the following steps are updated and iterated based on the initial position of the vehicle. Step B2: Based on the constructed state equation and observation equation, using the Kalman filter, iteratively execute steps B2.1 to B2.4 until the vehicle positioning iteration is completed, and the inertial measurement unit installation parameters, as well as the real-time position, attitude and speed of the vehicle are obtained:

[0128] Step B2.1: Based on the preset sampling period, obtain the sampling time t k The output data of the inertial measurement unit, where the three-axis gyroscope output data of the inertial measurement unit minus the gyroscope zero bias; t k =kT, T is the sampling period, k is a positive integer;

[0129] Step B2.2: Perform time update, predict the state quantity through the state equation, calculate the Jacobian matrix of the state equation to predict the state error covariance matrix, specifically: based on the sampling time t k-1 The estimated value of the state quantity is obtained through the state equation, and the sampling time t is obtained k The predicted value of the state quantity; and based on the sampling time t k-1 The estimated value of the state error covariance matrix and the Jacobian matrix of the state equation are used to obtain the sampling time t k The corresponding state error covariance matrix prediction value;

[0130] Step B2.3: Perform measurement update, calculate the Jacobian matrix of the observation equation to calculate the gain matrix, estimate the state quantity and state error covariance matrix, specifically: based on the sampling time t k The predicted value of the state error covariance matrix, the observation noise variance matrix and the Jacobian matrix of the observation equation are used to obtain the sampling time t k Gain matrix; Based on the gain matrix, sampling time tk The predicted value of the state quantity, sampling time t k Observed quantity data and the observation equation to obtain the sampling time t k The estimated value of the state quantity; and based on the gain matrix, the Jacobian matrix of the observation equation, the sampling time t k The predicted value of the state error covariance matrix and the observation noise variance matrix to obtain the sampling time t k The estimated value of the state error covariance matrix;

[0131] Step B2.4: Based on the estimated value of the state quantity at sampling time t k To obtain the vehicle attitude at sampling time t k of the vehicle Vehicle position Eccentricity r and the installation attitude matrix of the inertial measurement unit relative to the wheel Update the position and IMU installation attitude parameters; return to step B2.1 and enter the iteration of the next sampling time.

[0132] In this embodiment, according to the actual application scenario, the Kalman filter can adopt not only the extended Kalman filter, but also the unscented Kalman filter, or the cubature Kalman filter, etc. When the installation parameter estimation of the inertial measurement unit is completed, the installation parameters of the inertial measurement unit can be used as known quantities for the next vehicle positioning, without estimating the installation parameters of the inertial measurement unit every time. This can reduce the processes required for vehicle positioning, reduce the model complexity, and improve the calculation efficiency.

[0133] Figure 4 Display the eccentricity r and The corresponding estimated results of the installation azimuth angle over time, and the estimated results tend to be stable in a short time. Figure 5 Display the positioning results of the vehicle in this embodiment and compare them with the true value trajectory obtained by the GPS / IMU integrated navigation system Mti-G-710. Figure 6 Display the schematic diagram of the vehicle azimuth estimation error in this embodiment, Figure 7 Display the schematic diagram of the vehicle speed estimation results in this embodiment and compare them with the true value speed obtained by Mti-G-710. The technology of this embodiment has high positioning accuracy and can obtain high-precision vehicle attitude and speed.

[0134] Embodiment 2

[0135] Based on the known installation parameters of the inertial measurement unit, the installation parameters are the installation attitude matrix of the wheel IMU Given the wheel radius R and the eccentricity r, this embodiment provides a vehicle positioning method based on an inertial measurement unit installed on the wheel. When the installation parameters of the inertial measurement unit are known, step C is executed to obtain the real-time position, attitude, and speed of the vehicle:

[0136] Step C: Under the constraints of vehicle and wheel kinematics, based on the known installation parameters of the inertial measurement unit, various preset wheel parameters and various preset vehicle body attitude parameters are included in the state variables, and the output data of the inertial measurement unit and the vehicle body zero-level angle are included in the observed variables. Combining with the Kalman filter, the real-time position, attitude, and speed of the vehicle are obtained.

[0137] In the said step C, based on the known installation parameters of the inertial measurement unit, the following steps are specifically executed to further obtain the real-time position, attitude, and speed of the vehicle;

[0138] Step C1: Collect the output data of the three-axis gyroscope in the inertial measurement unit within a preset time period when the vehicle is in a stopped state, and take the average value of the output data of the three-axis gyroscope within the preset time period, which is denoted as the gyro zero bias. To reduce the influence of noise, the average value of the output of the three-axis gyroscope collected within 10 seconds is taken.

[0139] Step C2: For the state variables including various preset wheel parameters and various preset vehicle body attitude parameters, i.e., the state variables Construct the state equation; for the observed variables including the output data of the inertial measurement unit and the vehicle body zero-level angle, i.e., the observed variables Construct the observation equation; establish the Kalman filter, discretize the state equation and the observation equation, and set the initial values. Setting the initial values includes setting the initial value of the state variable X, setting the initial value of the state error covariance matrix as P(t0), setting the variance of the state noise as Q, setting the variance matrix of the observation noise V as R, and giving the initial azimuth angle and the initial longitude and latitude.

[0140] Among them, α is the wheel rotation angle, is the wheel angular velocity, is the wheel angular acceleration; γ, θ, and ψ are the vehicle roll angle, pitch angle, and azimuth angle respectively; is the output data of the three-axis accelerometer measured in the inertial measurement unit; is the output data of the z-axis in the projection of the output angular velocity of the three-axis gyroscope measured in the inertial measurement unit in the w coordinate system; the w coordinate system represents the wheel coordinate system, the origin of which is located at the wheel center, the x-axis is perpendicular to the wheel axis of rotation, and points from the wheel center to the center of the inertial measurement unit, the z-axis points to the left side of the vehicle body along the wheel axis of rotation, and the y-axis is determined by the right-hand coordinate system rule. The w coordinate system is fixed relative to the wheel; the inertial measurement unit coordinate system is the b coordinate system, and the three axes are respectively along the three sensitive directions of the inertial measurement unit, and the origin is located at the center of the inertial measurement unit, that is, the sensitive center of the three-axis accelerometer; the i coordinate system is the inertial coordinate system.

[0141] When the installation parameters of the inertial measurement unit are known, establishing the Kalman filter includes establishing the state equation and observation equation of the extended Kalman filter. The constructed state equation and observation equation are as follows:

[0142] Equation of state:

[0143]

[0144] Observation equation:

[0145]

[0146] in,

[0147]

[0148]

[0149]

[0150]

[0151] In this embodiment, the following equality is established

[0152] In the formula, R is the wheel radius; g represents the magnitude of gravitational acceleration; Represents the attitude transformation matrix from the p system to the q system. Both the p system and the q system refer to coordinate systems, satisfying The superscript 'T' means that the matrix is ​​transposed; is the projection of the motion s of the q system relative to the p system on the o system. The motion s is the angular velocity w, position p, velocity v or acceleration a. The body coordinate system is the v system, whose three axes point to the right, front and top of the vehicle respectively, and its origin is located at the contact point between the wheel where the IMU is installed and the ground; the navigation coordinate system is the n system, which coincides with the v system at the initial moment relative to the solidification of the earth; η and ε are the preset process noises; η is caused by the uncertainty of the motion of the moving body, and it is approximated as zero-mean Gaussian white noise; ignoring small quantities such as the rotation of the earth, there is Considering the gyro output noise, the actual gyro output is recorded as ε includes small quantities such as gyro output noise and earth rotation, which are approximated as zero-mean Gaussian white noise; ∈、ζ γ and θ All are observation noises, and all are approximately zero-mean Gaussian white noise, which is the same as in the first embodiment; is the actual output of the accelerometer, which includes output noise;

[0153] Based on the process of updating the state using the Kalman filter described above, combined with the established state equation and observation equation, the following iterative update process is carried out based on the initial position; Step C3: Based on the established state equation and observation equation, use the Kalman filter to iteratively execute Step C3.1 to Step C3.4 until the vehicle positioning ends the iteration, and then obtain the real-time position, attitude, and speed of the vehicle:

[0154] Step C3.1: Based on the preset sampling period, obtain the sampling time t k The output data of the inertial measurement unit, where the output data of the three-axis gyroscope of the inertial measurement unit is subtracted by the gyro zero bias;

[0155] Step C3.2: Perform time update, predict the state quantity through the state equation, and calculate the Jacobian matrix of the state equation to predict the state error covariance matrix. Specifically: Based on the estimated value of the state quantity at sampling time t k-1 , obtain the predicted value of the state quantity at sampling time t k through the state equation; and based on the estimated value of the state error covariance matrix at sampling time t k-1 and the Jacobian matrix of the state equation, obtain the predicted value of the state error covariance matrix at sampling time t k corresponding thereto;

[0156] Step C3.3: Perform measurement update, calculate the Jacobian matrix of the observation equation to calculate the gain matrix, and estimate the state quantity and the state error covariance matrix. Specifically: Based on the predicted value of the state error covariance matrix at sampling time t k , the observation noise variance matrix, and the Jacobian matrix of the observation equation, obtain the gain matrix at sampling time t k ; based on the gain matrix, the predicted value of the state quantity at sampling time t k , the observed quantity data at sampling time t k and the observation equation, obtain the estimated value of the state quantity at sampling time t k ; and based on the gain matrix, the Jacobian matrix of the observation equation, the predicted value of the state error covariance matrix at sampling time t k and the observation noise variance matrix, obtain the estimated value of the state error covariance matrix at sampling time t k ;

[0157] Step C3.4: Based on the estimated value of the state quantity at sampling time t k , obtain the vehicle attitude at sampling time t k Vehicle position Return to Step C3.1 and enter the iteration of the next sampling time.

[0158] ​In this embodiment, according to the actual application scenario, the Kalman filter can adopt, in addition to the extended Kalman filter, the unscented Kalman filter, the cubature Kalman filter, etc.

[0159] The present invention designs a vehicle positioning method based on an inertial measurement unit mounted on a wheel. The present invention fully considers the influence of IMU installation parameters, vehicle body movement, and wheel movement on the output of the wheel IMU, establishes a high-precision vehicle positioning model based on the wheel-mounted IMU, constructs a state equation and an observation equation based on this model, and performs information fusion through the Kalman filtering algorithm to effectively estimate the vehicle position, speed, attitude, and wheel IMU installation parameters, improving the vehicle positioning accuracy based on the wheel-mounted IMU.

[0160] The above are only the preferred embodiments of the present invention, but do not limit the patent scope of the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, for those skilled in the art, they can still modify the technical solutions recorded in the foregoing specific embodiments, or perform equivalent replacements on some of the technical features. Any equivalent structure made by using the content of the specification and drawings of the present invention, directly or indirectly applied in other related technical fields, is similarly within the scope of the patent protection of the present invention.

Claims

1. A vehicle positioning method based on an inertial measurement unit mounted on a wheel, characterized in that: Based on an inertial measurement unit installed at a preset position of a vehicle wheel, when the installation parameters of the inertial measurement unit are unknown, steps A - B are executed to obtain the installation parameters of the inertial measurement unit, as well as the real - time position, attitude, and speed of the vehicle; when the installation parameters of the inertial measurement unit are known, step C is executed to obtain the real - time position, attitude, and speed of the vehicle: Step A: Based on the output data of the inertial measurement unit, a preliminary installation attitude matrix of the inertial measurement unit relative to the wheel is obtained by using a preset attitude solution algorithm. The above step A includes the following steps: Step A1: Collect the output data of the triaxial accelerometer in the inertial measurement unit within a preset time period when the vehicle is in a stopped state, and take the average value of the output data of the triaxial accelerometer within the preset time period, denoted as ; Step A2: Collect the output data of the triaxial gyroscope within a preset time period during vehicle driving, and then obtain the average value of the output data of the triaxial gyroscope within the preset time period, denoted as ; Based on and , the initial installation attitude matrix of the inertial measurement unit relative to the wheel is obtained by using an attitude solution algorithm , denotes the attitude transformation matrix from the coordinate system to the coordinate system, where the coordinate system is the inertial measurement unit coordinate system, and the coordinate system is the coordinate system that coincides with the axis after rotating by according to the right-hand rule and is the coordinate system representing the wheel coordinate system; Step B: Based on the preliminary installation attitude matrix of the inertial measurement unit relative to the wheel, first include the preset parameters of various types of wheels, the preset parameters of various types of vehicle body postures, and the preset parameters related to the installation of the inertial measurement unit in the state variables. For the state variables , construct a state equation; include the output data of the inertial measurement unit and the body zero-level angle in the observation variables. For the observation variables , construct an observation equation; and set the initial values; then, based on the state equation and the observation equation, use the Kalman filter to obtain the installation parameters of the inertial measurement unit and the real-time position, posture, and speed of the vehicle; Among them, is the wheel steering angle, is the wheel angular velocity, is the wheel angular acceleration; is the distance between the center of the inertial measurement unit and the wheel axis, i.e., the eccentricity; 、 and are the vehicle roll angle, pitch angle and azimuth angle respectively; is the output data of the triaxial accelerometer in the inertial measurement unit measured; is the projection of the output angular velocity of the triaxial gyroscope in the inertial measurement unit measured on the axis in the system; Step C: Based on the known installation parameters of the inertial measurement unit, first include the preset parameters of each type of wheel and the preset parameters of each type of vehicle body attitude in the state variables, and for the state variables , construct a state equation; include the output data of the inertial measurement unit and the vehicle body zero-level angle in the observation variables, and for the observation variables , construct an observation equation; set the initial value; then, based on the state equation and the observation equation, use the Kalman filter to obtain the real-time position, attitude, and speed of the vehicle; Wherein, is the wheel steering angle, is the wheel angular velocity, is the wheel angular acceleration; , and are respectively the vehicle roll angle, pitch angle and azimuth angle; is the output data of the three-axis accelerometer in the inertial measurement unit measured; is the projection of the output angular velocity of the three-axis gyroscope in the inertial measurement unit in the system on the axis of the output data; The system represents the wheel coordinate system, the origin of which is located at the wheel center, and its axis is perpendicular to the wheel rotation axis and points from the wheel center to the inertial measurement unit center, axis points to the left side of the vehicle body along the wheel rotation axis direction, axis is determined by the right-hand coordinate system rule, and the system is fixed relative to the wheel; the inertial measurement unit coordinate system is the system, and the three axes are respectively along the three sensitive directions of the inertial measurement unit, and the origin is located at the inertial measurement unit center, that is, the sensitive center of the three-axis accelerometer; The system is the inertial coordinate system.

2. The vehicle positioning method based on an inertia measurement unit mounted on a wheel according to claim 1, characterized in that: In the said step A, the following steps are specifically executed to obtain the preliminary installation attitude matrix of the inertial measurement unit relative to the wheel: Step A1: Collect the output data of the triaxial accelerometer in the inertial measurement unit within a preset time period when the vehicle is in a stopped state, and take the average value of the output data of the triaxial accelerometer within the preset time period, denoted as ; collect the output data of the triaxial gyroscope in the inertial measurement unit within a preset time period when the vehicle is in a stopped state, and take the average value of the output data of the triaxial gyroscope within the preset time period, denoted as the gyro zero bias; Step A2: Collect the output data of the three-axis gyroscope within a preset time period during vehicle driving, subtract the gyro zero bias from the output data of the three-axis gyroscope, and then obtain the average value of the output data of the three-axis gyroscope within the preset time period, denoted as ; Step A3: Based on the average value data obtained in Step A1 and Step A2, obtain the preliminary installation attitude matrix of the inertial measurement unit relative to the wheel through the attitude calculation algorithm shown in the following formula : ; Wherein, ， ； ， ； ； Among them, represents the attitude transformation matrix related to the system, that is, the preliminary installation attitude matrix of the inertial measurement unit relative to the wheel; The system is rotated by along its axis according to the right-hand rule and then coincides with the system. The origin of the system coincides with that of the system. At the initial moment of the system, its axis points upward along the vertical direction; The system represents the wheel coordinate system, whose origin is located at the wheel center. Its axis is perpendicular to the wheel rotation axis and points from the wheel center to the inertial measurement unit center. The axis is determined by the right-hand coordinate system rule; Both the system and the system are fixed relative to the wheel. The inertial measurement unit coordinate system is the system, and the three axial directions are respectively along the three sensitive directions of the inertial measurement unit. Its origin is located at the inertial measurement unit center, that is, the sensitive center of the three-axis accelerometer; ; represents the attitude transformation matrix from the obtained coordinate system after the original coordinate system is rotated by along its axis according to the right-hand rule to the original coordinate system. represents or axis; represents the pitch angle of the inertial measurement unit in the system, represents the roll angle of the inertial measurement unit in the system, represents the azimuth angle of the inertial measurement unit in the system.

3. The vehicle positioning method based on an inertial measurement unit mounted on a wheel according to claim 2, wherein, In step B, based on the state equation and the observation equation, a Kalman filter is used to obtain the installation parameters of the inertial measurement unit and the real - time position, attitude, and speed of the vehicle, which specifically includes the following: Based on the state equation and the observation equation, using a Kalman filter, steps B2.1 to B2.4 are iteratively executed until the vehicle positioning ends and the iteration ends, thereby obtaining the installation parameters of the inertial measurement unit and the real - time position, attitude, and speed of the vehicle: Step B2.1: Obtain the sampling moment based on a preset sampling period Output data of the inertial measurement unit, where the output data of the three-axis gyroscope of the inertial measurement unit is subtracted by the gyro zero bias; Step B2.2: Perform the Kalman filter time update, specifically: Based on the estimated value of the state quantity at the sampling moment , the predicted value of the state quantity at the sampling moment is obtained through the state equation; and based on the estimated value of the state error covariance matrix at the sampling moment and the state equation, the predicted value of the state error covariance matrix at the sampling moment is obtained; Step B2.3: Perform Kalman filter measurement update, specifically: Based on the predicted value of the state error covariance matrix, the observation noise variance matrix, and the observation equation at the sampling time , obtain the gain matrix at the sampling time ; Based on the gain matrix, the predicted value of the state quantity at the sampling time , the observed quantity data at the sampling time , and the observation equation, obtain the estimated value of the state quantity at the sampling time ; And based on the gain matrix, the predicted value of the state error covariance matrix at the sampling time , the observation noise variance matrix, and the observation equation, obtain the estimated value of the state error covariance matrix at the sampling time . Step B2.4: Based on the state quantity estimation value at the sampling moment , obtain the vehicle attitude , vehicle position , eccentricity , and the installation attitude matrix of the inertial measurement unit relative to the wheel , where ; is the wheel radius; Return to step B2.1 and enter the iteration at the next sampling moment.

4. The vehicle positioning method based on an inertia measurement unit mounted on a wheel according to claim 3, characterized in that: The constructed state equation and observation equation are as follows: Equation of state: ; Observation equation: ; Wherein, ； ； ； ； ； ； ； ； ； ； ； ； wherein is the wheel radius; represents the magnitude of the acceleration due to gravity; represents the attitude transformation matrix related to the system, the system and the system both refer to coordinate systems, satisfying , and the superscript 'T' represents taking the transpose of the matrix; is the motion quantity of the system relative to the system projected in the system, and the motion quantity is the angular velocity , position , velocity or acceleration , ; is the output angular velocity of the three-axis gyroscope in the measured inertial measurement unit; the body coordinate system is the system, and its three axes point to the right side of the vehicle, the front of the vehicle, and above the vehicle in sequence, and its origin is located at the contact point between the wheel where the IMU is installed and the ground; the navigation coordinate system is the system, fixed relative to the earth, the system coincides with the system at the initial moment; and are the preset process noises; , , and are all preset observation noises.

5. The vehicle positioning method based on an inertia measurement unit mounted on a wheel according to claim 1, wherein, In step C, based on the state equation and the observation equation, a Kalman filter is used to obtain the real - time position, attitude, and speed of the vehicle, which specifically includes the following: First, the output data of the three - axis gyroscope in the inertial measurement unit within a preset time period when the vehicle is in a stopped state is collected, and the average value of the output data of the three - axis gyroscope within the preset time period is taken, denoted as the gyro zero - bias. Then, based on the constructed state equation and observation equation, using a Kalman filter, steps C3.1 to C3.4 are iteratively executed until the vehicle positioning ends and the iteration ends, thereby obtaining the real - time position, attitude, and speed of the vehicle: Step C3.1: Obtain the sampling moment based on a preset sampling period Output data of the inertial measurement unit, where the output data of the three-axis gyroscope of the inertial measurement unit is subtracted by the gyro zero bias; Step C3.2: Perform the Kalman filter time update, specifically: Based on the estimated value of the state quantity at the sampling time , obtain the predicted value of the state quantity at the sampling time through the state equation; and based on the estimated value of the state error covariance matrix at the sampling time and the state equation, obtain the predicted value of the state error covariance matrix at the sampling time ; Step C3.3: Perform Kalman filter measurement update, specifically: Based on the predicted value of the state error covariance matrix, the observation noise variance matrix, and the observation equation at the sampling time , obtain the gain matrix at the sampling time ; Based on the gain matrix, the predicted value of the state quantity at the sampling time , the observed quantity data at the sampling time , and the observation equation, obtain the estimated value of the state quantity at the sampling time ; And based on the gain matrix, the predicted value of the state error covariance matrix at the sampling time , the observation noise variance matrix, and the observation equation, obtain the estimated value of the state error covariance matrix at the sampling time . Step C3.4: Based on the state quantity estimation value at the sampling moment , obtain the vehicle attitude and vehicle position at the sampling moment , where is the wheel radius; Return to step C3.1 and enter the iteration at the next sampling moment.

6. The vehicle positioning method based on an inertia measurement unit mounted on a wheel according to claim 5, wherein: The constructed state equation and observation equation are as follows: State equation: ； Observation equation: ； Wherein, ； ； ； ； ； ； ； ； ； ； In the formula, is the wheel radius; represents the magnitude of gravitational acceleration; represents the attitude transformation matrix related to the system, the system and the system both refer to coordinate systems; is the motion quantity of the system relative to the system under the projection of the system, the motion quantity is the angular velocity , position , velocity or acceleration , ; is the output angular velocity of the three-axis gyroscope in the measured inertial measurement unit; the body coordinate system is the system, its three axes point to the right of the vehicle, in front of the vehicle, and above the vehicle in sequence, and its origin is located at the contact point between the wheel where the IMU is installed and the ground; the navigation coordinate system is the system, fixed relative to the earth, the system coincides with the system at the initial moment; and are the preset process noises; , , and are all preset observation noises.

7. The vehicle positioning method based on an inertia measurement unit mounted on a wheel according to claim 1, characterized in that: The installation parameters of the inertial measurement unit include the attitude transformation matrix of the inertial measurement unit coordinate system relative to the wheel coordinate system , and the distance between the center of the inertial measurement unit and the wheel rotation axis .

8. The vehicle positioning method based on an inertial measurement unit mounted on a wheel according to claim 1, characterized in that: The Kalman filter uses an extended Kalman filter, an unscented Kalman filter, or a cubature Kalman filter.

Images