Vehicle-mounted IMU online calibration method and apparatus, electronic device, and storage medium

Abstract

A vehicle-mounted IMU online calibration method and apparatus, an electronic device, and a storage medium. The method comprises: acquiring GNSS data and dynamic IMU data collected in a traveling process of a vehicle, wherein the traveling process at least comprises a straight-going process and a turning process (S110); identifying, from the GNSS data and the dynamic IMU data, turning IMU data and turning GNSS data corresponding to the turning process of the vehicle, and calculating a roll angle observed value of an IMU on the basis of the turning IMU data and the turning GNSS data (S120); identifying, from the GNSS data and the dynamic IMU data, straight-going IMU data and straight-going GNSS data corresponding to the straight-going process of the vehicle, and on the basis of the straight-going IMU data and the straight-going GNSS data, calculating a pitch angle observed value and a yaw angle observed value of the IMU (S130); and for parameters to be calibrated, inputting the corresponding observed values into optimizers of corresponding types, and obtaining, on the basis of the optimizers, predicted values of the parameters to be calibrated as calibration results of the parameters to be calibrated, wherein the parameters to be calibrated comprise: a roll angle, a pitch angle, and a yaw angle, and the optimizers comprise: a roll angle optimizer, a pitch angle optimizer, and a yaw angle optimizer (S140).

Description

A method, apparatus, electronic device and storage medium for online calibration of vehicle-mounted IMU Technical Field

[0001] This application relates to the field of IMU calibration technology, and more specifically, to an online calibration method, apparatus, electronic device, and storage medium for vehicle-mounted IMUs. Background Technology

[0002] An IMU (Inertial Measurement Unit) is an electronic component used to measure the inertial motion of a moving object. It outputs information such as three-axis acceleration and three-axis angular velocity, and is commonly used for measuring attitude angles and motion paths. When used in vehicles, an IMU can measure parameters such as acceleration and attitude during vehicle operation. In vehicle applications, online calibration is sometimes necessary to improve the accuracy of subsequent measurements. Online calibration (OLC) is the process by which mass-produced vehicles or internally developed vehicles continuously collect data and perform calculations during manual driving to obtain the latest calibration parameters. These parameters are automatically updated when conditions are met, and the entire process is seamless for the user. Specifically, the IMU's roll, pitch, and yaw angles need to be calibrated.

[0003] Known online IMU calibration methods only require the IMU's own data to estimate various extrinsic parameters. However, this method is only suitable for high-precision IMUs. Typically, vehicle-mounted IMUs have poor accuracy, leading to significant errors in the calibration results. Therefore, improving the accuracy of online calibration for vehicle-mounted IMUs has become a pressing technical problem. Summary of the Invention

[0004] This application provides a method, apparatus, electronic device, and storage medium for online calibration of vehicle-mounted IMUs, which can improve the accuracy of online calibration of vehicle-mounted IMUs. The specific technical solution is as follows.

[0005] In a first aspect, embodiments of this application provide an online calibration method for an inertial measurement unit (IMU) mounted on a vehicle, wherein the IMU is installed in a vehicle and the vehicle is also equipped with at least a Global Navigation Satellite System (GNSS). The method includes:

[0006] Acquire GNSS data and dynamic IMU data collected during the vehicle's driving process; the driving process includes at least a straight-line process and a turning process;

[0007] In the GNSS data and dynamic IMU data, identify the turning IMU data and turning GNSS data corresponding to the vehicle turning process, and calculate the roll angle observation value of the IMU based on the turning IMU data and turning GNSS data;

[0008] In the GNSS data and dynamic IMU data, identify the straight-moving IMU data and straight-moving GNSS data corresponding to the straight-moving process of the vehicle, and calculate the pitch angle observation value and yaw angle observation value of the IMU based on the straight-moving IMU data and straight-moving GNSS data;

[0009] For each parameter to be calibrated, the corresponding observation value is input into the corresponding type of optimizer, and the predicted value of the parameter to be calibrated is obtained based on the optimizer, which is used as the calibration result of the parameter to be calibrated; the parameters to be calibrated include: roll angle, pitch angle and yaw angle, and the optimizers include: roll angle optimizer, pitch angle optimizer and yaw angle optimizer.

[0010] In this embodiment, GNSS (Global Navigation Satellite System) and IMU can be fused for positioning, thereby enabling online calibration of the IMU. This method can satisfy the calibration of IMU extrinsic parameters of arbitrary accuracy. Therefore, for vehicle-mounted IMUs with low accuracy, their online calibration accuracy can be improved.

[0011] Optionally, the step of calculating the pitch angle observation value and yaw angle observation value of the IMU based on the straight-line IMU data and the straight-line GNSS data includes:

[0012] Based on the straight-line IMU data and the straight-line GNSS data, estimate the world attitude of the IMU in the world coordinate system;

[0013] Based on the world attitude, calculate the IMU velocity in the IMU coordinate system;

[0014] Based on the rotational extrinsic parameters of the IMU and the vehicle, the IMU velocity is converted into the vehicle velocity of the IMU in the vehicle coordinate system;

[0015] Based on the vehicle speed and according to non-integrity constraints, the pitch angle and yaw angle observations of the IMU are calculated.

[0016] Optionally, the step of estimating the world attitude of the IMU in the world coordinate system based on the straight-line IMU data and the straight-line GNSS data includes:

[0017] The Kalman state estimation filter is predicted based on the straight-line IMU data, and the Kalman state estimation filter is updated by observation based on the straight-line GNSS data to obtain the world attitude of the IMU in the world coordinate system; the state vector of the Kalman state estimation filter includes at least the world attitude of the IMU in the world coordinate system.

[0018] Optionally, the step of calculating the pitch angle and yaw angle observations of the IMU based on the vehicle speed and according to non-integrity constraints includes:

[0019] The pitch bias and yaw bias of the IMU are calculated using the following formulas:

[0020] Among them, v x Let v be the component of the vehicle's velocity along the x-direction. y Let v be the component of the vehicle's velocity along the y-direction. z Let be the component of the vehicle's velocity along the z-direction.

[0021] Optionally, the step of calculating the roll angle observation of the IMU based on the turning IMU data and the turning GNSS data includes:

[0022] Calculate the actual centripetal acceleration of the vehicle based on the turning GNSS data;

[0023] The lateral acceleration of the vehicle is identified from the turning IMU data;

[0024] The roll angle observation of the IMU is calculated based on the actual centripetal acceleration and the lateral acceleration.

[0025] Optionally, the step of calculating the roll angle observation of the IMU based on the actual centripetal acceleration and the lateral acceleration includes:

[0026] The roll angle observation θ of the IMU is calculated using the following formula:

[0027] Among them, a c Let a be the actual centripetal acceleration. y Let g be the lateral acceleration, and g be the gravitational acceleration.

[0028] Optionally, the optimizer includes a roll angle Kalman filter, a pitch angle Kalman filter, and a yaw angle Kalman filter, and the state vector of each Kalman filter remains unchanged during the prediction process.

[0029] Secondly, embodiments of this application provide an online calibration device for an in-vehicle IMU, wherein the IMU is installed in a vehicle, and the vehicle is at least equipped with a Global Navigation Satellite System (GNSS). The device includes:

[0030] The data acquisition module is used to acquire GNSS data and dynamic IMU data collected during the vehicle's driving process; the driving process includes at least a straight-line process and a turning process;

[0031] The first observation module is used to identify the turning IMU data and turning GNSS data corresponding to the vehicle turning process in the GNSS data and dynamic IMU data, and to calculate the roll angle observation value of the IMU based on the turning IMU data and turning GNSS data.

[0032] The second observation module is used to identify the straight-moving IMU data and straight-moving GNSS data corresponding to the straight-moving process of the vehicle in the GNSS data and dynamic IMU data, and to calculate the pitch angle observation value and yaw angle observation value of the IMU based on the straight-moving IMU data and straight-moving GNSS data.

[0033] The parameter optimization module is used to input the corresponding observation values ​​into the corresponding type of optimizer for each parameter to be calibrated, and obtain the predicted value of the parameter to be calibrated based on the optimizer, which is used as the calibration result of the parameter to be calibrated; the parameters to be calibrated include: roll angle, pitch angle and yaw angle, and the optimizers include: roll angle optimizer, pitch angle optimizer and yaw angle optimizer.

[0034] Optionally, the second observation module is specifically used for:

[0035] Based on the straight-line IMU data and the straight-line GNSS data, estimate the world attitude of the IMU in the world coordinate system;

[0036] Based on the world attitude, calculate the IMU velocity in the IMU coordinate system;

[0037] Based on the rotational extrinsic parameters of the IMU and the vehicle, the IMU velocity is converted into the vehicle velocity of the IMU in the vehicle coordinate system;

[0038] Based on the vehicle speed and according to non-integrity constraints, the pitch angle and yaw angle observations of the IMU are calculated.

[0039] Optionally, the second observation module is specifically used for:

[0040] The Kalman state estimation filter is predicted based on the straight-line IMU data, and the Kalman state estimation filter is updated by observation based on the straight-line GNSS data to obtain the world attitude of the IMU in the world coordinate system; the state vector of the Kalman state estimation filter includes at least the world attitude of the IMU in the world coordinate system.

[0041] Optionally, the second observation module is specifically used for:

[0042] The pitch bias and yaw bias of the IMU are calculated using the following formulas:

[0043] Among them, v x Let v be the component of the vehicle's velocity along the x-direction. y Let v be the component of the vehicle's velocity along the y-direction. z Let be the component of the vehicle's velocity along the z-direction.

[0044] Optionally, the first observation module is specifically used for:

[0045] Calculate the actual centripetal acceleration of the vehicle based on the turning GNSS data;

[0046] The lateral acceleration of the vehicle is identified from the turning IMU data;

[0047] The roll angle observation of the IMU is calculated based on the actual centripetal acceleration and the lateral acceleration.

[0048] Optionally, the first observation module is specifically used for:

[0049] The roll angle observation θ of the IMU is calculated using the following formula:

[0050] Among them, a c Let a be the actual centripetal acceleration. y Let g be the lateral acceleration, and g be the gravitational acceleration.

[0051] Optionally, the optimizer includes a roll angle Kalman filter, a pitch angle Kalman filter, and a yaw angle Kalman filter, and the state vector of each Kalman filter remains unchanged during the prediction process.

[0052] Thirdly, embodiments of this application provide a computer device, including: a memory and a processor, wherein the memory and the processor are coupled together;

[0053] The memory is used to store one or more computer instructions;

[0054] The processor is used to execute one or more computer instructions to implement the online calibration method for the vehicle-mounted IMU as described in the first aspect.

[0055] Fourthly, embodiments of this application provide a computer-readable storage medium storing one or more computer instructions that are executed by a processor to implement the online calibration method for an on-board IMU as described in the first aspect above.

[0056] Fifthly, this application provides a computer program product, which includes a computer program that, when executed by a processor, implements the vehicle-mounted IMU online calibration method described in the first aspect. Attached Figure Description

[0057] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0058] Figure 1 is a flowchart illustrating an online calibration method for an on-board IMU provided in an embodiment of this application;

[0059] Figure 2 is a schematic diagram of the structure of an on-board IMU online calibration device provided in an embodiment of this application;

[0060] Figure 3 is a schematic diagram of a computer device provided in an embodiment of this application. Detailed Implementation

[0061] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0062] It should be noted that the terms "comprising" and "having," and any variations thereof, in the embodiments and accompanying drawings of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the steps or units listed, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0063] This application discloses an online calibration method, apparatus, electronic device, and storage medium for vehicle-mounted IMUs, which can improve the accuracy of online calibration of vehicle-mounted IMUs. The embodiments of this application are described in detail below.

[0064] Online calibration (OLC) refers to the process where mass-produced vehicles or internally developed vehicles continuously collect and calculate data during manual driving to obtain the latest calibration parameters. These parameters are automatically updated when conditions are met, without the user noticing. Specifically, users can filter relevant data from their previous driving data; turning data can be used for roll calibration, while straight-line data can be used for pitch and yaw calibration.

[0065] As shown in Figure 1, the online calibration method for vehicle-mounted IMUs provided in this embodiment may include the following steps:

[0066] S110: Acquire GNSS data and dynamic IMU data collected during vehicle operation; the operation process includes at least straight-line driving and turning.

[0067] S120: In GNSS data and dynamic IMU data, identify the turning IMU data and turning GNSS data corresponding to the vehicle turning process, and calculate the IMU roll angle observation value based on the turning IMU data and turning GNSS data;

[0068] S130: In GNSS data and dynamic IMU data, identify the straight-moving IMU data and straight-moving GNSS data corresponding to the straight-moving process of the vehicle, and calculate the pitch angle observation value and yaw angle observation value of the IMU based on the straight-moving IMU data and straight-moving GNSS data;

[0069] S140: For each parameter to be calibrated, the corresponding observation value is input into the corresponding type of optimizer, and the predicted value of the parameter to be calibrated is obtained based on the optimizer, which is used as the calibration result of the parameter to be calibrated; the parameters to be calibrated include: roll angle, pitch angle and yaw angle, and the optimizers include: roll angle optimizer, pitch angle optimizer and yaw angle optimizer.

[0070] The GNSS data mentioned above may include: GNSS latitude and longitude, altitude, GNSS status, GNSS hdop value (horizontal accuracy), number of GNSS satellites, and GNSS velocity in the northeast-northeast coordinate system; IMU data may include: IMU's three-axis acceleration data acc_x, acc_y, acc_z, and three-axis gyroscope data gyro_x, gyro_y, gyro_z.

[0071] During vehicle operation, the system can detect the vehicle. When online calibration is required, the system can filter the necessary data from the vehicle's historical driving data, including GNSS data and dynamic IMU data. Specifically, it can filter turning data and straight-ahead data. Turning data is used for roll calibration, while straight-ahead data is used for pitch and yaw calibration. There are no speed requirements for the filtered data, but the IMU's gyro_z data must meet certain ranges. For example, the IMU's gyro_z data for turning data needs to be greater than 0.15 radians / second and less than 0.35 radians / second; the gyro_z data for straight-ahead data needs to be less than 0.005 radians / second.

[0072] After acquiring GNSS and dynamic IMU data, the turning IMU data and turning GNSS data corresponding to the vehicle's turning process can be identified from the GNSS and dynamic IMU data. Based on the turning IMU data and turning GNSS data, the IMU roll angle observation value can be calculated. Similarly, the straight-moving IMU data and straight-moving GNSS data corresponding to the vehicle's straight-moving process can be identified from the GNSS and dynamic IMU data. Based on the straight-moving IMU data and straight-moving GNSS data, the IMU pitch angle observation value and yaw angle observation value can be calculated. For each parameter to be calibrated, the corresponding observation value is input into the corresponding type of optimizer. Based on the optimizer, the predicted value of the parameter to be calibrated is obtained, which is used as the calibration result for that parameter. The parameters to be calibrated include: roll angle, pitch angle, and yaw angle. The optimizers include: roll angle optimizer, pitch angle optimizer, and yaw angle optimizer.

[0073] In this embodiment, the optimizer may include a roll angle Kalman filter, a pitch angle Kalman filter, and a yaw angle Kalman filter. During prediction, the state vectors of each Kalman filter remain unchanged. Specifically, each Kalman filter can be an error state Kalman filter (ESKF). That is, a small Kalman filter can be maintained for each quantity to be optimized, used to optimize each quantity. During optimization, the result of a single measurement is input into the Kalman filter as an observation, and the Kalman filter estimator outputs the optimal estimate.

[0074] In one implementation, the roll angle can be estimated using the vehicle trajectory measured by GNSS and the lateral acceleration measured by IMU when the vehicle is turning. The basic principle is as follows: Ideally, without roll angle error, the lateral acceleration measured by the IMU should be equal to the centripetal acceleration calculated by GNSS. If roll angle error exists, the lateral acceleration measured by the IMU will deviate. Therefore, the calculation of the acceleration difference is the basis for calculating the roll angle error.

[0075] Specifically, the roll angle calculation process for a single instance, that is, the steps for calculating the roll angle observation value of the IMU, may include: calculating the actual centripetal acceleration of the vehicle based on the turning GNSS data; identifying the lateral acceleration of the vehicle in the turning IMU data; and calculating the roll angle observation value of the IMU based on the actual centripetal acceleration and lateral acceleration.

[0076] Vehicle speed can be obtained from GNSS data. Angular velocity can be obtained through GNSS and IMU fusion positioning. Therefore, the actual centripetal acceleration of the vehicle can be obtained by multiplying the vehicle speed by its angular velocity using a variation of the centripetal acceleration formula. IMU data includes the vehicle's lateral acceleration, so the IMU roll angle observation θ can be calculated using the following formula:

[0077] Among them, a c For the actual centripetal acceleration, a y denoted as lateral acceleration, and g as gravitational acceleration.

[0078] The NHC (Non-Holonomic Constraint) assumes that land vehicles neither jump off the ground nor slide along it; therefore, both of the vehicle's velocity components in the plane perpendicular to the direction of travel (e.g., the x-axis) are zero. Thus, the lateral (y-axis) and vertical (z-axis) velocity components in the vehicle frame can be expressed as: v y ≈0, v z ≈0, where v y and v z These represent the velocity components of the vehicle in the plane perpendicular to the direction of travel (x-axis).

[0079] Based on NHC constraints, in this embodiment of the application, the pitch angle and yaw angle observations of the IMU can be calculated based on the straight-moving IMU data and straight-moving GNSS data corresponding to the straight-moving process of the vehicle.

[0080] Specifically, in one implementation, the steps of calculating the pitch and yaw angle observations of the IMU based on the straight-moving IMU data and the straight-moving GNSS data may include: estimating the world attitude of the IMU in the world coordinate system based on the straight-moving IMU data and the straight-moving GNSS data; calculating the IMU velocity in the IMU coordinate system based on the world attitude; converting the IMU velocity into the vehicle velocity in the vehicle coordinate system based on the rotational extrinsic parameters of the IMU and the vehicle; and calculating the pitch and yaw angle observations of the IMU based on the vehicle velocity and according to non-integrity constraints.

[0081] In other words, GNSS can measure the vehicle's velocity in the world coordinate system. Then, through GNSS and IMU fusion positioning, the IMU's attitude in the world coordinate system can be estimated. Thus, the velocity of GNSS in the world coordinate system can be converted into the velocity of IMU in the IMU coordinate system through the attitude positioning of IMU in the world coordinate system. At the same time, through the rotational extrinsic parameters of IMU and vehicle body, it can be further converted into the velocity of IMU in vehicle body coordinate system. Finally, through the NHC constraint that the lateral and longitudinal velocities of the vehicle body are normally 0, the observed values ​​of the IMU's installation angle pitch and yaw can be obtained.

[0082] In this embodiment, an extended Kalman filter, also known as a Kalman state estimation filter, can be used to estimate the attitude of the IMU. For example, the filter can include the following 15-dimensional state vector: x = [position(3), orientation(3), velocity(3), bias_acc(3), bias_gyro(3)], where position represents the position of the IMU in the world coordinate system, orientation represents the attitude of the IMU in the world coordinate system, velocity represents the velocity of the IMU in the world coordinate system, bias_acc represents the zero bias of acceleration, and bias_gyro represents the zero bias of gyroscope. Each parameter is a 3-dimensional vector.

[0083] Using straight-line IMU data and straight-line GNSS data, the extended Kalman filter is updated for prediction, ultimately estimating the IMU's attitude (orientation_w2i) and velocity (v_w) in the world coordinate system. Further, v_i = orientation_w2i * v_w is calculated to obtain the IMU velocity v_i in the IMU coordinate system. Therefore, based on the IMU's rotational extrinsic parameter orientation_design_i2b relative to the vehicle, v_b = orientation_design_i2b * v_i is calculated to obtain the vehicle's velocity v_b in the vehicle coordinate system. Here, the origin of the vehicle coordinate system is the center of the vehicle's rear axle, the vehicle's forward direction is the positive x-axis, the left side of the vehicle is the positive y-axis, and the vertically upward direction is the positive z-axis.

[0084] After obtaining the vehicle velocity of the IMU in the vehicle coordinate system, the pitch-bias and yaw-bias of the IMU can be calculated using the following formulas:

[0085] Among them, v x Let v be the component of the vehicle's velocity along the x-direction. yLet v be the component of the vehicle's velocity along the y-direction. z Let z be the component of the vehicle's velocity along the z-direction.

[0086] In this embodiment, each parameter to be calibrated, namely roll, pitch, and yaw, can have its own corresponding optimizer, such as a Kalman filter. Therefore, after calculating each observation, the corresponding optimizer can be used to optimize the parameter to be calibrated. For example, for the pitch optimizer, its state vector x = [pitch], the prediction process x remains unchanged, the update process z = pitch_bias, H = 1, and then the EKF (Error Kalman Filter) update formula can be used.

[0087] In this embodiment, GNSS (Global Navigation Satellite System) and IMU can be fused for positioning, thereby enabling online calibration of the IMU. This method can satisfy the calibration of IMU extrinsic parameters of arbitrary accuracy. Therefore, for vehicle-mounted IMUs with low accuracy, their online calibration accuracy can be improved.

[0088] Figure 2 shows a schematic diagram of a vehicle-mounted IMU online calibration device according to an embodiment of this application. The IMU is installed in a vehicle, and the vehicle is at least equipped with a Global Navigation Satellite System (GNSS). The device includes:

[0089] The data acquisition module 210 is used to acquire GNSS data and dynamic IMU data collected during the vehicle's driving process; the driving process includes at least a straight-line process and a turning process;

[0090] The first observation module 220 is used to identify the turning IMU data and turning GNSS data corresponding to the vehicle turning process in the GNSS data and dynamic IMU data, and to calculate the roll angle observation value of the IMU based on the turning IMU data and turning GNSS data.

[0091] The second observation module 230 is used to identify the straight-moving IMU data and straight-moving GNSS data corresponding to the straight-moving process of the vehicle in the GNSS data and dynamic IMU data, and to calculate the pitch angle observation value and yaw angle observation value of the IMU based on the straight-moving IMU data and straight-moving GNSS data.

[0092] The parameter optimization module 240 is used to input the corresponding observation value into the corresponding type of optimizer for each parameter to be calibrated, and obtain the predicted value of the parameter to be calibrated based on the optimizer, which is used as the calibration result of the parameter to be calibrated; the parameters to be calibrated include: roll angle, pitch angle and yaw angle, and the optimizer includes: roll angle optimizer, pitch angle optimizer and yaw angle optimizer.

[0093] Optionally, the second observation module is specifically used for:

[0094] Based on the straight-line IMU data and the straight-line GNSS data, estimate the world attitude of the IMU in the world coordinate system;

[0095] Based on the world attitude, calculate the IMU velocity in the IMU coordinate system;

[0096] Based on the rotational extrinsic parameters of the IMU and the vehicle, the IMU velocity is converted into the vehicle velocity of the IMU in the vehicle coordinate system;

[0097] Based on the vehicle speed and according to non-integrity constraints, the pitch angle and yaw angle observations of the IMU are calculated.

[0098] Optionally, the second observation module is specifically used for:

[0099] The Kalman state estimation filter is predicted based on the straight-line IMU data, and the Kalman state estimation filter is updated by observation based on the straight-line GNSS data to obtain the world attitude of the IMU in the world coordinate system; the state vector of the Kalman state estimation filter includes at least the world attitude of the IMU in the world coordinate system.

[0100] Optionally, the second observation module is specifically used for:

[0101] The pitch bias and yaw bias of the IMU are calculated using the following formulas:

[0102] Among them, v x Let v be the component of the vehicle's velocity along the x-direction. y Let v be the component of the vehicle's velocity along the y-direction. z Let be the component of the vehicle's velocity along the z-direction.

[0103] Optionally, the first observation module is specifically used for:

[0104] Calculate the actual centripetal acceleration of the vehicle based on the turning GNSS data;

[0105] The lateral acceleration of the vehicle is identified from the turning IMU data;

[0106] The roll angle observation of the IMU is calculated based on the actual centripetal acceleration and the lateral acceleration.

[0107] Optionally, the first observation module is specifically used for:

[0108] The roll angle observation θ of the IMU is calculated using the following formula:

[0109] Among them, a c Let a be the actual centripetal acceleration. y Let g be the lateral acceleration, and g be the gravitational acceleration.

[0110] Optionally, the optimizer includes a roll angle Kalman filter, a pitch angle Kalman filter, and a yaw angle Kalman filter, and the state vector of each Kalman filter remains unchanged during the prediction process.

[0111] In this embodiment, GNSS (Global Navigation Satellite System) and IMU can be fused for positioning, thereby enabling online calibration of the IMU. This method can satisfy the calibration of IMU extrinsic parameters of arbitrary accuracy. Therefore, for vehicle-mounted IMUs with low accuracy, their online calibration accuracy can be improved.

[0112] The following describes a computer device provided in an embodiment of this application. Please refer to Figure 3, which is a schematic diagram of the structure of a computer device provided in an embodiment of this application. The computer device includes:

[0113] One or more processors 40;

[0114] The processor 40 is coupled to a storage device 41, which is used to store one or more programs.

[0115] When the one or more programs are executed by the one or more processors 40, the electronic device implements the technical solution of the vehicle-mounted IMU online calibration method as described above.

[0116] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the technical solution of the online calibration method for an on-board IMU as described above.

[0117] This application provides a computer program product, which includes a computer program that, when executed by a processor, implements the technical solution of the online calibration method for an on-board IMU as described above.

[0118] Those skilled in the art will understand that the accompanying drawings are merely schematic diagrams of one embodiment, and the modules or processes shown in the drawings are not necessarily essential for implementing this application.

[0119] Those skilled in the art will understand that the modules in the apparatus of the embodiments can be distributed in the apparatus of the embodiments as described in the embodiments, or they can be located in one or more devices different from this embodiment with corresponding changes. The modules of the above embodiments can be combined into one module, or they can be further divided into multiple sub-modules.

[0120] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for online calibration of a vehicle-mounted inertial measurement unit (IMU), characterized in that, The IMU is installed in the vehicle, and the vehicle is also equipped with at least a Global Navigation Satellite System (GNSS). The method includes: Acquire GNSS data and dynamic IMU data collected during the vehicle's driving process; the driving process includes at least a straight-line process and a turning process; In the GNSS data and dynamic IMU data, identify the turning IMU data and turning GNSS data corresponding to the vehicle turning process, and calculate the roll angle observation value of the IMU based on the turning IMU data and turning GNSS data; In the GNSS data and dynamic IMU data, identify the straight-moving IMU data and straight-moving GNSS data corresponding to the straight-moving process of the vehicle, and calculate the pitch angle observation value and yaw angle observation value of the IMU based on the straight-moving IMU data and straight-moving GNSS data; For each parameter to be calibrated, the corresponding observation value is input into the corresponding type of optimizer, and the predicted value of the parameter to be calibrated is obtained based on the optimizer, which is used as the calibration result of the parameter to be calibrated; the parameters to be calibrated include: roll angle, pitch angle and yaw angle, and the optimizers include: roll angle optimizer, pitch angle optimizer and yaw angle optimizer.

2. The method according to claim 1, characterized in that, The step of calculating the pitch angle and yaw angle observations of the IMU based on the straight-line IMU data and the straight-line GNSS data includes: Based on the straight-line IMU data and the straight-line GNSS data, estimate the world attitude of the IMU in the world coordinate system; Based on the world attitude, calculate the IMU velocity in the IMU coordinate system; Based on the rotational extrinsic parameters of the IMU and the vehicle, the IMU velocity is converted into the vehicle velocity of the IMU in the vehicle coordinate system; Based on the vehicle speed and according to non-integrity constraints, the pitch angle and yaw angle observations of the IMU are calculated.

3. The method according to claim 2, characterized in that, The step of estimating the world attitude of the IMU in the world coordinate system based on the straight-line IMU data and the straight-line GNSS data includes: The Kalman state estimation filter is predicted based on the straight-line IMU data, and the Kalman state estimation filter is updated by observation based on the straight-line GNSS data to obtain the world attitude of the IMU in the world coordinate system; the state vector of the Kalman state estimation filter includes at least the world attitude of the IMU in the world coordinate system.

4. The method according to claim 2, characterized in that, The step of calculating the pitch angle and yaw angle observations of the IMU based on the vehicle speed and according to non-integrity constraints includes: The pitch bias and yaw bias of the IMU are calculated using the following formulas: Among them, v x Let v be the component of the vehicle's velocity along the x-direction. y Let v be the component of the vehicle's velocity along the y-direction. z Let be the component of the vehicle's velocity along the z-direction.

5. The method according to any one of claims 1-4, characterized in that, The step of calculating the roll angle observation value of the IMU based on the turning IMU data and the turning GNSS data includes: Calculate the actual centripetal acceleration of the vehicle based on the turning GNSS data; The lateral acceleration of the vehicle is identified from the turning IMU data; The roll angle observation of the IMU is calculated based on the actual centripetal acceleration and the lateral acceleration.

6. The method according to claim 5, characterized in that, The step of calculating the roll angle observation of the IMU based on the actual centripetal acceleration and the lateral acceleration includes: The roll angle observation θ of the IMU is calculated using the following formula: Among them, a c Let a be the actual centripetal acceleration. y Let g be the lateral acceleration, and g be the gravitational acceleration.

7. The method according to any one of claims 1-4, characterized in that, The optimizer includes a roll angle Kalman filter, a pitch angle Kalman filter, and a yaw angle Kalman filter. During the prediction process, the state vector of each Kalman filter remains unchanged.

8. A vehicle-mounted IMU online calibration device, characterized in that, The IMU is installed in the vehicle, which is also equipped with at least a GNSS receiver. The device includes: The data acquisition module is used to acquire GNSS data and dynamic IMU data collected during the vehicle's driving process; the driving process includes at least a straight-line process and a turning process; The first observation module is used to identify the turning IMU data and turning GNSS data corresponding to the vehicle turning process in the GNSS data and dynamic IMU data, and to calculate the roll angle observation value of the IMU based on the turning IMU data and turning GNSS data. The second observation module is used to identify the straight-moving IMU data and straight-moving GNSS data corresponding to the straight-moving process of the vehicle in the GNSS data and dynamic IMU data, and to calculate the pitch angle observation value and yaw angle observation value of the IMU based on the straight-moving IMU data and straight-moving GNSS data. The parameter optimization module is used to input the corresponding observation values ​​into the corresponding type of optimizer for each parameter to be calibrated, and obtain the predicted value of the parameter to be calibrated based on the optimizer, which is used as the calibration result of the parameter to be calibrated; the parameters to be calibrated include: roll angle, pitch angle and yaw angle, and the optimizers include: roll angle optimizer, pitch angle optimizer and yaw angle optimizer.

9. A computer device, characterized in that, include: The memory and the processor are coupled; The memory is used to store one or more computer instructions; The processor is used to execute one or more computer instructions to implement the online calibration method for an on-board IMU as described in any one of claims 1 to 7.

10. A readable storage medium having stored thereon one or more computer instructions, characterized in that, The instruction is executed by the processor to implement the online calibration method for an on-board IMU as described in any one of claims 1 to 7.

Images