Vehicle-mounted IMU (Inertial Measurement Unit) data compensation method for intelligent driving inertial navigation

Abstract

The invention relates to the technical field of positioning navigation, in particular to a vehicle-mounted IMU (inertial measurement unit) data compensation method for intelligent driving inertial navigation, which comprises the following steps: acquiring various attitude data of a vehicle-mounted IMU through a rotating platform, acquiring a reference quantity, establishing a target function, and acquiring an offline calibrated rotating error matrix and a scale factor matrix by adopting an optimization algorithm; the method comprises the following steps: acquiring operation original measurement in a vehicle operation process, defining a system state vector, analyzing a Taylor expansion approximation relation between a cross multiplication matrix corresponding to an error small angle and a rotation error matrix, and constructing a discrete time state equation and an observation equation based on an IMU kinematics model; acquiring an error small angle after each update by adopting extended Kalman filtering; and carrying out real-time compensation on the vehicle-mounted IMU data in combination with the operation original measurement at each moment. The invention aims to improve the precision and robustness of inertial navigation and meet the requirement of intelligent driving for continuous and stable positioning.

Description

Technical Field

[0001] The present application relates to the field of positioning and navigation technology, and specifically to a vehicle-mounted IMU data compensation method for inertial navigation of intelligent driving. Background Art

[0002] The on-board inertial measurement unit (IMU) is a core sensor in a vehicle. Composed of an accelerometer and a gyroscope, it is used to measure the vehicle's three-dimensional acceleration and angular velocity in real time, thereby sensing the vehicle's state of motion. Intelligent driving inertial navigation technology uses data provided by the IMU to calculate the vehicle's real-time position, velocity, and attitude through inertial navigation algorithms. It can continue to provide accurate motion status even if the GNSS signal is lost. However, because the IMU is affected by hardware processing accuracy and installation errors, its raw measurement data may deviate from the actual motion state. These errors can quickly accumulate during the integration process, causing navigation results to drift. In severe cases, this can affect intelligent driving decisions. Therefore, error compensation technology is needed to eliminate these deviations and ensure the high accuracy and reliability of inertial navigation.

[0003] The physical installation of the three-axis accelerometer and gyroscope within the IMU is often not completely orthogonal. Furthermore, the entire device may be slightly rotated or tilted when mounted on the vehicle. This inter-axis non-orthogonality and installation error can cause the motion component of one axis to interfere with the measurement of other axes, thereby affecting the accuracy of attitude estimation and position integration. Traditional methods typically obtain fixed error parameters through offline calibration (such as the six-plane method). However, these methods fail to fully account for vehicle dynamics, causing the error parameters to drift over time or under complex road conditions. The mismatch between offline calibration and actual operating conditions ultimately makes it difficult for the compensation model to continuously track error changes. This significantly reduces the accuracy of inertial navigation positioning, especially during long-term operation or under complex road conditions, making it difficult to meet the continuous high-precision positioning requirements of intelligent driving systems. Summary of the Invention

[0004] In view of the above, it is necessary to provide an on-board IMU data compensation method for intelligent driving inertial navigation to solve the above problems.

[0005] One embodiment of the present application provides a vehicle-mounted IMU data compensation method for intelligent driving inertial navigation, the method comprising: The rotating platform is used to obtain each set of calibration raw measurements for various postures of the vehicle-mounted IMU, and the corresponding calibration reference quantities are obtained. Based on the difference between the calibration raw measurements and the corresponding reference quantities, combined with the inter-axis non-orthogonality of the vehicle-mounted IMU, an objective function is established, and an optimization algorithm is used to obtain the rotation error matrix and scale factor matrix for offline calibration. During vehicle operation, the original measurements of the onboard IMU at each moment are obtained, and the system state vector including the small error angle is defined. The Taylor expansion approximate relationship between the cross product matrix corresponding to the small error angle and the rotation error matrix is ​​analyzed. Based on the IMU kinematic model, the discrete-time state equation and observation equation are constructed to predict the system state vector at the next moment. Based on the difference between the actual measurement value and the predicted value, the extended Kalman filter is used to obtain the small error angle after each update. Based on the scale factor matrix obtained by offline calibration and the cross product matrix corresponding to the small error angle parameter after each update, combined with the original running measurements at each moment, the on-board IMU data is compensated in real time.

[0006] The objective function formula is specifically: ;Where, min represents the minimization function; represents the rotation error matrix; S represents the scale factor matrix; N represents the number of data sets collected for each posture; represents the i-th set of calibration raw measurements for each pose; Represents the calibration reference corresponding to the i-th group of calibration raw measurements for each posture.

[0007] The constraint condition of the objective function is that the rotation error matrix is ​​an orthogonal matrix.

[0008] The elements of the system state vector also include: a vehicle position vector, an instantaneous velocity vector, and an attitude angle vector.

[0009] The construction of the discrete-time state equation and the observation equation is specifically as follows: Discrete-time state equation: ,in, 、 They represent the system state vector and IMU operation raw measurement at time k respectively; represents the state transition function; represents the overall system noise; represents the system state vector at the predicted k+1 time; Observation equation: ,in, represents the observation vector at time k, represents the observation matrix, represents the observation noise, represents the system state vector at time k.

[0010] Among them, the update formula of the instantaneous velocity vector in the system state vector is: Where: Indicates vehicle posture The corresponding rotation matrix, represents the acceleration due to gravity, Indicates the sampling interval of IMU, It represents the acceleration after error compensation when IMU measures acceleration k. represents the instantaneous velocity vector of the vehicle at time k, Represents the predicted instantaneous velocity vector of the vehicle at time k+1.

[0011] Among them, the update process of the position vector in the system state vector is: obtain the product of the instantaneous velocity vector of the vehicle body at each moment and the IMU sampling interval, add it to the vehicle body position vector at each moment, and obtain the predicted vehicle body position vector at the next moment.

[0012] Among them, the update formula of the attitude vector in the system state vector is: ;in: It represents the angular velocity after error compensation when the IMU measures the angular velocity at time k; Represents the predicted vehicle posture vector at time k+1; Represents the vehicle posture vector at time k; Indicates vehicle posture The corresponding rotation matrix; Indicates the sampling interval of the IMU.

[0013] Among them, the updating process of the error small angle vector in the system state vector is: the sum of the Gaussian white noise and the error small angle vector at each moment is used as the predicted error small angle vector at the next moment.

[0014] The real-time compensation of the vehicle-mounted IMU data is specifically as follows: Calculate the sum of the cross product matrix corresponding to the small error angle obtained in each update and the identity matrix, and multiply the result with the scale factor matrix obtained by offline calibration; Multiplying the multiplication result by the angular velocity in the original measurement of the IMU operation to obtain the compensated angular velocity; The multiplication result is multiplied by the acceleration in the original measurement of the IMU operation to obtain the compensated acceleration.

[0015] This application has at least the following beneficial effects: This application obtains each set of calibration original measurements of various postures of the vehicle-mounted IMU through a rotating platform, and obtains the corresponding calibration reference quantities, which helps to provide calibration data and determine the relationship between the IMU measurement value and the reference quantity. The calibration original measurements of different postures provide data of various angles and directions, which are the basis for establishing models and calculating errors; based on the difference between the calibration original measurements and the corresponding reference quantities, combined with the inter-axis non-orthogonality of the vehicle-mounted IMU, an objective function is established, and an optimization algorithm is used to obtain the rotation error matrix and the scale factor matrix of the offline calibration, which helps to correct the system error of the IMU through the optimization algorithm, and obtain the rotation error matrix and the scale factor matrix. These matrices represent the non-ideal errors in the IMU measurement, and the optimization algorithm is used to obtain the rotation error matrix and the scale factor matrix. The matrix after quantization provides an accurate reference for subsequent error compensation. During vehicle operation, the original IMU measurement at each moment is obtained, the system state vector containing the small error angle is defined, and the Taylor expansion approximate relationship between the cross product matrix corresponding to the small error angle and the rotation error matrix is ​​analyzed. The beneficial effect is that by analyzing the small error angle and combining the Taylor expansion approximate relationship, the complex error calculation process is simplified, providing an effective theoretical framework for subsequent Kalman filtering and system state updates. Based on the IMU kinematic model, a discrete-time state equation and observation equation are constructed, and the IMU kinematic model is used to predict the system state, providing a preliminary estimate of the state at the next moment in the absence of real-time observation data. The prediction step provides the basis for the subsequent update step. Based on the difference between the actual measurement value and the predicted value, the extended Kalman filter is used to obtain the small error angle after each update. By processing the difference between the actual measurement value and the predicted value, the system state estimate can be corrected to obtain the small error angle after each update. EKF, combined with nonlinear system characteristics, can optimize state estimation in real time and improve accuracy; finally, based on the scale factor matrix obtained by offline calibration and the cross product matrix corresponding to the error small angle parameters after each update, combined with the original operation measurements at each moment, the on-board IMU data is compensated in real time. Its beneficial effect is that the error of the IMU is corrected to the minimum through real-time compensation, providing high-precision attitude estimation. This step combines the results of offline calibration and real-time updated data to dynamically calibrate the IMU measurement and ensure the accuracy of the compensated IMU data. This application solves the systematic static error through offline calibration, and uses online EKF self-calibration to track dynamic error drift in real time, taking into account calibration accuracy and environmental adaptability; the small angle parameterized design reduces the computational complexity and makes real-time estimation feasible; the full life cycle covers the error sources from factory to operation, effectively eliminates the influence of crosstalk between axes, significantly improves the accuracy and robustness of inertial navigation, and meets the needs of intelligent driving for continuous and stable positioning. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1Flowchart of the vehicle-mounted IMU data compensation method for intelligent driving inertial navigation provided in this application; Figure 2 Schematic diagram of obtaining the compensated vehicle-mounted IMU data provided in this application. DETAILED DESCRIPTION

[0017] In the description of the embodiments of this application, words such as "exemplary," "or," and "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of this application should not be construed as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary," "or," and "for example" is intended to present the relevant concepts in a concrete manner.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art in the art of this application. The terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application.

[0019] It should also be noted that the terms "first" and "second" in this application and its accompanying drawings are used to distinguish similar objects, rather than to describe a specific order or precedence. The methods disclosed in the embodiments of this application or the methods shown in the flowcharts include one or more steps for implementing the methods. Without departing from the scope of protection of this application, the order of execution of multiple steps can be interchanged with each other, and some steps can also be deleted.

[0020] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0021] This application proposes a vehicle-mounted IMU data compensation method for intelligent driving inertial navigation, which is applied in the field of positioning and navigation technology. Figure 1 , the method comprises the following steps: Step 1: Obtain each set of calibration raw measurements of the vehicle-mounted IMU in multiple postures through the rotating platform, and obtain the corresponding calibration reference quantities; Based on the difference between the calibration raw measurements and the corresponding reference quantities, combined with the inter-axis non-orthogonality of the vehicle-mounted IMU, establish the objective function, and use the optimization algorithm to obtain the offline calibration rotation error matrix and scale factor matrix.

[0022] During the production, processing, and assembly process, the physical axis systems of the three-axis accelerometers and gyroscopes within the IMU cannot be completely orthogonal due to mechanical precision limitations. Furthermore, when the entire unit is installed on the vehicle, there are fixed initial deviations. These systematic, static errors can cause crosstalk between the motion components of different axes in the measured values. Without pre-correction, this can directly amplify deviations in subsequent attitude estimation and position integration. Because these errors are stable and relatively fixed for a short period after the device leaves the factory, they are difficult to eliminate all at once through online dynamic calibration. Therefore, offline precision calibration is required before commissioning to lay the foundation for error compensation.

[0023] Fix the IMU to be calibrated on the 3D rotating platform with a rigid fixture to ensure that the relative position of the IMU and the platform coordinate system is fixed; connect the IMU data acquisition module to the rotating platform control system to achieve synchronous recording of the original measurement and the platform posture.

[0024] Design at least 30 reference poses covering the entire pose space, including: Static posture: For example, horizontal placement, pitch angle range from arrive , Roll angle range from arrive , yaw angle range from arrive These attitudes should ensure that the accelerometer can detect the projection of gravity on the different axis systems.

[0025] Dynamic posture: Rotate the platform to arrive The gyroscope can detect the known angular velocity by rotating around the x / y / z axis at a uniform speed.

[0026] Control the rotating platform to switch to the preset postures in sequence, collect data after each posture is stable for 5 seconds, and record the original measurement of the IMU calibration in the i-th group in each posture , including: the accelerometer outputs the i-th group of acceleration measurement vectors for each posture , the gyroscope outputs the i-th group of angular velocity measurement vectors for each attitude , recorded as ,in Represents transposition. Synchronously record the calibration reference corresponding to the i-th group of calibration raw measurements of each posture output by the rotating platform , including: Under static posture, the i-th group of acceleration reference vectors for each posture Determined by the projection of the gravity vector in each direction of the platform coordinate system; during dynamic rotation, the i-th group of acceleration reference vectors for each posture It is determined by the rotation angular velocity set in each direction of the platform coordinate system, which is recorded as , it should be noted that the calibration of the original measurement , calibration reference Both are 3×2 matrices.

[0027] All calibration raw measurements of each group of postures are filtered using a time window mean. In this application, the window length is 100ms to eliminate high-frequency noise, obtain smoothed calibration raw measurements, and remove deviations from the reference value. The above outliers ensure the validity of the data used for calibration.

[0028] Based on the principle that the original calibration measurement should be close to the reference value after the error matrix transformation, this application solves the error matrix through constrained optimization. Since the original IMU calibration measurement is affected by the non-orthogonality between axes, that is, the rotation error and the scale factor error, the correction relationship is: ,in, Represents the calibration reference of each attitude group i, S represents the scale factor matrix, which is used to describe the scale factor error of the IMU sensor (accelerometer and gyroscope), that is, the proportional deviation between the original output value of the sensor and the real physical quantity. The scale factor matrix is ​​a 3×3 matrix; Represents the rotation error matrix, which is used to describe the non-orthogonality and installation deviation between axes, satisfying the constraints (Orthogonality), because the rotation error matrix must ensure that the length and angle of the vector remain unchanged during the transformation process, the rotation error matrix is ​​also a 3×3 matrix.

[0029] The optimization goal is to achieve optimal calibration by minimizing the sum of squares of the deviations between the corrected measured values ​​and the reference values. That is, the objective function is ; ; This application uses the constrained Levenberg-Marquardt algorithm to iteratively solve, initializing (ideally orthogonal), (ideal ratio); in each iteration, the rotation error matrix is ​​parameterized into axis-angle form (3 degrees of freedom), combined with the 3 diagonal elements of the scale factor matrix, a total of 6 parameters to be estimated; the objective function is minimized by gradient descent, while forcing the rotation error matrix to satisfy the orthogonality constraint; after convergence, the rotation error matrix is ​​obtained and the scale factor matrix .

[0030] Will and The initial compensation matrix is ​​written into the IMU preprocessing module and serves as the initial compensation matrix. This matrix is ​​used for hardware mapping correction before the equipment is first commissioned. It eliminates most systematic static errors at the source and initially maps the raw IMU calibration measurements to the ideal axis system. This provides a low-error basis for subsequent online dynamic calibration, preventing excessive initial errors from causing difficulty in online algorithm convergence or reduced accuracy.

[0031] Step 2: During vehicle operation, obtain the original operation measurements of the on-board IMU at each moment, define the system state vector including the small error angle, analyze the Taylor expansion approximate relationship between the cross product matrix corresponding to the small error angle and the rotation error matrix, and construct the discrete-time state equation and observation equation based on the IMU kinematic model to predict the system state vector at the next moment; based on the difference between the actual measurement value and the predicted value, use the extended Kalman filter to obtain the small error angle after each update.

[0032] While offline calibration can determine the initial compensation matrix, in actual vehicle operation, environmental factors such as temperature fluctuations, continuous vibration, and minor collisions can cause IMU axis non-orthogonality and installation errors to drift slowly over time. These dynamic changes cannot be fully accounted for by static laboratory calibration. If only initial compensation is relied upon, the accumulated errors over long periods of operation can significantly reduce navigation accuracy. Therefore, online methods are needed to track and correct these dynamic errors in real time to adapt to the error variations under complex operating conditions.

[0033] The coupling relationship between IMU measurement value and attitude and installation error is nonlinear. Extended Kalman Filter (EKF) can realize recursive estimation of state by linearizing and approximating the nonlinear system. At the same time, it can integrate multi-source observation information such as GNSS and odometer as constraints to effectively suppress error accumulation. Compared with traditional EKF, this application uses the error small angle parameter By incorporating it into the system state vector as an extended state, it achieves joint dynamic estimation of attitude, position, velocity and installation error parameters, rather than just estimating conventional motion states. This enables real-time tracking of error changes caused by factors such as temperature and vibration, and provides online self-calibration capabilities. Traditional EKFs usually do not use such error parameters as state quantities, making it difficult to adapt to dynamic drift of errors.

[0034] Specifically, the system state vector is defined as , where x represents the system state vector; Represents the position vector of the vehicle body (three-dimensional), represents the instantaneous velocity vector of the vehicle body (three-dimensional), Represents the vehicle's attitude angle vector (three-dimensional), Represents the small-angle error vector (three-dimensional), which is used to correspond to small-angle disturbances of non-orthogonality between axes and installation deviation.

[0035] When performing preliminary compensation on the original measurement values ​​of the IMU operation, a small angle error parameter is introduced Using the principle of small angle rotation approximation, that is, when In radians, the rotation error matrix can be approximated by Taylor expansion as , where represents a unit vector, express The cross product matrix of , which means that the vector cross product operation is converted into matrix multiplication, ;in, 、 、 They represent the components of the small error angle on the x-axis, y-axis, and z-axis in the IMU coordinate system respectively.

[0036] Based on the IMU kinematic model and the dynamic characteristics of the error parameters, the discrete-time state equation is established: ,in, represents the system state vector at the predicted k+1 time; 、 They represent the system state vector at time k and the original measurement of the operation. In the initial system state vector, the position is set to the coordinate output by GNSS, the velocity is set to the 0 vector, the attitude angle is solved by the gravity vector measured by the accelerometer in the static state to calculate the pitch angle and roll angle, the yaw angle is set to 0, and the initial value of the small angle is obtained by extracting the installation rotation error matrix of the offline calibration The antisymmetric part of is obtained, and the subsequent state vector is updated through Kalman filtering; Represents the overall system noise, which obeys a zero-mean Gaussian distribution, and its covariance matrix is ​​preset to the variance of the inherent noise in the IMU; Represents the state transfer function, which describes the evolution of the state over time. The core is derived based on the corrected IMU measurements.

[0037] For the update of each element of the system state vector, the speed The update satisfies: ,in: Indicates vehicle posture The corresponding rotation matrix can be used to convert the IMU measurement value from the IMU coordinate system to the navigation coordinate system through Euler angle or quaternion conversion. represents the acceleration due to gravity, Indicates the sampling interval of IMU, It represents the acceleration after error compensation when IMU measures acceleration k. represents the instantaneous velocity vector of the vehicle at time k, The predicted instantaneous velocity vector of the vehicle at time k+1 is obtained by integrating the effective acceleration within the sampling interval, and finally the velocity is updated from the current velocity to the next moment.

[0038] Location and posture The update is also based on the velocity integral and angular velocity integral derivation, the formulas are: 、 ,in: represents the vehicle position vector at time k; Represents the predicted vehicle position vector at time k+1; It represents the angular velocity after error compensation when the IMU measures the angular velocity at time k; Represents the predicted vehicle posture vector at time k+1; Represents the vehicle posture vector at time k.

[0039] Error parameters The update of needs to consider the slow drift of error, which is approximately a random walk process. ,in Represents Gaussian white noise, with a variance preset to , reflecting the slowness of error drift, making The estimation can track the real slow changes without being disturbed by high-frequency noise; Represents the small angle vector of error at time k; Represents the small angle vector of the predicted error at time k+1.

[0040] The state estimation is constrained by using the observation values ​​provided by the external sensor. In this embodiment, the speed measured by GNSS is selected As the observed value, the observation equation is ,in represents the observation vector at time k, that is ; The implementer can also choose any other observation value for analysis; Represents the observation matrix, extracts the velocity component in the system state vector and associates it with the observation value, that is, , corresponding to the three-dimensional position, velocity, attitude angle and error respectively, and 0 means no correlation. represents the observation noise, which follows a zero-mean Gaussian distribution and whose covariance matrix is ​​preset to the variance of the inherent noise in the GNSS sensor.

[0041] Then, the EKF iteration is used to predict and update the error, and the state equation is used to predict the The prior state at the moment, the prior covariance matrix is ​​calculated using the Jacobian matrix; in the update phase, the Kalman filter is first calculated, and then the observation value is used Update the state, obtain the posterior estimate, and finally update the covariance matrix. The formula of EKF iteration belongs to the well-known technology and the specific content will not be repeated here.

[0042] Step 3: Based on the scale factor matrix obtained by offline calibration and the cross product matrix corresponding to the small error angle parameter after each update, combined with the original running measurements at each moment, the on-board IMU data is compensated in real time.

[0043] After each EKF iteration, the updated error small angle parameter is extracted from the posterior state. , as the input of the next round of IMU measurement correction, through continuous iteration, the error parameters will gradually converge to the true error value, realizing dynamic tracking and calibration of the installation error, incorporating the installation error parameters into the real-time estimation closed loop, and finally realizing adaptive correction of the error with the vehicle's operating status.

[0044] By using the EKF extended state to estimate dynamically changing small-angle error parameters in real time, the offline initial compensation can be dynamically corrected to ensure that the IMU measurement values ​​are always consistent with the real physical motion throughout the vehicle's life cycle (including environmental disturbances, hardware aging and other scenarios). Ultimately, through the online convergence of the error parameters, high-precision IMU correction data is provided for subsequent attitude solution and position integration, reducing navigation drift caused by dynamic errors and improving the long-term stability and accuracy of the vehicle navigation system under complex working conditions.

[0045] After offline calibration and online EKF dynamic estimation, the parameters required for error compensation have been obtained: the scale factor matrix and the real-time small-angle error vector. However, if these parameters are not applied to the IMU raw data in real time, inter-axis non-orthogonality and installation errors will continue to cause measurement cross-axis, resulting in deviations in attitude and position solutions. Furthermore, when the IMU operates alone, errors accumulate over time, requiring the integration of observations from external sensors to constrain drift. Therefore, real-time compensation must be used to correct the raw data and integrate external information to achieve the error suppression effect in the actual navigation output.

[0046] Furthermore, the scale factor matrix obtained by offline calibration makes the original measurement match the real physical quantity. When the modulus of the error angle is much less than 1 radian, the rotation error matrix is ​​approximately , the running raw measurement is rotated and mapped from the actual axis system with deviation to the ideal orthogonal axis system, eliminating the crosstalk between axes, so that the measurement value of each axis only reflects the real motion component in the corresponding direction.

[0047] The acceleration in the raw measurement of the calibrated IMU run is: , the angular velocity is: ,in, represents the scale factor matrix obtained by offline calibration, represents the cross product matrix of small angle errors, represents the acceleration measurement vector at time i; Represents the angular velocity measurement vector at time i.

[0048] The corrected IMU data is input into the main navigation filter, and external observation information is introduced as a constraint. In this embodiment, the speed output of GNSS is used. ,During the fusion process, the main navigation filter suppresses the error accumulation when ,the IMU works alone through state prediction and observation ,update.

[0049] After fusion is completed, the filter outputs real-time high-precision navigation parameters: vehicle position ,speed and posture , as the positioning and attitude reference of the intelligent driving system. Among them, the schematic diagram of obtaining the compensated vehicle IMU data is as follows Figure 2 shown.

[0050] The core purpose of real-time compensation is to apply the error parameters of offline calibration and online dynamic estimation to the IMU raw data in real time, eliminate the measurement deviation caused by inter-axis non-orthogonality and installation errors, and ensure that the IMU data input to the navigation filter accurately reflects the motion state of the vehicle; the fusion output solves the error accumulation problem when the IMU works alone through the observation constraints of external sensors, and ultimately provides continuous, stable, and high-precision attitude and position information for the inertial navigation of the intelligent driving system, meeting the vehicle's stringent requirements for positioning accuracy and robustness under complex road conditions.

[0051] The flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the systems, methods and computer program products according to the embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of the code, and the part of the module, program segment or code contains one or more executable instructions for realizing the specified logical function. In some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, which can depend on the functions involved. In the description corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different boxes can also occur in an order different from that disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, which can depend on the functions involved. Each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented by a dedicated hardware-based system that performs the specified function or action, or may be implemented by a combination of dedicated hardware and computer instructions.

[0052] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. A vehicle-mounted IMU data compensation method for intelligent driving inertial navigation, characterized in that: The method comprises the following steps: The rotating platform is used to obtain each set of calibration raw measurements for various postures of the vehicle-mounted IMU, and the corresponding calibration reference quantities are obtained. Based on the difference between the calibration raw measurements and the corresponding reference quantities, combined with the inter-axis non-orthogonality of the vehicle-mounted IMU, an objective function is established, and an optimization algorithm is used to obtain the rotation error matrix and scale factor matrix for offline calibration. During vehicle operation, the original measurements of the onboard IMU at each moment are obtained, and the system state vector including the small error angle is defined. The Taylor expansion approximate relationship between the cross product matrix corresponding to the small error angle and the rotation error matrix is ​​analyzed. Based on the IMU kinematic model, the discrete-time state equation and observation equation are constructed to predict the system state vector at the next moment. Based on the difference between the actual measurement value and the predicted value, the extended Kalman filter is used to obtain the small error angle after each update. Based on the scale factor matrix obtained by offline calibration and the cross product matrix corresponding to the small error angle parameters after each update, combined with the original running measurements at each moment, the on-board IMU data is compensated in real time.

2. The vehicle-mounted IMU data compensation method for intelligent driving inertial navigation according to claim 1, characterized in that: The formula of the objective function is specifically: ;Where, min represents the minimization function; represents the rotation error matrix; S represents the scale factor matrix; N represents the number of data sets collected for each posture; represents the i-th set of calibration raw measurements for each pose; Represents the calibration reference corresponding to the i-th group of calibration raw measurements for each posture.

3. The vehicle-mounted IMU data compensation method for intelligent driving inertial navigation according to claim 2, characterized in that: The constraint condition of the objective function is that the rotation error matrix is ​​an orthogonal matrix.

4. The vehicle-mounted IMU data compensation method for intelligent driving inertial navigation according to claim 1, characterized in that: The elements of the system state vector also include: a vehicle position vector, an instantaneous velocity vector, and an attitude angle vector.

5. The vehicle-mounted IMU data compensation method for intelligent driving inertial navigation according to claim 1, characterized in that: The construction of the discrete-time state equation and observation equation is specifically as follows: Discrete-time state equation: ,in, 、 They represent the system state vector and IMU operation raw measurement at time k respectively; represents the state transition function; represents the overall system noise; represents the system state vector at the predicted k+1 time; Observation equation: ,in, represents the observation vector at time k, represents the observation matrix, represents the observation noise, represents the system state vector at time k.

6. The vehicle-mounted IMU data compensation method for intelligent driving inertial navigation according to claim 4, characterized in that: In the system state vector, the update formula of the instantaneous velocity vector is: Where: Indicates vehicle posture The corresponding rotation matrix, represents the acceleration due to gravity, Indicates the sampling interval of IMU, It represents the acceleration after error compensation when IMU measures acceleration k. represents the instantaneous velocity vector of the vehicle at time k, Represents the predicted instantaneous velocity vector of the vehicle at time k+1.

7. The vehicle-mounted IMU data compensation method for intelligent driving inertial navigation according to claim 4, characterized in that: In the system state vector, the update process of the position vector is: obtain the product of the instantaneous velocity vector of the vehicle at each moment and the IMU sampling interval, add it to the vehicle position vector at each moment, and obtain the predicted vehicle position vector at the next moment.

8. The vehicle-mounted IMU data compensation method for intelligent driving inertial navigation according to claim 4, characterized in that: In the system state vector, the update formula of the attitude vector is: ;in: It represents the angular velocity after error compensation when the IMU measures the angular velocity at time k; Represents the predicted vehicle posture vector at time k+1; Represents the vehicle posture vector at time k; Indicates vehicle posture The corresponding rotation matrix; Indicates the sampling interval of the IMU.

9. The vehicle-mounted IMU data compensation method for intelligent driving inertial navigation according to claim 1, characterized in that: In the system state vector, the updating process of the error small angle vector is as follows: the sum of the Gaussian white noise and the error small angle vector at each moment is used as the predicted error small angle vector at the next moment.

10. The vehicle-mounted IMU data compensation method for intelligent driving inertial navigation according to claim 1, characterized in that: The real-time compensation of the vehicle-mounted IMU data is specifically as follows: Calculate the sum of the cross product matrix corresponding to the small error angle obtained in each update and the identity matrix, and multiply it with the scale factor matrix obtained by offline calibration; Multiplying the multiplication result by the angular velocity in the original measurement of the IMU operation to obtain the compensated angular velocity; The multiplication result is multiplied by the acceleration in the original measurement of the IMU operation to obtain the compensated acceleration.

Images