A vehicle integrated navigation method, device, equipment and medium

Abstract

This application discloses a vehicle integrated navigation method, apparatus, device, and medium, relating to the field of intelligent driving navigation technology. The method includes: acquiring wheel speed data for the current epoch from a wheel speed measurement unit, and acquiring inertial measurement data for the current epoch from an inertial measurement unit; inputting a non-integrity constraint prediction model to obtain non-integrity constraint components for the current epoch; constructing three-dimensional motion constraint observations based on the non-integrity constraint components and wheel speed data; updating the kinematic observation equations of the Kalman filter through measurement; obtaining the updated state estimation results; and outputting the positioning results for the current epoch. This application dynamically predicts lateral and vertical velocity components using a non-integrity constraint prediction model, constructs three-dimensional motion constraint observations, and optimizes the state estimation of the Kalman filter. This effectively alleviates the problem of performance degradation in vehicle motion constraints under complex motion conditions, thereby improving the accuracy of integrated navigation.

Description

Technical Field

[0001] This application relates to the field of intelligent driving navigation technology, and in particular to a vehicle integrated navigation method, device, equipment and medium. Background Technology

[0002] The application of intelligent vehicles in environmental perception, obstacle avoidance, and path planning all require accurate pose estimation as a prerequisite. However, existing multi-sensor fusion methods struggle to meet the demands of continuous, high-precision state estimation in complex observation scenarios and under varying motion conditions. Although AI has brought new capabilities to perception and planning, its potential for improving pose estimation has not yet been fully explored.

[0003] Currently, traditional multi-sensor fusion navigation methods primarily rely on the absolute positioning capabilities of Global Navigation Satellite Systems (GNSS) and the relative positioning capabilities of systems such as Inertial Navigation Systems (INS) to achieve state estimation across all scenarios. For vehicles, the Controller Area Network Bus (CAN Bus) typically integrates a speed measurement interface, which can be used to monitor the vehicle's motion status in real time and can also be used as an odometer (ODO) to couple with GNSS and INS. Furthermore, based on rigid body kinematics, applying non-holonomic constraints (NHC) at the vehicle's drive wheels can create three-dimensional motion constraints within the vehicle's body frame (b-frame). The NHC constraint principle of these traditional kinematic constraints is simple and effective, and therefore widely used in vehicle navigation.

[0004] However, the combined navigation methods described above also have significant drawbacks. The calculated NHC constraints are based on strong lateral and vertical zero-velocity assumptions. In reality, due to complex road conditions and the vehicle's dynamic states such as steering and vibration, it's impossible to guarantee that the zero-velocity assumption in non-motion directions will always be satisfied. Therefore, applying incorrect motion constraints to the vehicle can lead to a degradation of the combined navigation system's constraint performance and even result in erroneous state estimations.

[0005] Therefore, there is currently a lack of a method that can predict and improve NHC constraints based on the vehicle's actual lateral and vertical velocities, in order to alleviate the problem of performance degradation of vehicle motion constraints under complex motion conditions and improve the accuracy of integrated navigation of the vehicle under complex motion conditions. Summary of the Invention

[0006] This application provides a vehicle integrated navigation method, apparatus, device, and medium to address the deficiencies of the aforementioned related technologies. The technical solution is as follows:

[0007] In a first aspect, embodiments of this application provide a vehicle integrated navigation method, including:

[0008] Obtain the wheel speed data of the current epoch output by the wheel speed measurement unit of the vehicle, and obtain the inertial measurement data of the current epoch output by the inertial measurement unit of the vehicle.

[0009] The inertial measurement data and the wheel speed data are input into the non-integrity constraint prediction model, and the non-integrity constraint components of the current epoch are calculated through the non-integrity constraint prediction model.

[0010] Three-dimensional motion constraint observations are constructed based on the non-integrity constraint components and the wheel speed data.

[0011] The kinematic observation equations of the Kalman filter are updated based on the three-dimensional motion constraint observations to obtain the state estimation results obtained from the Kalman filter update. Based on the state estimation results, the positioning results of the vehicle in the current epoch are output.

[0012] The non-integrity constraint prediction model is trained based on a given sample set.

[0013] In one alternative embodiment of the first aspect, the step of inputting the inertial measurement data and the wheel speed data into the non-integrity constraint prediction model, and calculating the non-integrity constraint components of the current epoch through the non-integrity constraint prediction model, includes:

[0014] The inertial measurement data of the current epoch output by the inertial measurement unit of the vehicle includes the angular velocity output by the three-axis gyroscope and the specific force output by the three-axis accelerometer.

[0015] The angular velocity output by the triaxial gyroscope, the specific force output by the triaxial accelerometer, and the wheel speed data of the current epoch are input into the non-integrity constraint prediction model so that the non-integrity constraint prediction model can predict the non-integrity constraint components of the current epoch based on the input data. The non-integrity constraint components include lateral velocity components and vertical velocity components.

[0016] The three-dimensional motion constraint observations are constructed based on the non-integrity constraint components and the wheel speed data, and the formula is applied:

[0017] ;

[0018] in, The three-dimensional motion constraint observation value, The wheel speed data, The lateral velocity component, The vertical velocity component is denoted by t, which represents the corresponding epoch, and the superscript b indicates that the coordinate system of the corresponding parameter is the carrier coordinate system.

[0019] In one alternative embodiment of the first aspect, before updating the kinematic observation equations of the Kalman filter based on the three-dimensional motion constraint observations, the method further includes:

[0020] The velocity of the center of the inertial measurement unit in the navigation coordinate system is calculated based on the inertial measurement data, and the velocity in the navigation coordinate system is reduced to the wheel speed measurement unit to obtain the reduced velocity in the carrier coordinate system.

[0021] The measurement update of the kinematic observation equations of the Kalman filter based on the three-dimensional motion constraint observations includes:

[0022] The prior residual of the kinematic observation equation is calculated by subtracting the three-dimensional motion constraint observation value from the reduced velocity.

[0023] The prior residual is used as the velocity information observation of the Kalman filter, and the state estimation result of the current epoch is obtained by updating the Kalman filter.

[0024] In one alternative embodiment of the first aspect, the training process of the non-integrity constraint prediction model includes the following steps:

[0025] The vehicle is driven for the typical convergence time of the GNSS solution strategy under the preset GNSS observation conditions, and then the vehicle is driven for the preset time under the preset GNSS observation conditions.

[0026] During the preset driving time, the inertial measurement data of each epoch output by the inertial measurement unit of the vehicle, the wheel speed data of the corresponding epoch output by the wheel speed measurement unit, and the GNSS positioning, speed measurement, and attitude measurement data of the corresponding epoch output by the GNSS signal receiving unit are acquired during the preset driving time.

[0027] A sample set is constructed based on the inertial measurement data of each epoch, the wheel speed data of the corresponding epoch, and the GNSS positioning, velocimetry, and attitude measurement data of the corresponding epoch to train the non-integrity constraint prediction model.

[0028] The non-integrity constraint prediction model is trained online based on the sample set, and the weight parameters of the converged non-integrity constraint prediction model are determined to obtain the trained non-integrity constraint prediction model.

[0029] In one alternative embodiment of the first aspect, before the vehicle travels under preset GNSS observation conditions for the typical convergence time of the GNSS solution strategy, the method further includes:

[0030] The linkage information from the inertial measurement unit to the drive wheel is obtained through calibration;

[0031] When the vehicle is stationary and meets the preset GNSS observation conditions, the initial GNSS positioning, velocity and attitude measurement data output by the GNSS signal receiving unit and the initial inertial measurement data output by the inertial measurement unit are acquired.

[0032] The navigation state of the vehicle is initialized based on the lever information, initial GNSS observations, and initial inertial measurement data, and the initial state estimation results of the Kalman filter are obtained, including the initial position information, initial velocity information, and initial attitude information of the vehicle.

[0033] In one alternative embodiment of the first aspect, the sample set for training the non-integrity constraint prediction model, constructed based on inertial measurement data of each epoch, wheel velocity data of the corresponding epoch, and GNSS positioning, velocimetry, and attitude measurement data of the corresponding epoch, includes:

[0034] Based on the GNSS positioning, velocity and attitude measurement data, the velocity of the vehicle in the corresponding epoch of the navigation coordinate system is calculated. The velocity in the navigation coordinate system is transformed to the vehicle coordinate system to obtain the velocity in the vehicle coordinate system. The lateral velocity component and the vertical velocity component of the velocity in the vehicle coordinate system are calculated.

[0035] The sample set is constructed by using wheel speed data, angular velocity, and specific force from the same epoch as input to the sample, and using the lateral velocity component and vertical velocity component from the same epoch as sample labels.

[0036] In one alternative embodiment of the first aspect, the preset GNSS observation conditions include:

[0037] The number of available GNSS satellites observed by the GNSS signal receiving unit of the vehicle is greater than a preset threshold, and the accuracy attenuation factor of the GNSS positioning result is less than a preset attenuation factor threshold, and the sum of the main diagonal elements of the position covariance matrix of the GNSS positioning result is less than a preset threshold.

[0038] After obtaining the trained incompleteness constraint prediction model, the method further includes:

[0039] During the vehicle's operation, under the preset GNSS observation conditions, wheel speed data, inertial measurement data, and GNSS positioning, velocity, and attitude measurement data for each epoch are acquired.

[0040] New training samples were constructed based on wheel speed data, inertial measurement data, and GNSS positioning, velocimetry, and attitude measurement data from the same epoch.

[0041] The new training samples are added to the sample set, and the non-integrity constraint prediction model is trained online using the supplemented sample set to update the weight parameters of the non-integrity constraint prediction model, thereby obtaining the updated non-integrity constraint prediction model.

[0042] The step of calculating the non-integrity constraint components of the current epoch based on the updated non-integrity constraint prediction model is performed.

[0043] Secondly, embodiments of this application also provide a vehicle navigation system, comprising:

[0044] The data acquisition unit is used to acquire the wheel speed data of the current epoch output by the wheel speed measurement unit of the vehicle, and to acquire the inertial measurement data of the current epoch output by the inertial measurement unit of the vehicle.

[0045] The constraint calculation unit is used to input the inertial measurement data and the wheel speed data into the non-integrity constraint prediction model, and calculate the non-integrity constraint components of the current epoch through the non-integrity constraint prediction model.

[0046] The constraint calculation unit is also used to construct three-dimensional motion constraint observations based on the non-integrity constraint components and the wheel speed data;

[0047] The positioning result calculation unit is used to measure and update the kinematic observation equation of the Kalman filter based on the three-dimensional motion constraint observation values, obtain the state estimation result obtained by the Kalman filter update, and output the positioning result of the vehicle in the current epoch based on the state estimation result.

[0048] The non-integrity constraint prediction model is trained based on a given sample set.

[0049] Thirdly, embodiments of this application also provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method provided by the first aspect or any implementation thereof of the embodiments of this application.

[0050] Fourthly, this application also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method provided by the first aspect of the embodiments of this application or any implementation thereof.

[0051] The beneficial effects of the technical solutions provided in some embodiments of this application include at least the following:

[0052] This application provides a vehicle integrated navigation method, device, electronic device, and storage medium. By dynamically predicting lateral and vertical velocity components through a non-integrity constraint prediction model, constructing three-dimensional motion constraint observations, and optimizing the state estimation of the Kalman filter, it can effectively alleviate the problem of performance degradation of vehicle motion constraints under complex motion conditions. It avoids the shortcomings of related technologies where the calculated NHC constraints are based on strong lateral and vertical zero-velocity assumptions and cannot reflect the true motion constraints of the vehicle under complex motion conditions, thereby improving the accuracy of integrated navigation. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in this application or related technologies, the drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0054] Figure 1 This is a schematic flowchart of a vehicle integrated navigation method provided in an embodiment of this application;

[0055] Figure 2 This is a schematic diagram of the structure of a vehicle-mounted navigation device provided in an embodiment of this application;

[0056] Figure 3 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0058] The terms "comprising" and "having," and any variations thereof, in the specification, claims, and accompanying drawings of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or modules is not limited to the steps or modules listed, but may optionally include steps or modules not listed, or may optionally include other steps or modules inherent to such process, method, product, or apparatus.

[0059] It should be noted that the terms "first" and "second" used in this application are merely to distinguish similar objects and do not represent a specific ordering of the objects. It is understood that "first" and "second" can be interchanged in a specific order or sequence where permitted. It should be understood that the objects distinguished by "first" and "second" can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in an order other than those described or illustrated herein.

[0060] It should be noted that the vehicle can include various types of vehicles, such as cars, buses, and trucks. Generally, these vehicles are equipped with an INS system and a GNSS signal receiving unit. GNSS provides accurate position and velocity information, but the signal may be affected by interference or unavailable in certain environments (such as tunnels, indoors, or urban areas with tall buildings). The INS system, on the other hand, uses accelerometers and gyroscopes in the Inertial Measurement Unit (IMU) to calculate the vehicle's position and orientation changes. Although its errors accumulate over time, it can provide continuous navigation information when GNSS signals are lost. Combining the two systems leverages their respective advantages to provide a more stable and reliable navigation service, achieving integrated navigation.

[0061] However, the NHC constraints calculated by related technologies in integrated navigation are based on strong lateral and vertical zero-velocity assumptions. They do not take into account the complex road conditions in reality, such as the vehicle's steering and vibration during driving. They cannot guarantee that the zero-velocity assumption in the non-movement direction will always be met. In other words, the related technologies do not take into account the vehicle's actual lateral and vertical velocities under complex motion conditions, which can easily lead to the accumulation of errors and thus inaccurate navigation and positioning results.

[0062] Based on this, the inventors discovered a fundamental contradiction between the traditional zero-velocity assumption and the actual motion state. By analyzing the correlation of multi-source sensor data, they found that the angular velocity and specific force information output by the inertial measurement unit can reflect the changing patterns of the vehicle's motion mode. Considering the fitting ability of machine learning models to nonlinear relationships, they proposed using historical GNSS positioning, velocity, and attitude measurement data to construct a supervisory signal (as sample labels) to train a prediction model for non-holonomic constraint components. This model, by sensing the vehicle's motion state in real time and dynamically adjusting the parameters of the constraint equations, can establish a three-dimensional observation model that more closely conforms to actual kinematics, providing technical support for precise integrated navigation of vehicles in complex environments and under complex motion conditions.

[0063] The present application will now be described in detail with reference to specific embodiments.

[0064] Next, combine Figure 1 This application describes a vehicle-integrated navigation method, apparatus, device, and medium provided by embodiments of this application. For details, please refer to... Figure 1 , Figure 1 A schematic flowchart of a vehicle integrated navigation method provided in an embodiment of this application is shown. Figure 1 As shown, the method includes the following steps:

[0065] S101, Obtain the wheel speed data of the current epoch output by the wheel speed measurement unit of the vehicle, and obtain the inertial measurement data of the current epoch output by the inertial measurement unit of the vehicle.

[0066] S102, input the inertial measurement data and the wheel speed data into the non-integrity constraint prediction model, and calculate the non-integrity constraint components of the current epoch through the non-integrity constraint prediction model;

[0067] S103, Based on the non-integrity constraint components and the wheel speed data, three-dimensional motion constraint observations are constructed;

[0068] S104, based on the three-dimensional motion constraint observations, the kinematic observation equations of the Kalman filter are measured and updated to obtain the state estimation results obtained from the Kalman filter update, and the positioning results of the vehicle in the current epoch are output based on the state estimation results.

[0069] Specifically, in S101, the inertial measurement unit outputs inertial measurement data including specific force and angular velocity. The inertial measurement unit typically measures the specific force using an onboard accelerometer, which represents the acceleration relative to the inertial reference frame. The inertial measurement unit typically measures the angular velocity using a gyroscope. The wheel speed measurement unit outputs wheel speed data representing the speed of the vehicle's drive wheels.

[0070] It should be noted that the measurement data output by the inertial measurement unit is relative to the body frame (b-frame), and the specific force at the corresponding epoch t is denoted as... The angular velocity corresponding to epoch t is .

[0071] Specifically, the wheel speed measurement unit may include an odometer (OD) mounted on the vehicle's drive wheels or non-steering wheels, through which the wheel speed in the vehicle's coordinate system b can be obtained. Wheel speed can also be obtained through the wheel speed acquisition interface on the CAN bus.

[0072] In some embodiments, S102, the inertial measurement data and the wheel speed data are input into the non-integrity constraint prediction model, and the non-integrity constraint components of the current epoch are calculated through the non-integrity constraint prediction model, specifically including:

[0073] The inertial measurement unit of the vehicle outputs inertial measurement data for the current epoch t, ​​including the angular velocity output by the three-axis gyroscope. The specific force output of the triaxial accelerometer .

[0074] Specifically, the angular velocity output by the three-axis gyroscope can be... The specific force output by the triaxial accelerometer and the wheel speed data of the current epoch. Constructing the vector Then, it is used as input to the non-integrity constraint prediction model.

[0075] Furthermore, the incompleteness constraint prediction model is based on the input data. The prediction yields the non-integrity constraint components for the current epoch, specifically including the lateral velocity components. and vertical velocity components .

[0076] Specifically, in S103, based on the lateral velocity component and vertical velocity components Wheel speed data The three-dimensional motion-constrained observations are constructed, and the specific application formula is as follows:

[0077] ;

[0078] in, These are three-dimensional motion-constrained observations. For wheel speed data, The lateral velocity component, The vertical velocity component is represented by t, which indicates the corresponding epoch. The superscript b indicates that the coordinate system of the corresponding parameter is the carrier coordinate system.

[0079] In some embodiments, before performing the step of measuring and updating the kinematic observation equations of the Kalman filter based on the three-dimensional motion constraint observations in S104, the following steps may be performed:

[0080] The velocity of the center of the inertial measurement unit in the navigation coordinate system n is calculated based on the inertial measurement data. and the center velocity The reduced speed in the carrier coordinate system is obtained by referring to the wheel speed measurement unit. .

[0081] Specifically, the earth-centered earth-fixed frame (e-frame) can be chosen as the navigation reference coordinate system (n-frame). This can be understood as follows: when the e-frame is chosen as the navigation coordinate system, the velocity of the inertial measurement unit center in the n-frame is expressed as... .

[0082] For example, taking the e-frame as the navigation coordinate system n-frame, the steps for calculating the velocity of the inertial measurement unit center in the navigation coordinate system are as follows:

[0083] Mechanical arrangement is performed based on inertial measurement data, and the pose information of the inertial measurement unit is output, specifically according to the comparison force. and angular velocity For mechanical choreography, the error state differential equation can be expressed as:

[0084] ;

[0085] The pose information obtained from the inertial measurement unit includes: position ,speed ,attitude .

[0086] Furthermore, based on the relationship between the linear velocities of the center of the circle and points on the circle during circular motion, the velocity of the center of the inertial measurement unit can be determined. The reduced speed in the vehicle coordinate system is obtained by referring to the wheel speed measurement unit. Apply the formula:

[0087] ;

[0088] in, This refers to the position information of the inertial measurement unit in the geocentric coordinate system. For location information The differential, Differential representing location information The error, This refers to the velocity information of the inertial measurement unit in the geocentric coordinate system. Represents speed information The error, For speed information The differential, Differential representing velocity information The error, This refers to the attitude error of the inertial measurement unit. To address attitude error The differential, This is the Earth's rotational angular velocity. Let be the rotation matrix from the carrier coordinate system to the geocentric Earth-fixed coordinate system. It is the acceleration due to gravity. The error is due to gravitational acceleration. The velocity information is in the carrier coordinate system. The specific force in the carrier coordinate system. The error is due to the specific force. The angular velocity in the carrier coordinate system The error is the angular velocity in the carrier coordinate system. Let T be the rotation matrix from the carrier coordinate system to the vehicle coordinate system, and let T denote the transpose matrix. This refers to the lever arm information from the inertial measurement unit to the drive wheel.

[0089] Specifically, the linkage information from the inertial measurement unit to the drive wheels can be measured before the vehicle moves. Alternatively, the lever arm information can be found in the product manual or other instruction documents of the vehicle. This application does not limit the method of obtaining lever arm information and installation angle error.

[0090] Specifically, in S104, the kinematic observation equations of the Kalman filter are updated based on the three-dimensional motion constraint observations, including:

[0091] Three-dimensional motion constraint observations With calculation speed The prior residuals of the kinematic observation equations are calculated by taking the difference.

[0092] Specifically, based on three-dimensional motion constraint observations The observation equations calculated for updating ODO / NHC measurements can be expressed as follows:

[0093] ;

[0094] in, The prior residuals of the above motion-constrained observation equations are... for n Tie b The rotation matrix of the system, for n Speed ​​error under system, for n The attitude error under the system, The pole arm vector is calibrated offline. This refers to the zero bias error of the inertial measurement unit gyroscope. For motion-constrained observation noise, Represents the antisymmetric matrix transformation operation of vectors.

[0095] Specifically, the distance from the IMU center to the GNSS antenna phase center (Antenna Reference Point, ARP) can be measured. b Projection under the system And the lever arm from the IMU center to the drive wheel in b Projection under the system This allows for offline calibration.

[0096] After calculating the prior residual for the current epoch, the prior residual can be used as the velocity information observation of the Kalman filter. The state estimation result for the current epoch is then obtained by updating the Kalman filter, specifically including:

[0097] Let the observation equation be:

[0098] ;

[0099] in, , Specifically, the observation vector in the Kalman filter at time k. , To include positional errors in the n-system Speed ​​error Attitude error and gyroscope zero bias accelerometer zero bias The state variables included.

[0100] The coefficient matrix corresponding to each state variable can be obtained from the ODO / NHC observation equation. Those skilled in the art know the specific form of the ODO / NHC observation equation, and this application does not limit it.

[0101] For observation noise, the corresponding covariance matrix is: Let the time update result before the measurement update be denoted as . The corresponding covariance is The measurement update process can then be represented as:

[0102] ;

[0103] ;

[0104] ;

[0105] Calculate the above intermediate variables Then, update the filter variables:

[0106] ;

[0107] ;

[0108] in, The state estimation results are updated based on ODO / NHC measurements. The corresponding covariances together constitute the navigation output of the Kalman filter, which can then output the positioning result of the vehicle in the current epoch based on the state estimation result.

[0109] In some embodiments, the training process of the non-integrity constraint prediction model in S102 includes the following steps:

[0110] First, perform static navigation initialization on the vehicle, specifically including:

[0111] Obtain the linkage information from the inertial measurement unit to the drive wheel obtained from the calibration. The specific calibration process is described in the foregoing embodiments and will not be repeated here.

[0112] When the vehicle is stationary and meets the preset GNSS observation conditions, the initial GNSS positioning, velocity and attitude measurement data output by the GNSS signal receiving unit and the initial inertial measurement data output by the inertial measurement unit are acquired.

[0113] The navigation state of the vehicle is initialized based on the lever information, initial GNSS observations, and initial inertial measurement data. The initial state estimation results of the Kalman filter are then obtained, including the initial position information of the vehicle. Initial velocity information and initial attitude information .

[0114] It's important to note that the Kalman filter is a recursive algorithm that relies on the state estimate from the previous time step to infer the current state. Therefore, before the actual calculation, it's helpful to provide an initial state estimate under static conditions. This ensures the recursive process runs correctly, accelerates the filter's convergence speed, improves overall performance, and prevents the filter from diverging due to excessive initial errors.

[0115] The specific training process includes the following steps:

[0116] S1021, the vehicle is driven for the typical convergence time of the GNSS solution strategy under the preset GNSS observation conditions, and then the vehicle is driven for the preset time under the preset GNSS observation conditions.

[0117] Specifically, preset GNSS observation conditions may include the following conditions:

[0118] The number of available GNSS satellites observed by the GNSS signal receiving unit of the vehicle is greater than a preset threshold, and the accuracy attenuation factor of the GNSS positioning result is less than a preset attenuation factor threshold, and the sum of the main diagonal elements of the position covariance matrix of the GNSS positioning result is less than a preset threshold.

[0119] The number of available GNSS satellites observed by the GNSS signal receiving unit is specifically the number of satellites (NSAT). Generally, at least four satellites are required to achieve three-dimensional positioning (longitude, latitude, and altitude). This application embodiment does not limit the value of the preset number threshold, and can set it considering different accuracy requirements and observation conditions.

[0120] It should be noted that the typical convergence time of the GNSS solution strategy is... This refers to the time required from when the receiver starts receiving satellite signals until the positioning calculation results reach the required level of accuracy.

[0121] Specifically, the preset duration depends on the size of the sample set used for training, and can be denoted as s.

[0122] S1022, during the preset driving time, acquire the inertial measurement data of each epoch output by the inertial measurement unit of the vehicle, the wheel speed data of the corresponding epoch output by the wheel speed measurement unit, and the GNSS positioning, speed measurement, and attitude measurement data of the corresponding epoch output by the GNSS signal receiving unit during the preset driving time.

[0123] Specifically, the preset driving duration corresponds to a time interval. ,in, The termination time of the typical convergence time for the GNSS solution strategy is given. It is possible to calculate the values ​​of each epoch t based on GNSS positioning, velocity, and attitude measurement data. b Vehicle speed Obtain the epochal sequence t. b The angular velocity output by the IMU three-axis gyroscope corresponding to the vehicle's speed The specific force output of the triaxial accelerometer Wheel speed output by the wheel speed measurement unit .

[0124] S1023, Construct a sample set for training the non-integrity constraint prediction model based on the inertial measurement data of each epoch, the wheel speed data of the corresponding epoch, and the GNSS positioning, velocimetry, and attitude measurement data of the corresponding epoch.

[0125] Specifically, the process of constructing the sample set includes:

[0126] Based on the GNSS positioning, velocity, and attitude measurement data, the velocity of the vehicle in the corresponding epoch of the navigation coordinate system is calculated. This velocity is then transformed to the vehicle coordinate system to obtain the velocity in the vehicle coordinate system. The lateral and vertical velocity components of the vehicle coordinate system velocity are then calculated, specifically including:

[0127] Obtain the GNSS measurement update post-hoc solution velocity ,attitude Calculated direction cosine matrix The calculated velocity of the b-system was obtained after verification. Apply the formula:

[0128] .

[0129] based on The lateral velocity components were calculated. With vertical velocity component The wheel speed data of the same epoch angular velocity Comparison As input to the sample, the lateral velocity components of the same epoch are used. With vertical velocity component The sample labels are used to construct the sample set. .

[0130] S1024, Based on the sample set, perform online training on the non-integrity constraint prediction model, determine the weight parameters of the converged non-integrity constraint prediction model, and obtain the trained non-integrity constraint prediction model.

[0131] Specifically, taking the Long Short-Term Memory (LSTM) network model as an example, the training process uses the following formula:

[0132] ;

[0133] Among them, subscript This indicates the time corresponding to an epoch, with subscript i representing the input gate, subscript f representing the forget gate, subscript c representing the cell state, and subscript o representing the output gate. Represents the weight matrix. U Let V represent the weight matrix, and b represent the bias vector. x Represents the input vector. This represents the hidden state. The functions involved include: the Sigmoid activation function. Hyperbolic tangent activation function . This is for Adama product operations.

[0134] In some embodiments, after obtaining the trained incompleteness constraint prediction model in S1024, the method further includes:

[0135] During the vehicle's operation, under the preset GNSS observation conditions, wheel speed data, inertial measurement data, and GNSS positioning, velocity, and attitude measurement data for each epoch are acquired.

[0136] New training samples were constructed based on wheel speed data, inertial measurement data, and GNSS positioning, velocimetry, and attitude measurement data from the same epoch.

[0137] The new training samples are added to the sample set, and the non-integrity constraint prediction model is trained online using the supplemented sample set to update the weight parameters of the non-integrity constraint prediction model, thereby obtaining the updated non-integrity constraint prediction model.

[0138] The step of calculating the non-integrity constraint components of the current epoch based on the updated non-integrity constraint prediction model is performed.

[0139] Specifically, this approach expands the sample size of the sample set and allows for continuous updates to the non-integrity constraint prediction model as the vehicle operates. This enables the trained model to adapt to the specific mechanical conditions of the vehicle and the driver's driving habits, avoiding the limitation of batch-trained non-integrity constraint prediction models that can only provide accurate prediction results in the initial stage. The model parameters can be adjusted according to the specific mechanical conditions of the vehicle and the different driving habits of the driver, thereby improving the model's computational accuracy and ultimately enhancing the navigation effect of the integrated navigation system.

[0140] In some embodiments, the weight parameters of the non-integrity constraint prediction model trained for each vehicle at the corresponding time can also be saved, and a mapping relationship between each vehicle and the weight parameters of the non-integrity constraint prediction model can be established. This allows the driver to adjust the non-integrity constraint prediction model in real time according to the mapping relationship between each vehicle and the weight parameters of the non-integrity constraint prediction model when changing to different vehicles, so as to adapt to the changes in vehicles, improve the adaptability of the non-integrity constraint prediction model, and minimize the impact caused by changes in vehicle hardware conditions.

[0141] The following are apparatus embodiments of this application, which can be used to execute the method embodiments of this application. For details not disclosed in the apparatus embodiments of this application, please refer to the method embodiments of this application.

[0142] Please see below. Figure 2 This is a schematic diagram of a vehicle navigation device provided as an exemplary embodiment of this application. The device can be implemented as all or part of a terminal through software, hardware, or a combination of both, or it can be integrated as an independent module on a server. The vehicle navigation device in this embodiment can be applied to a terminal or the cloud. The device 20 includes a data acquisition unit 201, a constraint calculation unit 202, and a positioning result calculation unit 203, wherein:

[0143] The data acquisition unit 201 is used to acquire the wheel speed data of the current epoch output by the wheel speed measurement unit of the vehicle, and to acquire the inertial measurement data of the current epoch output by the inertial measurement unit of the vehicle.

[0144] The constraint calculation unit 202 is used to input the inertial measurement data and the wheel speed data into the non-integrity constraint prediction model, and calculate the non-integrity constraint components of the current epoch through the non-integrity constraint prediction model;

[0145] The constraint calculation unit 202 is also used to construct three-dimensional motion constraint observations based on the non-integrity constraint components and the wheel speed data;

[0146] The positioning result calculation unit 203 is used to measure and update the kinematic observation equation of the Kalman filter based on the three-dimensional motion constraint observation values, obtain the state estimation result obtained by the Kalman filter update, and output the positioning result of the vehicle in the current epoch based on the state estimation result;

[0147] The non-integrity constraint prediction model is trained based on a given sample set.

[0148] It should be noted that the device 20 provided in the above embodiments is only illustrated by the division of the above functional modules when executing a vehicle combined navigation method. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the device provided in the above embodiments and the vehicle combined navigation method embodiments belong to the same concept, and the implementation process can be found in the method embodiments, which will not be repeated here.

[0149] This application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of any of the methods described above.

[0150] Please see Figure 3 This is a structural block diagram of an electronic device provided in an embodiment of this application.

[0151] like Figure 3 As shown, the electronic device 300 includes a processor 301 and a memory 302.

[0152] In this embodiment, the processor 301 is the control center of the computer system, and can be a processor of a physical machine or a processor of a virtual machine. The processor 301 may include one or more processing cores, such as a 4-core processor or an 8-core processor. The processor 301 can be implemented using at least one hardware form selected from DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array).

[0153] Processor 301 may also include a main processor and a coprocessor. The main processor is a processor used to process data in the wake-up state, also known as a CPU (Central Processing Unit). The coprocessor is a low-power processor used to process data in the standby state.

[0154] Memory 302 may include one or more computer-readable storage media, which may be non-transitory. Memory 302 may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices or flash memory devices. In some embodiments of this application, the non-transitory computer-readable storage media in memory 302 is used to store at least one instruction, which is executed by processor 301 to implement the method in the embodiments of this application.

[0155] In some embodiments, the electronic device 300 further includes a peripheral device interface 303 and at least one peripheral device 304. The processor 301, memory 302, and peripheral device interface 303 can be connected via a bus or signal line. Each peripheral device 304 can be connected to the peripheral device interface 303 via a bus, signal line, or circuit board. Specifically, the peripheral device 304 includes: a display screen, a camera, and audio circuitry. The peripheral device interface 303 can be used to connect at least one I / O (Input / Output) related peripheral device to the processor 301 and memory 302.

[0156] In some embodiments of this application, the processor 301, memory 302, and peripheral device interface 303 are integrated on the same chip or circuit board; in other embodiments of this application, any one or two of the processor 301, memory 302, and peripheral device interface 303 can be implemented on separate chips or circuit boards. This application does not specifically limit the implementation in this regard.

[0157] The block diagram of the electronic device shown in the embodiments of this application does not constitute a limitation on the electronic device 300. The electronic device 300 may include more or fewer components than shown, or combine certain components, or use different component arrangements.

[0158] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the methods in any of the foregoing embodiments. The computer-readable storage medium may include, but is not limited to, any type of disk, including floppy disks, optical disks, DVDs, CD-ROMs, microdrives, as well as magneto-optical disks, ROMs, RAMs, EPROMs, EEPROMs, DRAMs, VRAMs, flash memory devices, magnetic cards or optical cards, nanosystems (including molecular memory ICs), or any type of medium or device suitable for storing instructions and / or data.

[0159] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the parts that contribute to the related technology, can be embodied in the form of software products. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0160] 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 vehicle integrated navigation method, characterized in that, include: Obtain the wheel speed data of the current epoch output by the wheel speed measurement unit of the vehicle, and obtain the inertial measurement data of the current epoch output by the inertial measurement unit of the vehicle. The inertial measurement data and the wheel speed data are input into the non-integrity constraint prediction model, and the non-integrity constraint components for the current epoch are calculated using the non-integrity constraint prediction model, including: The inertial measurement data of the current epoch output by the inertial measurement unit of the vehicle includes the angular velocity output by the three-axis gyroscope and the specific force output by the three-axis accelerometer. The angular velocity output by the triaxial gyroscope, the specific force output by the triaxial accelerometer, and the wheel speed data of the current epoch are input into the non-integrity constraint prediction model so that the non-integrity constraint prediction model can predict the non-integrity constraint components of the current epoch based on the input data. The non-integrity constraint components include lateral velocity components and vertical velocity components. Based on the non-integrity constraint components and the wheel speed data, three-dimensional motion constraint observations are constructed, and the following formula is applied: ； in, The three-dimensional motion constraint observation value, The wheel speed data, The lateral velocity component, The vertical velocity component is denoted by t, which represents the corresponding epoch, and the superscript b indicates that the coordinate system of the corresponding parameter is the carrier coordinate system. The kinematic observation equations of the Kalman filter are updated based on the three-dimensional motion constraint observations to obtain the state estimation results obtained from the Kalman filter update. Based on the state estimation results, the positioning results of the vehicle in the current epoch are output. The non-integrity constraint prediction model is trained based on a given sample set.

2. The vehicle integrated navigation method according to claim 1, characterized in that, Before updating the kinematic observation equations of the Kalman filter based on the three-dimensional motion constraint observations, the method further includes: The velocity of the center of the inertial measurement unit in the navigation coordinate system is calculated based on the inertial measurement data, and the velocity in the navigation coordinate system is reduced to the wheel speed measurement unit to obtain the reduced velocity in the carrier coordinate system. The measurement update of the kinematic observation equations of the Kalman filter based on the three-dimensional motion constraint observations includes: The prior residual of the kinematic observation equation is calculated by subtracting the three-dimensional motion constraint observation value from the reduced velocity. The prior residual is used as the velocity information observation of the Kalman filter, and the state estimation result of the current epoch is obtained by updating the Kalman filter.

3. The vehicle integrated navigation method according to claim 2, characterized in that, The training process of the non-integrity constraint prediction model includes the following steps: The vehicle is driven for the typical convergence time of the GNSS solution strategy under the preset GNSS observation conditions, and then the vehicle is driven for the preset time under the preset GNSS observation conditions. During the preset driving time, the inertial measurement data of each epoch output by the inertial measurement unit of the vehicle, the wheel speed data of the corresponding epoch output by the wheel speed measurement unit, and the GNSS positioning, speed measurement, and attitude measurement data of the corresponding epoch output by the GNSS signal receiving unit are acquired during the preset driving time. A sample set is constructed based on the inertial measurement data of each epoch, the wheel speed data of the corresponding epoch, and the GNSS positioning, velocimetry, and attitude measurement data of the corresponding epoch to train the non-integrity constraint prediction model. The non-integrity constraint prediction model is trained online based on the sample set, and the weight parameters of the converged non-integrity constraint prediction model are determined to obtain the trained non-integrity constraint prediction model.

4. The vehicle integrated navigation method according to claim 3, characterized in that, Before the vehicle is driven under preset GNSS observation conditions for the GNSS solution strategy to reach its typical convergence time, the following steps are also included: The linkage information from the inertial measurement unit to the drive wheel is obtained through calibration; When the vehicle is stationary and meets the preset GNSS observation conditions, the initial GNSS positioning, velocity and attitude measurement data output by the GNSS signal receiving unit and the initial inertial measurement data output by the inertial measurement unit are acquired. The navigation state of the vehicle is initialized based on the lever information, initial GNSS observations, and initial inertial measurement data, and the initial state estimation results of the Kalman filter are obtained, including the initial position information, initial velocity information, and initial attitude information of the vehicle.

5. A vehicle integrated navigation method according to claim 3, characterized in that, The sample set used to train the non-integrity constraint prediction model is constructed based on the inertial measurement data of each epoch, the wheel speed data of the corresponding epoch, and the GNSS positioning, velocimetry, and attitude measurement data of the corresponding epoch. This sample set includes: Based on the GNSS positioning, velocity and attitude measurement data, the velocity of the vehicle in the corresponding epoch of the navigation coordinate system is calculated. The velocity in the navigation coordinate system is transformed to the vehicle coordinate system to obtain the velocity in the vehicle coordinate system. The lateral velocity component and the vertical velocity component of the velocity in the vehicle coordinate system are calculated. The sample set is constructed by using wheel speed data, angular velocity, and specific force from the same epoch as input to the sample, and using the lateral velocity component and vertical velocity component from the same epoch as sample labels.

6. A vehicle integrated navigation method according to any one of claims 3-5, characterized in that, The preset GNSS observation conditions include: The number of available GNSS satellites observed by the GNSS signal receiving unit of the vehicle is greater than a preset threshold, and the accuracy attenuation factor of the GNSS positioning result is less than a preset attenuation factor threshold, and the sum of the main diagonal elements of the position covariance matrix of the GNSS positioning result is less than a preset threshold. After obtaining the trained incompleteness constraint prediction model, the method further includes: During the vehicle's operation, under the preset GNSS observation conditions, wheel speed data, inertial measurement data, and GNSS positioning, velocity, and attitude measurement data for each epoch are acquired. New training samples were constructed based on wheel speed data, inertial measurement data, and GNSS positioning, velocimetry, and attitude measurement data from the same epoch. The new training samples are added to the sample set, and the non-integrity constraint prediction model is trained online using the supplemented sample set to update the weight parameters of the non-integrity constraint prediction model, thereby obtaining the updated non-integrity constraint prediction model. The step of calculating the non-integrity constraint components of the current epoch based on the updated non-integrity constraint prediction model is performed.

7. A vehicle navigation device based on the vehicle navigation method according to any one of claims 1-6, characterized in that, include: The data acquisition unit is used to acquire the wheel speed data of the current epoch output by the wheel speed measurement unit of the vehicle, and to acquire the inertial measurement data of the current epoch output by the inertial measurement unit of the vehicle. The constraint calculation unit is used to input the inertial measurement data and the wheel speed data into the non-integrity constraint prediction model, and calculate the non-integrity constraint components of the current epoch through the non-integrity constraint prediction model. The constraint calculation unit is also used to construct three-dimensional motion constraint observations based on the non-integrity constraint components and the wheel speed data; The positioning result calculation unit is used to measure and update the kinematic observation equation of the Kalman filter based on the three-dimensional motion constraint observation values, obtain the state estimation result obtained by the Kalman filter update, and output the positioning result of the vehicle in the current epoch based on the state estimation result. The non-integrity constraint prediction model is trained based on a given sample set.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 6.

Images