Vehicle positioning method and related device

Abstract

The invention discloses a vehicle positioning method and a related device, and the method comprises the steps: determining a slip angle corresponding to each wheel through employing a Transform wheel speed slip angle estimation model according to the wheel speed pulses corresponding to four wheels in obtained wheel speed sensor data, then determining a first relative position and a first attitude change of the target vehicle at the current moment relative to the previous moment according to the obtained inertial sensor data by using an inertial measurement unit pre-integration model; determining a second relative position and a second attitude change of the target vehicle at the current moment relative to the previous moment through a wheel speed pulse pre-integration measurement model according to the data of the wheel speed sensor and the slip angle corresponding to each wheel; and determining a target position and a target attitude of the target vehicle at the current moment according to the first relative position, the first attitude change, the second relative position and the second attitude change. Therefore, the vehicle positioning precision is improved.

Description

Technical Field

[0001] The present application relates to the field of vehicle technology, and in particular to a vehicle positioning method and related devices. Background Art

[0002] Vehicle positioning refers to the process of determining the position and posture of a vehicle on a map. In the application of autonomous driving and advanced driver assistance systems (ADAS), the real-time positioning accuracy of vehicle positioning is required to be relatively high.

[0003] At present, the relevant technologies generally use one or more of the global positioning system (GPS), visual sensors, inertial sensors, and wheel speed sensors to locate the vehicle. Among them, vehicle positioning based on inertial sensors and wheel speed sensors has the advantages of low cost, high reliability, and strong real-time performance.

[0004] However, the inertial sensor data obtained based on the inertial sensor may be affected by factors such as noise, temperature, zero bias, etc., resulting in low accuracy of the inertial sensor data, and further resulting in low positioning accuracy; the wheel speed sensor data obtained based on the wheel speed sensor may be affected by vehicle actions such as turning, acceleration, and braking, resulting in low accuracy of the wheel speed sensor data, and further resulting in low positioning accuracy. Summary of the invention

[0005] Based on the above problems, the present application provides a vehicle positioning method and related devices to solve the problem of low positioning accuracy caused by low data accuracy.

[0006] The embodiments of the present application disclose the following technical solutions:

[0007] In a first aspect, an embodiment of the present application provides a vehicle positioning method, the method comprising:

[0008] Obtain the inertial sensor data and wheel speed sensor data of the target vehicle at the current moment;

[0009] Based on the Transformer wheel speed slip angle estimation model, the slip angle corresponding to each wheel is determined according to the wheel speed pulses corresponding to each of the four wheels in the wheel speed sensor data;

[0010] Determine a first relative position and a first posture change of the target vehicle at a current moment relative to a previous moment according to the inertial measurement unit pre-integration model and the inertial sensor data;

[0011] The second relative position and the second posture change of the target vehicle at the current moment relative to the previous moment are determined according to the wheel speed sensor data and the slip angles corresponding to each wheel through the wheel speed pulse pre-integration measurement model;

[0012] The target position and target posture of the target vehicle at the current moment are determined according to the first relative position, the first posture change, the second relative position and the second posture change through the invariant Kalman filter model.

[0013] Optionally, determining the slip angle corresponding to each wheel according to the wheel speed pulses corresponding to each of the four wheels in the wheel speed sensor data includes:

[0014] Get the target vehicle's driving direction at the current moment;

[0015] Determine the wheel speed corresponding to each wheel according to the wheel speed sensor data;

[0016] Determine the linear speed corresponding to each wheel based on the wheel speed corresponding to each wheel;

[0017] The slip angle corresponding to each wheel is determined according to the driving direction and the linear speed corresponding to each wheel.

[0018] Optionally, determining a second relative position and a second posture change of the target vehicle at a current moment relative to a previous moment according to wheel speed sensor data and slip angles corresponding to each wheel through a wheel speed pulse pre-integration measurement model includes:

[0019] Obtain the historical wheel speed sensor data of the target vehicle at the previous moment;

[0020] Determine the wheel speed increment corresponding to each wheel according to the wheel speed sensor data and the historical wheel speed sensor data;

[0021] The wheel speed increment corresponding to each wheel is corrected using the slip angle corresponding to each wheel to obtain the target wheel speed increment corresponding to each wheel;

[0022] The second relative position and the second posture change of the target vehicle at the current moment relative to the previous moment are determined according to the target wheel speed increment through the wheel speed pulse pre-integration measurement model.

[0023] Optionally, using an inertial measurement unit pre-integration model, determining a first relative position and a first posture change of the target vehicle at a current moment relative to a previous moment according to inertial sensor data, includes:

[0024] Determine the linear acceleration of the target vehicle at the current moment according to the accelerometer data in the inertial sensor data;

[0025] Determine the angular velocity of the target vehicle at the current moment according to the gyroscope data in the inertial sensor data;

[0026] The inertial measurement unit pre-integration model is used to determine a first relative position and a first posture change of the vehicle at a current moment relative to a previous moment according to the linear acceleration and the angular velocity.

[0027] Optionally, determining a target position and a target posture of the target vehicle at a current moment according to the first relative position, the first posture change, the second relative position, and the second posture change through an invariant Kalman filter model includes:

[0028] According to the first relative position and the first posture change, the predicted position and the predicted posture of the target vehicle at the current moment are determined by a state transfer equation in an invariant Kalman filter model;

[0029] determining a constant Kalman gain based on the predicted position, the predicted attitude, the second relative position, and the second attitude change;

[0030] The predicted position and predicted attitude are updated using the invariant Kalman gain to obtain the target position and target attitude of the target vehicle at the current moment.

[0031] In a second aspect, an embodiment of the present application provides a vehicle positioning device, the device comprising:

[0032] An acquisition module is used to acquire the inertial sensor data and wheel speed sensor data of the target vehicle at the current moment;

[0033] The slip angle determination module is used to determine the slip angle corresponding to each wheel according to the wheel speed pulses corresponding to each of the four wheels in the wheel speed sensor data based on the Transformer wheel speed slip angle estimation model;

[0034] A first determination module is used to determine a first relative position and a first posture change of the target vehicle at a current moment relative to a previous moment according to inertial sensor data by using an inertial measurement unit pre-integration model;

[0035] A second determination module is used to determine a second relative position and a second posture change of the target vehicle at a current moment relative to a previous moment according to the wheel speed sensor data and the slip angle corresponding to each wheel through a wheel speed pulse pre-integration measurement model;

[0036] The vehicle positioning module is used to determine the target position and target posture of the target vehicle at the current moment according to the first relative position, the first posture change, the second relative position and the second posture change through an invariant Kalman filter model.

[0037] Optionally, the slip angle determination module includes:

[0038] A driving direction acquisition unit, used to acquire the driving direction of the target vehicle at the current moment;

[0039] A wheel speed determination unit, used to determine the wheel speed corresponding to each wheel according to wheel speed sensor data;

[0040] A linear speed determination unit, used to determine the linear speed corresponding to each wheel based on the wheel speed corresponding to each wheel;

[0041] The slip angle determination unit is used to determine the slip angle corresponding to each wheel according to the driving direction and the linear speed corresponding to each wheel.

[0042] Optionally, the second determining module includes:

[0043] A historical data acquisition unit, used to acquire historical wheel speed sensor data of the target vehicle at the previous moment;

[0044] A wheel speed increment acquisition unit, used to determine the wheel speed increment corresponding to each wheel according to the wheel speed sensor data and the historical wheel speed sensor data;

[0045] a target wheel speed increment determining unit, configured to correct the wheel speed increment corresponding to each wheel by using the slip angle corresponding to each wheel, so as to obtain the target wheel speed increment corresponding to each wheel;

[0046] The second determination unit is used to determine a second relative position and a second posture change of the target vehicle at a current moment relative to a previous moment according to the target wheel speed increment through the wheel speed pulse pre-integration measurement model.

[0047] In a third aspect, an embodiment of the present application provides a computer device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the vehicle positioning method as described in the first aspect is implemented.

[0048] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein instructions are stored in the computer-readable storage medium. When the instructions are executed on a terminal device, the terminal device executes the vehicle positioning method as described in the first aspect.

[0049] Compared with the prior art, this application has the following beneficial effects:

[0050] The vehicle positioning method provided in the embodiment of the present application determines the slip angle corresponding to each wheel based on the Transformer wheel speed slip angle estimation model according to the wheel speed pulses corresponding to each of the four wheels in the acquired wheel speed sensor data, and uses the inertial measurement unit pre-integration model to determine the first relative position and first posture change of the target vehicle at the current moment relative to the previous moment according to the acquired inertial sensor data, and then uses the wheel speed pulse pre-integration measurement model to determine the second relative position and second posture change of the target vehicle at the current moment relative to the previous moment according to the wheel speed sensor data and the slip angles corresponding to each wheel, and further uses the invariant Kalman filter model to determine the target position and target posture of the target vehicle at the current moment according to the first relative position, the first posture change, the second relative position and the second posture change. Among them, two relative positions and two posture changes are determined respectively through wheel speed sensor data and inertial sensor data, and the two relative positions and two posture changes are fused using the invariant Kalman filter model to determine the target position and target posture of the target vehicle, that is, the noise generated by the wheel speed sensor and the inertial sensor are optimized respectively through the wheel speed pulse pre-integration measurement model and the inertial measurement unit pre-integration model to improve the accuracy of relative position and posture changes, and further by using the invariant Kalman filter model to process the vehicle dynamic process with low error and high stability, the error caused by the wheel speed sensor and the inertial sensor is reduced, and the accuracy of vehicle positioning is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0052] Figure 1 A schematic diagram of a vehicle positioning method provided in an embodiment of the present application;

[0053] Figure 2 A schematic diagram of a sliding angle provided in an embodiment of the present application;

[0054] Figure 3 A schematic diagram of the structure of a vehicle positioning device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0055] As described above, in the research on vehicle positioning, it was found that vehicle positioning refers to the process of determining the position and posture of the vehicle on the map. In the application of autonomous driving and advanced driver assistance systems (ADAS), the real-time positioning accuracy of vehicle positioning is required to be relatively high.

[0056] At present, the relevant technologies generally use one or more of the global positioning system (GPS), visual sensors, inertial sensors, and wheel speed sensors to locate the vehicle. Among them, vehicle positioning based on inertial sensors and wheel speed sensors has the advantages of low cost, high reliability, and strong real-time performance.

[0057] However, the inertial sensor data obtained based on the inertial sensor may be affected by factors such as noise, temperature, zero bias, etc., resulting in low accuracy of the inertial sensor data, and further resulting in low positioning accuracy; the wheel speed sensor data obtained based on the wheel speed sensor may be affected by vehicle actions such as turning, acceleration, and braking, resulting in low accuracy of the wheel speed sensor data, and further resulting in low positioning accuracy.

[0058] In some hypothetical implementations, inertial sensor data and wheel speed sensor data can be fused and positioned by adopting Kalman filtering or its variants, but this method requires accurate system models and noise models to be established in advance, and these models may be difficult to obtain or adapt to different driving environments, resulting in low vehicle positioning accuracy.

[0059] In order to solve the above problems, the embodiment of the present application provides a vehicle positioning method and related devices. The method includes: determining the slip angle corresponding to each wheel according to the wheel speed pulses corresponding to each of the four wheels in the acquired wheel speed sensor data, and then using the inertial measurement unit pre-integration model to determine the first relative position and first posture change of the target vehicle at the current moment relative to the previous moment according to the acquired inertial sensor data, and further determining the second relative position and second posture change of the target vehicle at the current moment relative to the previous moment according to the wheel speed sensor data and the slip angles corresponding to each wheel through the wheel speed pulse pre-integration measurement model, and further determining the target position and target posture of the target vehicle at the current moment according to the first relative position, the first posture change, the second relative position and the second posture change through the invariant Kalman filter model.

[0060] In this way, two relative positions and two posture changes are determined respectively through the wheel speed sensor data and the inertial sensor data, that is, the noise generated by the wheel speed sensor and the inertial sensor are optimized respectively through the wheel speed pulse pre-integration measurement model and the inertial measurement unit pre-integration model to improve the accuracy of the relative position and posture changes, and the two relative positions and two posture changes are fused using the invariant Kalman filter model to determine the target position and target posture of the target vehicle, that is, by using the invariant Kalman filter model to process the vehicle dynamic process with low error and high stability, the errors caused by the wheel speed sensor and the inertial sensor are reduced, and the accuracy of vehicle positioning is improved.

[0061] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0062] See also Figure 1 , which is a flow chart of a vehicle positioning method provided by an embodiment of the present application, combined with Figure 1 As shown, the vehicle positioning method provided in the embodiment of the present application may include:

[0063] S101: Acquire inertial sensor data and wheel speed sensor data of the target vehicle at the current moment.

[0064] The target vehicle means the vehicle to be positioned. The inertial sensor data means the data collected by the inertial sensor of the target vehicle. In some possible implementations, the inertial sensor data includes but is not limited to accelerometer data and gyroscope data. The wheel speed sensor data means the data collected by the wheel speed sensor of the target vehicle. In some possible implementations, the wheel speed sensor data includes but is not limited to the wheel speed pulses of each wheel of the target vehicle.

[0065] An inertial sensor is a device that can measure the motion state of an object. Inertial sensors mainly include accelerometers (or acceleration sensors) and angular velocity sensors (gyroscopes) and their single, dual, and three-axis combined IMUs (inertial measurement units), AHRS (attitude reference systems including magnetic sensors). In the automotive industry, inertial sensors can also be used in electronic stability programs (ESP), adaptive cruise control (ACC), etc.

[0066] The wheel speed sensor is a sensor used to measure the rotation speed of the four wheels of the target vehicle and is an important component of the vehicle's electronic system.

[0067] It should be understood that in order to ensure the real-time and reliability of data and reduce the cost of vehicle positioning, in the embodiment of the present application, vehicle positioning is still performed through inertial sensors and wheel speed sensors.

[0068] S102: Based on the Transformer wheel speed slip angle estimation model, the slip angle corresponding to each wheel is determined according to the wheel speed pulses corresponding to each of the four wheels in the wheel speed sensor data.

[0069] Wheel speed pulses refer to the number of gears that rotate as the vehicle rotates, which is sensed by the wheel speed sensor through the electromagnetic induction principle or the Hall effect. These sensed pulse signals are counted and processed by the electronic control unit, so that the distance and speed of the wheel can be calculated.

[0070] The slip angle refers to the angle between the longitudinal direction of the tread of the target vehicle's wheel and the speed of the vehicle body connection point, combined with Figure 2 As shown, x w Indicates the longitudinal direction of the wheel tread, y w It represents the direction perpendicular to the horizontal direction of the wheel, v represents the direction of the speed of the vehicle body connection point, δ represents the angle between the longitudinal direction of the wheel tread and the centerline direction of the vehicle, and α represents the angle between the speed direction of the vehicle body connection point and the longitudinal direction of the wheel tread.

[0071] It should be understood that since the wheel speed pulses in the wheel speed sensor data can determine the direction and speed of the wheel, and the slip angle is related to the direction of the wheel rim of the target vehicle, the slip angle of the wheel can be determined by the wheel speed pulses.

[0072] In a possible implementation, step S102 may be obtained in the following manner:

[0073] A1: Get the target vehicle's current driving direction.

[0074] The driving direction refers to the direction of travel of the target vehicle. Since the slip angle is the angle between the direction of travel and the direction indicated by the wheel rim, it is necessary to obtain the direction of travel of the target vehicle. The driving direction of the vehicle can be determined in a variety of ways, such as using the vehicle's steering angle sensor data, GPS data, or by fusing information from multiple sensors, which is not specifically limited here.

[0075] A2: Determine the wheel speed corresponding to each wheel based on the wheel speed sensor data.

[0076] The wheel speed refers to the rotation speed of the wheel when it is in motion. In some possible implementations, the wheel speed can be calculated based on the wheel speed pulse frequency of the wheel speed sensor.

[0077] A3: Determine the linear speed corresponding to each wheel based on the wheel speed corresponding to each wheel.

[0078] The linear velocity of a wheel refers to the speed of any point on the wheel when it moves in a circular motion about a fixed axis. Specifically, the linear velocity is a physical quantity that describes the speed and direction of a particle moving in a curve. The direction of the linear velocity is along the tangent direction of the moving track, so it is also called the tangential velocity. In the calculation of the linear velocity, the formula is usually used. Where v represents the linear velocity, s represents the distance traveled by the object per unit time, and t represents the unit time.

[0079] Take a wheel as an example. Assuming the radius of the wheel is r and the wheel rotates n times per second, the distance the wheel travels per second is the circumference of the wheel multiplied by the number of revolutions, that is, s = 2πr × n. Then, substitute s into the linear velocity formula to get the linear velocity of the wheel. In other words, the radius of the wheel can be obtained, and the linear velocity of the wheel can be further calculated in combination with the wheel speed.

[0080] A4: Determine the slip angle of each wheel based on the driving direction and the linear speed of each wheel.

[0081] It should be understood that in the embodiment of the present application, the slip angle can be calculated by comparing the linear velocity vector of the wheel and the travel direction vector of the vehicle.

[0082] It should be noted that in actual applications, the wheel slip angle may be affected by many factors, such as road conditions, tire wear, vehicle load, etc. Therefore, when calculating the slip angle, it may be necessary to consider these additional factors and make corresponding corrections.

[0083] It should be understood that the trained Transformer wheel speed slip angle estimation model can be used to estimate the slip angle of each wheel of the target vehicle based on the wheel speed sensor data.

[0084] The Transformer model is a deep learning model based on the self-attention mechanism. It can capture the temporal and spatial characteristics of wheel speed sensor data. It is trained with a large amount of experimental data and can adapt to different road conditions and driving behaviors.

[0085] As an example, using the Transformer model, the slip angle of each wheel of the target vehicle is estimated based on the wheel speed sensor data, which can be:

[0086] B1: Preprocess the wheel speed sensor data.

[0087] The wheel speed sensor data of each wheel of the target vehicle is obtained, which includes information such as rotation speed and timestamp, and then the wheel speed sensor data is cleaned and standardized to ensure data quality and data consistency. In addition, additional features such as acceleration, speed change, etc. can also be calculated.

[0088] B2: Perform feature encoding on the preprocessed wheel speed sensor data to obtain a feature vector.

[0089] It should be understood that after the wheel speed sensor data is processed, it is necessary to further convert the pre-processed wheel speed sensor data into a format suitable for processing by the Transformer model. That is, the time series data is converted into a series of embedded vectors. In a possible implementation, position encoding (such as sine and cosine position encoding) can be considered to capture the order information in the time series of the pre-processed wheel speed sensor data.

[0090] B3: Construct a Transformer-based slip angle determination model.

[0091] A Transformer-based slip angle determination model is constructed based on the feature vector, and the slip angle determination model can process the sequence input of the wheel speed sensor data. The slip angle determination model may include multiple encoder layers, each encoder layer including a self-attention mechanism and a feedforward neural network. In a possible implementation, a regression layer may be added at the output end of the slip angle determination model to predict the slip angle.

[0092] B4: Training and optimization of the slip angle determination model.

[0093] Specifically, a labeled dataset can be prepared, which contains wheel speed sensor data and corresponding slip angle labels. The slip angle determination model is trained using a supervised learning method, and the model parameters are optimized by minimizing the error between the predicted slip angle and the actual slip angle. Furthermore, a suitable loss function (such as mean square error loss) and optimizer (such as Adam optimizer) can be used for training.

[0094] B5: Slip angle determination model evaluation and tuning.

[0095] Specifically, the slip angle determination model can be evaluated using a validation set to check the prediction performance of the slip angle determination model. The slip angle determination model can be tuned based on the evaluation results, including adjusting the model structure, hyperparameters, etc. Techniques such as cross-validation can be used to more comprehensively evaluate the performance of the slip angle determination model.

[0096] B6: Determine the slip angle of each wheel using the slip angle determination model.

[0097] Specifically, the new wheel speed sensor data is input into the trained slip angle determination model to determine the slip angle of each wheel.

[0098] That is, in the embodiment of the present application, by utilizing the Transformer model, the wheel slip angle can be effectively estimated, the reliability of the wheel speed sensor data can be improved, and at the same time, it can adapt to different road conditions and driving behaviors and improve the robustness of positioning.

[0099] S103: using an inertial measurement unit pre-integration model, and determining a first relative position and a first posture change of the target vehicle at a current moment relative to a previous moment according to inertial sensor data.

[0100] The inertial measurement unit (IMU) pre-integration model is a method of processing inertial sensor data. The inertial measurement unit pre-integration model aims to estimate the relative position and attitude changes of the target vehicle by integrating continuous IMU measurements without knowing the initial state (i.e. the position and attitude at the previous moment).

[0101] In a possible implementation, step S103 may include:

[0102] C1: Determine the linear acceleration of the target vehicle at the current moment based on the accelerometer data in the inertial sensor data.

[0103] The accelerometer data is mainly used to measure the acceleration of the target vehicle in a certain direction. Therefore, in the embodiment of the present application, the linear acceleration of the target vehicle at the current moment can be determined by the accelerometer data in the inertial sensor data.

[0104] In some possible implementations, the accelerometer data may be debiased to eliminate the effects of gravity acceleration and the bias of the inertial sensor itself.

[0105] C2: Determine the angular velocity of the target vehicle at the current moment based on the gyroscope data in the inertial sensor data.

[0106] The gyroscope data describes the speed and direction of rotation of the target vehicle. Therefore, in the embodiment of the present application, the angular velocity of the target vehicle at the current moment can be determined by the gyroscope data.

[0107] In some possible implementations, the gyroscope data may be denoised and filtered to reduce noise and interference.

[0108] C3: Using the inertial measurement unit pre-integration model, determine the first relative position and first posture change of the vehicle at the current moment relative to the previous moment according to the linear acceleration and angular velocity.

[0109] In one possible implementation, a time interval may be set, such as the time period from the previous moment to the current moment. Then, within this time period, the linear acceleration and angular velocity are pre-integrated. The pre-integration process is to fuse the inertial sensor data in continuous time to estimate the first relative position and the first attitude change of the target vehicle. For the pre-integration of the first relative position, the de-biased linear acceleration may be integrated to obtain the first relative position. For the pre-integration of the first attitude change, the angular velocity may be used to update the attitude change, such as by using an update equation of a quaternion or a rotation matrix.

[0110] In addition, the noise and bias of the inertial sensor data can also be considered during the pre-integration process, so they can be modeled by introducing random walk noise and bias terms, and estimated and compensated in the subsequent optimization process.

[0111] That is, in the embodiment of the present application, the characteristics of the pre-integration model of the inertial measurement unit are utilized, the influence of the noise and zero bias of the inertial sensor is considered, and the noise and zero bias are processed to improve the accuracy of the data.

[0112] S104: Determine a second relative position and a second posture change of the target vehicle at a current moment relative to a previous moment according to the wheel speed pulse pre-integration measurement model and the wheel speed sensor data and the slip angles corresponding to the wheels.

[0113] The current time refers to the current time point of the target vehicle, and the previous time is the time point between the current time. For example, if the current time is 12:30, the previous time can be 12:25. The relative position refers to the position of the target vehicle at the current time relative to the target vehicle at the previous time. For example, if the target vehicle is at point B at 12:30 and the target vehicle is at point A at 12:25, the relative position is the position of point B relative to point A. Posture change refers to the conversion of different states of the target vehicle at the current time and the previous time during its movement. For example, if the target vehicle is in posture X at 12:30 and in posture Y at 12:25, the posture change is the change of posture X relative to posture Y.

[0114] The wheel speed pulse pre-integration measurement model is a model for estimating the vehicle motion state, which combines wheel speed pulses and pre-integration technology. Pre-integration technology is a method for processing continuous time series data, which can accumulate data over a period of time to obtain more accurate estimation results. In the wheel speed pulse pre-integration measurement model, the wheel speed sensor data is first obtained through the wheel speed sensor and converted into wheel speed pulse signals, and then these pulse signals are accumulated and processed using pre-integration technology to eliminate noise and errors, and determine the second relative position and second posture change of the target vehicle.

[0115] In some possible implementations, step S104 may determine the second relative position and the second posture change in the following manner:

[0116] D1: Obtain the historical wheel speed sensor data of the target vehicle at the previous moment.

[0117] The historical wheel speed sensor data refers to the wheel speed sensor data of the target vehicle at the last moment.

[0118] D2: Determine the wheel speed increment corresponding to each wheel based on the wheel speed sensor data and the historical wheel speed sensor data.

[0119] Wheel speed increment means the amount of rotation of the wheel between the current moment and the last moment. In some possible implementations, the wheel speed increment of the wheel in the time between the current moment and the last moment can be obtained by comparing the wheel speed pulse number or wheel speed pulse frequency at the current moment and the last moment.

[0120] D3: Correct the wheel speed increment corresponding to each wheel using the slip angle corresponding to each wheel to obtain the target wheel speed increment corresponding to each wheel.

[0121] It should be understood that the influence of the slip angle is also considered in the embodiment of the present application. Therefore, the wheel speed increment needs to be corrected in combination with the slip angle of the wheel to obtain the target wheel speed increment of the wheel and improve data accuracy.

[0122] D4: Through the wheel speed pulse pre-integration measurement model, the second relative position and second posture change of the target vehicle at the current moment relative to the previous moment are determined according to the target wheel speed increment.

[0123] By using the wheel speed pulse pre-integration measurement model to process the target wheel speed increment, the wheel speed sensor data at the current moment and the previous moment can be effectively fused, thereby estimating the second relative position and second posture change of the target vehicle at the current moment relative to the previous moment.

[0124] That is, in the embodiment of the present application, by using the wheel speed pulse pre-integration measurement model, the second relative position and the second posture change from the current moment to the previous moment can be calculated according to the wheel speed sensor data and the slip angle of the four wheels. Compared with the speed constraint of the traditional wheel speed pulse, the wheel speed pulse pre-integration measurement model can more reasonably optimize the differential calculation noise amplification effect caused by the delay in the update phase, thereby effectively reducing the cumulative error and improving the measurement accuracy.

[0125] S105: Determine the target position and target posture of the target vehicle at the current moment according to the first relative position, the first posture change, the second relative position, and the second posture change through the invariant Kalman filter model.

[0126] The invariant Kalman filter model is a variant of the Kalman filter, which is mainly used in nonlinear systems or non-Gaussian noise. The core idea of ​​the invariant Kalman filter model is to transform the nonlinear system into a linear system for processing by introducing an invariant measurement model. This transformation is achieved by linearizing the state equation and measurement equation of the nonlinear system, making the Kalman filter applicable to nonlinear systems.

[0127] It should be understood that in the embodiment of the present application, by adopting the Invariant Kalman Filter (IEKF) model as the core of data fusion, compared with the traditional Kalman filter, the IEKF model is more robust in processing nonlinear systems, especially in high dynamic and high noise environments. The IEKF model is suitable for vehicle dynamic processes, can reduce the accumulation of errors, automatically adjust the filter gain, and reduce the accumulated error, thereby achieving the purpose of improving the vehicle positioning accuracy.

[0128] In a possible implementation, step S105 may include:

[0129] E1: According to the first relative position and the first posture change, the predicted position and predicted posture of the target vehicle at the current moment are determined through the state transfer equation in the invariant Kalman filter model.

[0130] In the invariant Kalman filter model, the state transfer equation describes the evolution of the target vehicle's state from the previous moment to the current moment. For nonlinear systems, the state transfer equation usually takes the following form:

[0131] x_k=F*x_{k-1}+B*u_k+w_k;

[0132] Where: x_k is the state vector at the current moment (moment k), x_{k-1} is the state vector at the previous moment (moment k-1), and F is the state transfer matrix. B is the control input matrix, which is used to map the control input u_k to the state space. u_k is the control input vector, which represents the external influence on the system. w_k is the system noise vector, which represents the state uncertainty caused by factors such as modeling errors and sensor noise.

[0133] In some possible implementations, before executing step E1, the following may be included: setting the initial position and initial attitude estimation value of the invariant Kalman filter model, and the corresponding covariance matrix. At the same time, the parameters of the invariant Kalman filter model are also set, including but not limited to the state transfer matrix, the control input matrix, the observation matrix, the process noise covariance matrix and the observation noise covariance matrix.

[0134] It should be understood that the predicted position and predicted posture of the target vehicle at the current moment can be predicted using the state transfer equation based on the first relative position, the first posture change, and the control input (if any, otherwise 0). At the same time, the covariance matrix of the predicted state is updated.

[0135] E2: Determine the invariant Kalman gain based on the predicted position, the predicted attitude, the second relative position and the second attitude change.

[0136] It should be understood that the second relative position and the second posture change can be used as observation values ​​of the invariant Kalman filter model, and the observation matrix of the invariant Kalman filter model can be further used to calculate the invariant Kalman gain in combination with the predicted position, predicted posture and observation values.

[0137] E3: Use the constant Kalman gain to update the predicted position and predicted attitude to obtain the target position and target attitude of the target vehicle at the current moment.

[0138] It should be understood that once the invariant Kalman gain is calculated, the invariant Kalman gain can be used to update the position and attitude of the target vehicle. The updating process is to perform a weighted fusion of the observed value and the predicted value, where the weight is determined by the invariant Kalman gain. At the same time, the covariance matrix of the estimated state is updated.

[0139] That is, the embodiment of the present application utilizes the invariant Kalman filter model, which can automatically adjust the filter gain, reduce the cumulative error, and improve the accuracy of vehicle positioning without the need to predict the system model and the noise model.

[0140] The vehicle positioning method provided in the embodiment of the present application determines two relative positions and two posture changes respectively through wheel speed sensor data and inertial sensor data, and fuses the two relative positions and two posture changes using an invariant Kalman filter model to determine the target position and target posture of the target vehicle, that is, the noise generated by the wheel speed sensor and the inertial sensor are optimized respectively through the wheel speed pulse pre-integration measurement model and the inertial measurement unit pre-integration model to improve the accuracy of relative position and posture changes, and further uses the invariant Kalman filter model to process the vehicle dynamic process with low error and high stability, thereby reducing the errors caused by the wheel speed sensor and the inertial sensor and improving the accuracy of vehicle positioning.

[0141] Based on the vehicle positioning method provided in the above embodiment, the present application also provides a vehicle positioning device, see Figure 3 , which is a schematic diagram of the structure of the vehicle positioning device provided in the embodiment of the present application, combined with Figure 3 As shown, the vehicle positioning device 300 provided in the embodiment of the present application may include:

[0142] The acquisition module 301 is used to acquire the inertial sensor data and wheel speed sensor data of the target vehicle at the current moment;

[0143] The slip angle determination module 302 is used to determine the slip angle corresponding to each wheel according to the wheel speed pulses corresponding to each of the four wheels in the wheel speed sensor data based on the Transformer wheel speed slip angle estimation model;

[0144] A first determination module 303 is used to determine a first relative position and a first posture change of the target vehicle at a current moment relative to a previous moment according to the inertial sensor data by using an inertial measurement unit pre-integration model;

[0145] A second determination module 304 is used to determine a second relative position and a second posture change of the target vehicle at a current moment relative to a previous moment according to the wheel speed sensor data and the slip angle corresponding to each wheel through a wheel speed pulse pre-integration measurement model;

[0146] The vehicle positioning module 305 is used to determine the target position and target posture of the target vehicle at the current moment according to the first relative position, the first posture change, the second relative position and the second posture change through the invariant Kalman filter model.

[0147] As an example, the slip angle determination module 302 includes:

[0148] A driving direction acquisition unit, used to acquire the driving direction of the target vehicle at the current moment;

[0149] A wheel speed determination unit, used to determine the wheel speed corresponding to each wheel according to wheel speed sensor data;

[0150] A linear speed determination unit, used to determine the linear speed corresponding to each wheel based on the wheel speed corresponding to each wheel;

[0151] The slip angle determination unit is used to determine the slip angle corresponding to each wheel according to the driving direction and the linear speed corresponding to each wheel.

[0152] As an example, the second determining module 304 includes:

[0153] A historical data acquisition unit, used to acquire historical wheel speed sensor data of the target vehicle at the previous moment;

[0154] A wheel speed increment acquisition unit, used to determine the wheel speed increment corresponding to each wheel according to the wheel speed sensor data and the historical wheel speed sensor data;

[0155] a target wheel speed increment determination unit, configured to correct the wheel speed increment corresponding to each wheel using the slip angle corresponding to each wheel, so as to obtain the target wheel speed increment corresponding to each wheel;

[0156] The second determination unit is used to determine a second relative position and a second posture change of the target vehicle at a current moment relative to a previous moment according to the target wheel speed increment through a wheel speed pulse pre-integration measurement model.

[0157] As an example, the first determining module 303 includes:

[0158] A linear acceleration determination unit, used to determine the linear acceleration of the target vehicle at a current moment according to the accelerometer data in the inertial sensor data;

[0159] An angular velocity determination unit, used to determine the angular velocity of the target vehicle at a current moment according to the gyroscope data in the inertial sensor data;

[0160] The first determination unit is used to determine a first relative position and a first posture change of the vehicle at a current moment relative to a previous moment according to the linear acceleration and the angular velocity by using an inertial measurement unit pre-integration model.

[0161] As an example, the vehicle positioning module 305 includes:

[0162] A prediction unit, configured to determine a predicted position and a predicted posture of the target vehicle at a current moment through a state transfer equation in an invariant Kalman filter model according to the first relative position and the first posture change;

[0163] A gain determination unit, configured to determine an invariant Kalman gain according to the predicted position, the predicted attitude, the second relative position, and the second attitude change;

[0164] The vehicle positioning unit is used to update the predicted position and predicted attitude using the invariant Kalman gain to obtain the target position and target attitude of the target vehicle at the current moment.

[0165] The vehicle positioning device provided in the embodiment of the present application has the same beneficial effects as the vehicle positioning method provided in the above embodiment, so they are not described in detail.

[0166] The embodiments of the present application also provide corresponding devices and computer storage media for implementing the solutions provided by the embodiments of the present application.

[0167] The device includes a memory and a processor, the memory is used to store instructions or codes, and the processor is used to execute the instructions or codes so that the device executes the vehicle positioning method described in any embodiment of the present application.

[0168] The computer storage medium stores codes, and when the codes are executed, the device executing the codes implements the vehicle positioning method described in any embodiment of the present application.

[0169] It should be noted that each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device and equipment embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments. The device and equipment embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components indicated as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. A person of ordinary skill in the art can understand and implement it without paying creative labor.

[0170] The "first" and "second" in the names such as "first" and "second" (if any) mentioned in the embodiments of the present application are only used as name identifiers and do not represent the first or second in order.

[0171] Through the description of the above implementation methods, it can be known that those skilled in the art can clearly understand that all or part of the steps in the above-mentioned embodiment method can be implemented by means of software plus a general hardware platform. Based on such an understanding, the technical solution of the present application can be embodied in the form of a software product, and the computer software product can be stored in a storage medium, such as a read-only memory (ROM) / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network communication device such as a router) to execute the methods described in each embodiment of the present application or some parts of the embodiments.

[0172] The data collection and processing in this application should be strictly in accordance with the relevant requirements of national laws, administrative regulations and mandatory national standards when applied in the examples, and the informed consent or separate consent of the subject of personal information (hereinafter referred to as the subject) should be obtained. Subsequent data use and processing can only be carried out within the scope of authorization of laws, regulations and subjects. Relevant requirements include but are not limited to:

[0173] (1) The individual consent of the subject must be obtained, and the subject may independently set the consent period;

[0174] (2) Under the premise of ensuring the driving safety of the vehicle, inform the subject of the status of data collection in an appropriate manner and provide convenience for the subject to terminate data collection;

[0175] (3) If the subject requests deletion of part or all of the data related to the subject, the processor shall delete the relevant data within ten working days;

[0176] (IV) A data processor may only collect data related to a subject if the data has the purpose of enhancing driving safety, has the purpose of directly serving the subject and / or is sufficiently necessary; data related to the subject includes but is not limited to biometric feature data (such as fingerprints, voiceprints, faces, heart rates, etc.), enhanced driving safety data, intelligent driving data, navigation data, etc.

[0177] (V) The necessity of data collection and its impact on the subject may be notified through prominent means such as user manuals, vehicle display panels, voice, text, and vehicle-related applications;

[0178] (VI) If this application involves the collection of the subject's voice, the voice will be converted into text in the vehicle and uploaded only in text form. Personal voiceprint information will not be involved, and the subject's personal information will be encrypted and uploaded.

[0179] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A vehicle positioning method, characterized in that: The method comprises: Obtain the inertial sensor data and wheel speed sensor data of the target vehicle at the current moment; Based on the Transformer wheel speed slip angle estimation model, the slip angle corresponding to each wheel is determined according to the wheel speed pulses corresponding to each of the four wheels in the wheel speed sensor data; Determine, using an inertial measurement unit pre-integration model, a first relative position and a first posture change of the target vehicle at a current moment relative to a previous moment according to inertial sensor data; Determine the second relative position and the second posture change of the target vehicle at the current moment relative to the previous moment according to the wheel speed sensor data and the slip angle corresponding to each of the wheels through the wheel speed pulse pre-integration measurement model; The target position and target posture of the target vehicle at the current moment are determined according to the first relative position, the first posture change, the second relative position and the second posture change through the invariant Kalman filter model.

2. The method according to claim 1, characterized in that: The step of determining the slip angle corresponding to each wheel according to the wheel speed pulses corresponding to each of the four wheels in the wheel speed sensor data comprises: Get the target vehicle's driving direction at the current moment; Determine the wheel speed corresponding to each of the wheels according to the wheel speed sensor data; Determining the linear speed corresponding to each of the wheels based on the wheel speed corresponding to each of the wheels; The slip angle corresponding to each of the wheels is determined according to the driving direction and the linear speed corresponding to each of the wheels.

3. The method according to claim 1, characterized in that The method of determining the second relative position and the second posture change of the target vehicle at the current moment relative to the previous moment according to the wheel speed sensor data and the slip angle corresponding to each of the wheels through the wheel speed pulse pre-integration measurement model includes: Acquire historical wheel speed sensor data of the target vehicle at the previous moment; Determine the wheel speed increment corresponding to each wheel according to the wheel speed sensor data and the historical wheel speed sensor data; Correcting the wheel speed increment corresponding to each wheel using the slip angle corresponding to each wheel to obtain the target wheel speed increment corresponding to each wheel; The wheel speed pulse pre-integration measurement model is used to determine a second relative position and a second posture change of the target vehicle at a current moment relative to a previous moment according to the target wheel speed increment.

4. The method according to claim 1, characterized in that: The method of using an inertial measurement unit pre-integration model to determine a first relative position and a first posture change of the target vehicle at a current moment relative to a previous moment according to inertial sensor data includes: Determining the linear acceleration of the target vehicle at a current moment according to the accelerometer data in the inertial sensor data; Determining the angular velocity of the target vehicle at the current moment according to the gyroscope data in the inertial sensor data; The inertial measurement unit pre-integration model is used to determine a first relative position and a first posture change of the vehicle at a current moment relative to a previous moment according to the linear acceleration and the angular velocity.

5. The method according to claim 1, characterized in that The method of determining the target position and target posture of the target vehicle at the current moment according to the first relative position, the first posture change, the second relative position, and the second posture change by using an invariant Kalman filter model includes: According to the first relative position and the first posture change, determining the predicted position and predicted posture of the target vehicle at the current moment through the state transfer equation in the invariant Kalman filter model; determining a constant Kalman gain according to the predicted position, the predicted attitude, the second relative position, and the second attitude change; The predicted position and the predicted attitude are updated by using the invariant Kalman gain to obtain the target position and the target attitude of the target vehicle at the current moment.

6. A vehicle positioning device, characterized in that: The device comprises: An acquisition module is used to acquire the inertial sensor data and wheel speed sensor data of the target vehicle at the current moment; A slip angle determination module, for determining the slip angle corresponding to each wheel according to the wheel speed pulses corresponding to each of the four wheels in the wheel speed sensor data based on the Transformer wheel speed slip angle estimation model; A first determination module is used to determine a first relative position and a first posture change of the target vehicle at a current moment relative to a previous moment according to inertial sensor data by using an inertial measurement unit pre-integration model; A second determination module is used to determine a second relative position and a second posture change of the target vehicle at a current moment relative to a previous moment according to wheel speed sensor data and slip angles corresponding to each of the wheels through a wheel speed pulse pre-integration measurement model; The vehicle positioning module is used to determine the target position and target posture of the target vehicle at the current moment according to the first relative position, the first posture change, the second relative position and the second posture change through an invariant Kalman filter model.

7. The device according to claim 6, characterized in that The slip angle determination module comprises: A driving direction acquisition unit, used to acquire the driving direction of the target vehicle at the current moment; A wheel speed determination unit, used to determine the wheel speed corresponding to each of the wheels according to the wheel speed sensor data; A linear speed determining unit, configured to determine a linear speed corresponding to each of the wheels based on a wheel speed corresponding to each of the wheels; The slip angle determination unit is used to determine the slip angle corresponding to each of the wheels according to the driving direction and the linear speed corresponding to each of the wheels.

8. The device according to claim 6, characterized in that The second determining module includes: A historical data acquisition unit, used to acquire historical wheel speed sensor data of the target vehicle at the previous moment; A wheel speed increment acquisition unit, used to determine the wheel speed increment corresponding to each wheel according to the wheel speed sensor data and the historical wheel speed sensor data; a target wheel speed increment determining unit, configured to correct the wheel speed increment corresponding to each wheel by using the slip angle corresponding to each wheel, so as to obtain the target wheel speed increment corresponding to each wheel; The second determination unit is used to determine a second relative position and a second posture change of the target vehicle at a current moment relative to a previous moment according to the target wheel speed increment through the wheel speed pulse pre-integration measurement model.

9. A computer device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the vehicle positioning method according to any one of claims 1 to 5 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, and when the instructions are executed on a terminal device, the terminal device executes the vehicle positioning method according to any one of claims 1 to 5.

Images