Vehicle state parameter estimation method and device

Abstract

The application provides a vehicle state parameter estimation method and device, and belongs to the technical field of automobile control. The method comprises the following steps: obtaining driving state data, a first process state x and a first process covariance Q of a vehicle. A second process covariance Q(k) and a first measurement covariance R(k) are determined according to the driving state data of the vehicle. A first measurement value y of a vehicle sensor is obtained h . A vehicle state parameter of the vehicle is determined according to the first measurement value y h , the first process state x, the first process covariance Q, the second process covariance Q(k) and the first measurement covariance R(k). In the estimation process of the vehicle state parameter, the process covariance and the measurement covariance are adaptively adjusted based on the driving state data, and then the adjusted process covariance and the measurement covariance are used for vehicle state estimation, so that the estimation accuracy of the vehicle state parameter can be improved.

Description

Technical Field

[0001] This application relates to the field of automotive control technology, and in particular to a method and apparatus for estimating vehicle state parameters. Background Technology

[0002] In recent years, with the development of intelligent vehicles, vehicle control systems, including Electronic Stability Program (ESP), Anti-lock Braking System (ABS), and Traction Control System (TCS), have been increasingly widely used in vehicles. Currently, to achieve automatic vehicle control, it is typically necessary to collect various vehicle status data to activate these control systems. Generally, this data can be collected by installing various sensors on the vehicle. However, some vehicle status data, including sideslip angle, usually require additional, expensive sensors for measurement. Considering the need to control the manufacturing costs of mass-produced vehicles, there is a growing tendency to estimate other parameters for vehicle control, such as sideslip angle, yaw rate, and longitudinal speed, using signals collected from existing onboard sensors. However, current methods for estimating vehicle status parameters have low accuracy and cannot meet the high-precision requirements of vehicle control. Summary of the Invention

[0003] This application provides a vehicle state parameter estimation method and a vehicle state parameter estimation device, which can improve the estimation accuracy of vehicle state parameters.

[0004] Firstly, this application provides a method for estimating vehicle state parameters, the method comprising:

[0005] Acquire the vehicle's driving status data, the first process state x, and the first process covariance Q;

[0006] The second process covariance Q(k) and the first measurement covariance R(k) are determined based on the vehicle's driving status data.

[0007] Acquire the first measurement value y from the vehicle sensor h ;

[0008] According to the first measured value y h The vehicle state parameters of the vehicle are determined by the first process state x, the first process covariance Q, the second process covariance Q(k), and the first measurement covariance R(k).

[0009] In the process of estimating vehicle state parameters, this application adaptively adjusts the process covariance and measurement covariance based on driving state data, and then uses the adjusted process covariance and measurement covariance for vehicle state estimation, which can improve the estimation accuracy of vehicle state parameters.

[0010] In one possible implementation, the driving state data includes one or more of the following: lateral acceleration change rate. Steering wheel speed lateral acceleration a y Road surface adhesion coefficient μ, and wheel speed ω of the front left wheel FL The wheel speed ω of the front right wheel FR The wheel speed ω of the rear left wheel RL Or the wheel speed ω of the rear right wheel RR .

[0011] In one possible implementation, determining the second process covariance Q(k) based on the vehicle's driving state data includes:

[0012] When the road surface adhesion coefficient μ is greater than or equal to the first preset road surface adhesion coefficient threshold μ TH And the absolute value of the rate of change of the lateral acceleration Greater than the preset lateral acceleration change rate threshold And / or, the absolute value of the steering wheel rotation speed. Greater than the preset steering wheel speed threshold When, obtain the first preset process covariance matrix as the second process covariance Q(k); or

[0013] When the road surface adhesion coefficient μ is greater than or equal to the first preset road surface adhesion coefficient threshold μ TH And the absolute value of the rate of change of the lateral acceleration Less than or equal to the preset lateral acceleration change rate threshold and the absolute value of the steering wheel rotation speed. Less than or equal to the preset steering wheel speed threshold At that time, according to the lateral acceleration a y Determine the covariance Q(k) of the second process; or

[0014] When the road surface adhesion coefficient μ is less than the first preset road surface adhesion coefficient threshold μ TH And the absolute value of the rate of change of the lateral acceleration Greater than the preset lateral acceleration change rate threshold And / or, the absolute value of the steering wheel rotation speed. Greater than the preset steering wheel speed threshold When, obtain the second preset process covariance matrix as the second process covariance Q(k); or

[0015] When the road surface adhesion coefficient μ is less than the first preset road surface adhesion coefficient threshold μ TH And the absolute value of the rate of change of the lateral acceleration Less than or equal to the preset lateral acceleration change rate threshold and the absolute value of the steering wheel rotation speed. Less than or equal to the preset steering wheel speed threshold At that time, the third preset process covariance matrix is ​​obtained as the second process covariance Q(k).

[0016] In one possible implementation, determining the first measurement covariance R(k) based on the vehicle's driving state data includes:

[0017] When the maximum absolute value of the wheel speed difference between different wheels is max|Δω i | Greater than the preset wheel speed difference threshold Δω TH And / or, the minimum absolute value of the wheel speed of different wheels, min|ω i | Less than the preset wheel speed threshold ω TH When the first preset measurement covariance matrix is ​​obtained, it is used as the first measurement covariance R(k);

[0018] When the maximum absolute value of the wheel speed difference between different wheels is max|Δω i | Less than or equal to the preset wheel speed difference threshold Δω TH And the minimum absolute value of the wheel speed of different wheels, min|ω i | Greater than or equal to the preset wheel speed threshold ω TH At that time, the first measurement covariance R(k) is determined based on the road surface adhesion coefficient μ.

[0019] In one possible implementation, determining the first measurement covariance R(k) based on the road surface adhesion coefficient μ includes:

[0020] When the road surface adhesion coefficient μ is greater than or equal to the second preset road surface adhesion coefficient threshold μ′ TH When the time is right, the second preset measurement covariance matrix is ​​obtained as the first measurement covariance R(k);

[0021] When the road surface adhesion coefficient μ is less than the second preset road surface adhesion coefficient threshold μ′ TH When the time comes, the third preset measurement covariance matrix is ​​obtained as the first measurement covariance R(k).

[0022] In one possible implementation, the first measurement value y of the vehicle sensor...h It includes measurements obtained from the vehicle's actual sensors and measurements acquired from virtual vehicle sensors, which are obtained through a neural network.

[0023] This application introduces the results obtained by other estimation methods (such as kinematic state estimation, neural network-based state estimation, and vision-based state estimation) as the measurement values ​​of the "virtual sensor," which expands the measurement values ​​obtained based on real sensor measurements and the calculated measurement covariance. The expanded measurement values ​​and measurement covariance are then used for vehicle state estimation, which can further improve the accuracy of vehicle state estimation.

[0024] In one possible implementation, the first process state x includes one or more of the following:

[0025] Longitudinal velocity v x Lateral velocity v y Yaw rate r, front left wheel speed ω FL The wheel speed ω of the front right wheel FR The wheel speed ω of the rear left wheel RL The wheel speed ω of the rear right wheel RR Or the road surface adhesion coefficient μ.

[0026] In one possible implementation, the first measured value y h Includes one or more of the following:

[0027] Longitudinal acceleration a x lateral acceleration a y Yaw rate r, front left wheel speed ω FL The wheel speed ω of the front right wheel FR The wheel speed ω of the rear left wheel RL The wheel speed ω of the rear right wheel RR Or the centroid side deflection angle β.

[0028] In one possible implementation, the vehicle state parameters include one or more of the following:

[0029] Longitudinal velocity v x Lateral velocity v y Yaw rate r, front left wheel speed ω FL The wheel speed ω of the front right wheel FR The wheel speed ω of the rear left wheel RL The wheel speed ω of the rear right wheel RR Road surface adhesion coefficient μ, longitudinal acceleration a x lateral acceleration a y , centroid side deflection angle β.

[0030] Secondly, this application provides a vehicle state parameter estimation device, the device comprising:

[0031] The transceiver unit is used to acquire the vehicle's driving status data, the first process state x, and the first process covariance Q;

[0032] The processing unit is used to determine the second process covariance Q(k) and the first measurement covariance R(k) based on the driving state data of the vehicle.

[0033] The transceiver unit is used to acquire the first measurement value y from the vehicle sensor. h ;

[0034] The processing unit is configured to process the first measured value y. h The vehicle state parameters of the vehicle are determined by the first process state x, the first process covariance Q, the second process covariance Q(k), and the first measurement covariance R(k).

[0035] In one possible implementation, the driving state data includes one or more of the following: lateral acceleration change rate. Steering wheel speed lateral acceleration a y Road surface adhesion coefficient μ, and wheel speed ω of the front left wheel FL The wheel speed ω of the front right wheel FR The wheel speed ω of the rear left wheel RL Or the wheel speed ω of the rear right wheel RR .

[0036] In one possible implementation, the processing unit is used for:

[0037] When the road surface adhesion coefficient μ is greater than or equal to the first preset road surface adhesion coefficient threshold μ TH And the absolute value of the rate of change of the lateral acceleration Greater than the preset lateral acceleration change rate threshold And / or, the absolute value of the steering wheel rotation speed. Greater than the preset steering wheel speed threshold When, obtain the first preset process covariance matrix as the second process covariance Q(k); or

[0038] When the road surface adhesion coefficient μ is greater than or equal to the first preset road surface adhesion coefficient threshold μ TH And the absolute value of the rate of change of the lateral acceleration Less than or equal to the preset lateral acceleration change rate threshold and the absolute value of the steering wheel rotation speed. Less than or equal to the preset steering wheel speed threshold At that time, according to the lateral acceleration a y Determine the covariance Q(k) of the second process; or

[0039] When the road surface adhesion coefficient μ is less than the first preset road surface adhesion coefficient threshold μ TH And the absolute value of the rate of change of the lateral acceleration Greater than the preset lateral acceleration change rate threshold And / or, the absolute value of the steering wheel rotation speed. Greater than the preset steering wheel speed threshold When, obtain the second preset process covariance matrix as the second process covariance Q(k); or

[0040] When the road surface adhesion coefficient μ is less than the first preset road surface adhesion coefficient threshold μ TH And the absolute value of the rate of change of the lateral acceleration Less than or equal to the preset lateral acceleration change rate threshold and the absolute value of the steering wheel rotation speed. Less than or equal to the preset steering wheel speed threshold At that time, the third preset process covariance matrix is ​​obtained as the second process covariance Q(k).

[0041] In one possible implementation, the processing unit is used for:

[0042] When the maximum absolute value of the wheel speed difference between different wheels is max|Δω i | Greater than the preset wheel speed difference threshold Δω TH And / or, the minimum absolute value of the wheel speed of different wheels, min|ω i | Less than the preset wheel speed threshold ω TH When the first preset measurement covariance matrix is ​​obtained, it is used as the first measurement covariance R(k);

[0043] When the maximum absolute value of the wheel speed difference between different wheels is max|Δω i | Less than or equal to the preset wheel speed difference threshold Δω TH And the minimum absolute value of the wheel speed of different wheels, min|ω i | Greater than or equal to the preset wheel speed threshold ω TH At that time, the first measurement covariance R(k) is determined based on the road surface adhesion coefficient μ.

[0044] In one possible implementation, the processing unit is used for:

[0045] When the road surface adhesion coefficient μ is greater than or equal to the second preset road surface adhesion coefficient threshold μ′ THWhen the time is right, the second preset measurement covariance matrix is ​​obtained as the first measurement covariance R(k);

[0046] When the road surface adhesion coefficient μ is less than the second preset road surface adhesion coefficient threshold μ′ TH When the time comes, the third preset measurement covariance matrix is ​​obtained as the first measurement covariance R(k).

[0047] In one possible implementation, the first measurement value y of the vehicle sensor... h It includes measurements obtained from the vehicle's actual sensors and measurements acquired from virtual vehicle sensors, which are obtained through a neural network.

[0048] In one possible implementation, the first process state x includes one or more of the following:

[0049] Longitudinal velocity v x Lateral velocity v y Yaw rate r, front left wheel speed ω FL The wheel speed ω of the front right wheel FR The wheel speed ω of the rear left wheel RL The wheel speed ω of the rear right wheel RR Or the road surface adhesion coefficient μ.

[0050] In one possible implementation, the first measured value y h Includes one or more of the following:

[0051] Longitudinal acceleration a x lateral acceleration a y Yaw rate r, front left wheel speed ω FL The wheel speed ω of the front right wheel FR The wheel speed ω of the rear left wheel RL The wheel speed ω of the rear right wheel RR Or the centroid side deflection angle β.

[0052] In one possible implementation, the vehicle state parameters include one or more of the following:

[0053] Longitudinal velocity v x Lateral velocity v y Yaw rate r, front left wheel speed ω FL The wheel speed ω of the front right wheel FR The wheel speed ω of the rear left wheel RL The wheel speed ω of the rear right wheel RR Road surface adhesion coefficient μ, longitudinal acceleration a x lateral acceleration a y , centroid side deflection angle β.

[0054] Thirdly, this application provides a vehicle state parameter estimation device. This device can be a terminal device, a device within a terminal device, or a device compatible with a terminal device. The vehicle state parameter estimation device can also be a chip system. This vehicle state parameter estimation device can execute the method described in the first aspect. The function of the vehicle state parameter estimation device can be implemented in hardware or by hardware executing corresponding software. The hardware or software includes one or more units or modules corresponding to the above functions. The unit or module can be software and / or hardware. The operation and beneficial effects performed by the vehicle state parameter estimation device can be found in the method described in the first aspect and its beneficial effects; repetitions will not be repeated.

[0055] Fourthly, this application provides a vehicle state parameter estimation device, which may be a terminal device. The vehicle state parameter estimation device includes a processor and a transceiver. The processor and the transceiver are used to execute at least one computer program or instructions stored in a memory to enable the device to implement the method as described in any of the first aspects.

[0056] Fifthly, this application provides a vehicle state parameter estimation device, which can be a terminal device. The vehicle state parameter estimation device includes a processor, a transceiver, and a memory. The processor, transceiver, and memory are coupled; the processor and transceiver are used to implement the method as described in any of the first aspects.

[0057] In a sixth aspect, this application provides a computer-readable storage medium storing a computer program or instructions that, when executed by a computer, implement the method as described in any of the first aspects.

[0058] In a seventh aspect, this application provides a computer program product including instructions, the computer program product including computer program code, which, when run on a computer, implements the method of any one of the first aspects.

[0059] Eighthly, a chip system is provided, including a processor and potentially a memory, for implementing methods of any of the first aspects and any possible designs. The chip system may be composed of chips or may include chips and other discrete devices. Attached Figure Description

[0060] Figure 1 This is a schematic diagram of vehicle state estimation based on wheel speed sensors provided in an embodiment of this application;

[0061] Figure 2 This is a schematic diagram of vehicle state estimation based on multiple sensors provided in an embodiment of this application;

[0062] Figure 3 This is a schematic diagram of the architecture of the vehicle state parameter estimation system provided in the embodiments of this application;

[0063] Figure 4 This is a flowchart illustrating a vehicle state parameter estimation method provided in an embodiment of this application;

[0064] Figure 5 This is a schematic diagram illustrating the determination of the second process covariance Q(k) based on driving state data, provided in an embodiment of this application.

[0065] Figure 6 This is a schematic diagram illustrating the determination of the first measurement covariance R(k) based on driving state data, provided in an embodiment of this application.

[0066] Figure 7 This is a flowchart illustrating another vehicle state parameter estimation method provided in an embodiment of this application;

[0067] Figure 8 This is a schematic diagram of the structure of a vehicle state parameter estimation device provided in an embodiment of this application;

[0068] Figure 9 This is a schematic diagram of another vehicle state parameter estimation device provided in an embodiment of this application;

[0069] Figure 10 This is a schematic diagram of the structure of a chip provided in an embodiment of this application. Detailed Implementation

[0070] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings.

[0071] In the description of this application, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. The "and / or" in the embodiments of this application is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Furthermore, "at least one" refers to one or more, and "multiple" refers to two or more. The terms "first," "second," etc., do not limit the quantity or execution order, and the terms "first," "second," etc., do not necessarily imply that they are different.

[0072] In this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or illustration. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0073] To facilitate understanding, some terms used in the embodiments of this application will be explained below.

[0074] 1. Unscented Kalman Filter (UKF): UKF is an alternative approach to solving nonlinear Kalman filtering problems. It utilizes unscented transformation to address the issue of nonlinear transformations of probability distributions. Unlike extended Kalman filtering, unscented Kalman filtering does not require calculating the Jacobian matrix, and with roughly the same computational cost, it can achieve more accurate nonlinear processing results.

[0075] 2. Inertial Measurement Unit (IMU): A device that measures the three-axis attitude angles and acceleration of an object. It generally includes three single-axis accelerometers and three single-axis gyroscopes.

[0076] 3. Steering angle sensor (SAS): Used to measure the rotation angle of the steering wheel when the vehicle is turning. It is mainly installed in the steering column below the steering wheel.

[0077] 4. Wheel speed sensor (WSS): A sensor used to measure the rotational speed of car wheels. Commonly used wheel speed sensors include magnetoelectric wheel speed sensors and Hall effect wheel speed sensors.

[0078] 5. Master cylinder pressure sensor (MPS): A sensor used to measure the pressure inside the master cylinder.

[0079] 6. Advanced Driver Assistance System (ADAS): Utilizing various sensors installed on the vehicle (millimeter-wave radar, lidar, cameras, and satellite navigation), it continuously senses the surrounding environment while the car is in motion, collects data, identifies, detects, and tracks static and dynamic objects, and combines this data with navigation map data to perform system calculations and analysis. This allows the driver to anticipate potential dangers, effectively increasing driving comfort and safety.

[0080] The system architecture and business scenarios of the embodiments of this application are described below. It should be noted that the system architecture and business scenarios described in this application are for the purpose of more clearly illustrating the technical solutions of this application, and do not constitute a limitation on the technical solutions provided in this application. As those skilled in the art will know, with the evolution of system architecture and the emergence of new business scenarios, the technical solutions provided in this application are also applicable to similar technical problems.

[0081] It should be noted that high-precision vehicle state parameter estimation (VSE) is a crucial prerequisite for vehicle dynamics control and autonomous driving motion control. With the widespread adoption of ADAS and artificial intelligence (AI) technologies, sensors such as vision, inertial navigation, radar, and lidar are increasingly used in vehicles, thus providing new opportunities for VSE. Generally, increasing the number of sensors can reduce the number of VSE operations, such as for center of gravity sideslip angle estimation and road adhesion coefficient estimation. For an example, please refer to [link to relevant documentation]. Figure 1 , Figure 1 This is a schematic diagram of vehicle state estimation based on wheel speed sensors provided in an embodiment of this application. Figure 1 As shown, when only wheel speed sensors (WSS) are available, the vehicle's center of gravity sideslip angle and road adhesion coefficient require four estimations to be obtained indirectly, and the estimation accuracy decreases with each estimation because each estimation introduces a certain estimation error. Specifically, the wheel speeds of the front left wheel, front right wheel, rear left wheel, and rear right wheel can be measured based on the WSS, and denoted as ω. FL ,ω FR ,ω RL ,ω RR For ease of description, the WSS measurement value is referred to as such. In the first estimation, longitudinal vehicle speed and longitudinal acceleration can be estimated based on the WSS measurement value. In the second estimation, slip ratio and yaw rate can be estimated based on the longitudinal vehicle speed; lateral acceleration can be estimated based on the longitudinal vehicle speed and the WSS measurement value; and master cylinder pressure can be estimated based on the longitudinal acceleration. In the third estimation, lateral vehicle speed is estimated based on yaw rate and lateral acceleration; vertical force and lateral force are estimated based on lateral acceleration; and longitudinal force is estimated based on lateral acceleration and master cylinder pressure. In the fourth estimation, sideslip angle is estimated based on the longitudinal vehicle speed obtained from the first estimation and the lateral vehicle speed obtained from the third estimation; and road adhesion coefficient is estimated based on the vertical force, lateral force, and longitudinal force obtained from the third estimation.

[0082] For example, please see Figure 2 , Figure 2 This is a schematic diagram of vehicle state estimation based on multiple sensors provided in an embodiment of this application. Figure 2As shown, when the wheel speed sensor is used in conjunction with the steering wheel angle sensor (SAS), inertial measurement unit (IMU), and master cylinder pressure sensor (MPS), the vehicle's sideslip angle and road adhesion coefficient need to be obtained through two estimations. Reducing the number of estimations improves the estimation accuracy of the sideslip angle and road adhesion coefficient. Specifically, the wheel speeds of the front left wheel, front right wheel, rear left wheel, and rear right wheel can be obtained based on the WSS measurement, denoted as ω. FL ,ω FR ,ω RL ,ω RR For ease of description, these are referred to as WSS measurements. Parameters such as steering wheel angle are obtained based on SAS measurements; for ease of description, these are referred to as SAS measurements. Parameters such as longitudinal acceleration, lateral acceleration, and yaw rate are obtained based on IMU measurements; for ease of description, these are referred to as IMU measurements. Parameters such as master cylinder pressure are obtained based on MPS measurements; for ease of description, these are referred to as MPS measurements. In the first estimation, the longitudinal vehicle speed can be estimated based on WSS measurements, the lateral vehicle speed based on SAS and IMU measurements, the road surface inclination angle, slope, vertical force, and lateral force based on IMU measurements, and the longitudinal force based on MPS measurements. In the second estimation, the slip ratio can be calculated based on the longitudinal vehicle speed obtained in the first estimation, the sideslip angle can be calculated based on the longitudinal vehicle speed and road surface inclination angle, and the road surface adhesion coefficient can be estimated based on the vertical force, lateral force, and longitudinal force.

[0083] The vehicle state parameter estimation system provided in the embodiments of this application is further described below.

[0084] For example, please see Figure 3 , Figure 3 This is a schematic diagram of the architecture of the vehicle state parameter estimation system provided in an embodiment of this application. Figure 3 As shown, the vehicle state parameter estimation system mainly comprises three modules: ① sensor measurement module, ② driving state adaptation module, and ③ UKF vehicle state estimation module. The ① sensor measurement module includes both real sensor measurements and virtual sensor measurements. These measurements serve as inputs to the UKF vehicle state estimation module, primarily for its posterior estimation submodule. Real sensor measurements mainly include information such as vehicle acceleration (e.g., lateral and longitudinal acceleration) and yaw rate measured by the inertial measurement unit (IMU), and wheel speeds of the four wheels measured by wheel speed sensors (WSS), etc., which are not limited here. Virtual sensor measurements mainly utilize the results of other state estimation methods as virtual measurement values ​​input to the UKF vehicle state estimation module. These mainly include kinematic-based state estimation results, neural network-based state estimation results, and vision-based state estimation results, etc., which are not limited here.

[0085] ② The adaptive driving state module acquires vehicle motion state information (such as acceleration, yaw rate, wheel speed, steering wheel angle, driving torque, and braking torque) and relevant environmental information (such as road adhesion coefficient) to perform driving state characteristic analysis. Based on the results of the driving state characteristic analysis, it determines adaptive strategies for process covariance and measurement covariance, i.e., adaptively adjusts the process covariance and measurement covariance. This module can address the dependence of UKF on the accuracy of the dynamic model, improving the estimation and fusion accuracy of the dynamic UKF.

[0086] ③ The UKF vehicle state estimation module uses an unscented Kalman filter to comprehensively estimate vehicle states, such as vehicle speed, center of gravity sideslip angle, tire force, slip ratio, and tire sideslip angle. This module mainly includes the following five sub-modules: Sigma points generation, unscented Sigma points transformation, prior estimation, posterior estimation, and output model. The functions of these five sub-modules are described below. Figure 4 The descriptions of each step in the process shown are not detailed here.

[0087] Please see Figure 4 , Figure 4 This is a flowchart illustrating a vehicle state parameter estimation method provided in an embodiment of this application. The method can be implemented using a vehicle state parameter estimation device and includes at least the following steps S401 to S404:

[0088] S401. Obtain the vehicle's driving status data, the first process state x, and the first process covariance Q.

[0089] In some feasible implementations, if vehicle state parameters need to be estimated, the vehicle's driving state data, first process state x, and first process covariance Q can first be obtained. The vehicle driving state data may include environmental information during vehicle operation and vehicle motion state information. For example, environmental information may include the road surface adhesion coefficient μ, and vehicle motion state information may include acceleration (e.g., lateral acceleration a). y Longitudinal acceleration a x ), yaw rate r, wheel speed (e.g., the wheel speed ω of the front left wheel of a vehicle), FL The wheel speed ω of the front right wheel FR The wheel speed ω of the rear left wheel RL The wheel speed ω of the rear right wheel RR Steering wheel speed The driving torque and braking torque are not limited here. It should be noted that in the embodiments of this application, the derivative of the yaw rate r is equal to the yaw acceleration. For ease of description, the yaw acceleration can be expressed as... The derivative of the steering wheel angle δ obtained from SAS measurements is equal to the steering wheel rotation speed. For ease of description, the steering wheel rotation speed can be expressed as... Among them, yaw acceleration With longitudinal velocity v x The product of these two factors equals the rate of change of lateral acceleration. For ease of description, the rate of change of lateral acceleration can be expressed as follows in this application: Where V = v x Therefore, the driving state data involved in the embodiments of this application may include one or more of the following parameters: lateral acceleration rate of change. Steering wheel speed lateral acceleration a y Road surface adhesion coefficient μ, and wheel speed ω of the front left wheel FL The wheel speed ω of the front right wheel FR The wheel speed ω of the rear left wheel RL Or the wheel speed ω of the rear right wheel RR .

[0090] The first process state x may include one or more of the following parameters: longitudinal velocity v x Lateral velocity v y Yaw rate r, front left wheel speed ω FL The wheel speed ω of the front right wheel FR The wheel speed ω of the rear left wheel RL The wheel speed ω of the rear right wheel RR Or the road surface adhesion coefficient μ, etc., are not limited here. That is to say, the first process state in the embodiments of this application can be defined as the following matrix x:

[0091] x = [v x ,v y ,r,ω FL ,ω FR ,ω RL ,ω RR ,μ]

[0092] Accordingly, the covariance Q(k) of the first process can be defined as a diagonal matrix Q1. Right now:

[0093]

[0094] in, The process covariance coefficient represents the longitudinal velocity. q represents the process covariance coefficient corresponding to the lateral velocity. r The process covariance coefficient represents the yaw rate. The process covariance coefficient representing the wheel speed of the front left wheel. The process covariance coefficient representing the wheel speed of the front right wheel. The process covariance coefficient representing the wheel speed of the rear left wheel. The process covariance coefficient q represents the wheel speed of the front right wheel. μ This represents the process covariance coefficient corresponding to the road surface adhesion coefficient.

[0095] S402. Determine the second process covariance Q(k) and the first measurement covariance R(k) based on the vehicle's driving status data.

[0096] In some feasible implementations, the second process covariance Q(k) and the first measurement covariance R(k) are determined based on the vehicle's driving state data. Specifically, determining the second process covariance Q(k) based on the vehicle's driving state data can be understood as: when the road surface adhesion coefficient μ is greater than or equal to the first preset road surface adhesion coefficient threshold μ... TH And the absolute value of the rate of change of lateral acceleration Greater than the preset lateral acceleration change rate threshold And / or, the absolute value of steering wheel rotation speed Greater than the preset steering wheel speed threshold When the first preset process covariance matrix is ​​obtained, it is used as the second process covariance Q(k). Alternatively, when the road surface adhesion coefficient μ is greater than or equal to the first preset road surface adhesion coefficient threshold μ TH And the absolute value of the rate of change of lateral acceleration Less than or equal to the preset lateral acceleration rate of change threshold and the absolute value of steering wheel speed Less than or equal to the preset steering wheel speed threshold At that time, according to the lateral acceleration a y Determine the covariance Q(k) of the second process. Alternatively, when the road surface adhesion coefficient μ is less than the first preset road surface adhesion coefficient threshold μ... TH And the absolute value of the rate of change of lateral acceleration Greater than the preset lateral acceleration change rate threshold And / or, the absolute value of steering wheel rotation speed Greater than the preset steering wheel speed threshold When the second preset process covariance matrix is ​​obtained, it is used as the second process covariance Q(k). Alternatively, when the road surface adhesion coefficient μ is less than the first preset road surface adhesion coefficient threshold μ TH And the absolute value of the rate of change of lateral acceleration Less than or equal to the preset lateral acceleration rate of change threshold and the absolute value of steering wheel speed Less than or equal to the preset steering wheel speed threshold At that time, the third preset process covariance matrix is ​​obtained as the second process covariance Q(k).

[0097] For example, please see Figure 5 , Figure 5 This is a schematic diagram illustrating the determination of the second process covariance Q(k) based on driving state data, provided in an embodiment of this application. For example... Figure 5 As shown, when the vehicle is traveling on the road surface, the road surface adhesion coefficient μ is greater than or equal to a given road surface adhesion coefficient threshold (i.e., the first preset road surface adhesion coefficient threshold μ). TH When on a road surface, if the absolute value of the rate of change of lateral acceleration is... The lateral acceleration rate of change is greater than a given threshold (i.e., a preset threshold for the lateral acceleration rate of change). ), and / or, the absolute value of steering wheel speed. The steering wheel speed is greater than a given threshold (i.e., the preset steering wheel speed threshold). If the second process covariance Q(k) can be taken as the first preset process covariance matrix, then the second process covariance Q(k) can be taken as the first preset process covariance matrix. For example, the first preset process covariance matrix can be the optimal process covariance matrix under transient conditions. For example, as shown in the figure... Figure 5 In State0, the first preset process covariance matrix q r =0.1215, (Right now If the absolute value of the rate of change of lateral acceleration Less than or equal to the preset lateral acceleration rate of change threshold And the absolute value of the steering wheel speed Less than or equal to the preset steering wheel speed threshold Then it can be further determined based on the vehicle's lateral acceleration. y The absolute value range is used to further adaptively adjust the covariance Q(k) of the second process, as follows: Figure 8 State1 through State5 in the context of [the system]. Specifically, as shown below... Figure 5 In State1, when |a y When the entry condition is |≤2m / s2, the covariance Q(k) of the second process takes the form Q1(k). For example, in Q1(k) q r =0.6197, (Right now When |a y When the exit condition is |> 2.5 m / s², the second process covariance Q(k) no longer takes the value Q1(k). That is, during the period when the second process covariance Q(k) takes the value Q1(k), when |a y |in (2, 2.5m / s 2 When the range fluctuates, Q(k) can still be equal to Q1(k) because during the actual operation of the vehicle, the lateral acceleration a of the vehicle... yIt may fluctuate continuously; therefore, to avoid frequent changes in the covariance Q(k) of the second process, it is necessary to set, for example... Figure 5 The exit condition in the text is such that during the period when Q(k) takes Q1(k), when |a y |>2.5m / s 2 When the exit condition is met, Q(k) will no longer take Q1(k). For example... Figure 5 In State2, when 2 < |a y ≤4m / s 2 When the entry condition is met, the covariance Q(k) of the second process takes the value Q2(k). For example, in Q2(k)... q r =1.305, (Right now When |a y |≤2 or|a y |>4.5m / s 2 When the exit condition is met, the covariance Q(k) of the second process no longer takes the value Q2(k). For example... Figure 5 In State3, when 4 < |a y ≤6m / s 2 When the entry condition is met, the covariance Q(k) of the second process takes the value Q3(k). For example, in Q3(k)... q r =0.8629, (Right now When |a y |≤4 or|a y |>6.5m / s 2 When the exit condition is met, the covariance Q(k) of the second process no longer takes the value Q3(k). For example... Figure 5 In State4, when 6 < |a y ≤8m / s 2 When the entry condition is met, the covariance Q(k) of the second process takes the value Q4(k). For example, in Q4(k)... q r =1.491, (Right now When |a y ≤6m / s 2 Or |a y |>8.5m / s 2 When the exit condition is met, the covariance Q(k) of the second process no longer takes the value Q4(k). For example... Figure 5 State5 in the context of |a y |>8m / s 2When the entry condition is met, the covariance Q(k) of the second process takes Q5(k). For example, in Q5(k)... q r =3.013, (Right now When |a y ≤8m / s 2 When the exit condition is met, the covariance Q(k) of the second process no longer takes the value Q5(k).

[0098] like Figure 5 When the vehicle is driving on a road surface, the coefficient of friction μ is less than the first preset road surface friction coefficient threshold μ. TH When on the road surface, if the absolute value of the rate of change of lateral acceleration is... Greater than the preset lateral acceleration change rate threshold And / or, the absolute value of steering wheel rotation speed Greater than the preset steering wheel speed threshold The second process covariance Q(k) can then be taken as the second preset process covariance matrix. The first preset process covariance matrix and the second preset process covariance matrix can be the same or different; this is not restricted here. For example, as shown... Figure 5 In State6, the second preset process covariance matrix q r =0.1215, (Right now If the absolute value of the rate of change of lateral acceleration Less than or equal to the preset lateral acceleration rate of change threshold And the absolute value of the steering wheel speed Less than or equal to the preset steering wheel speed threshold Then the covariance Q(k) of the second process is the covariance matrix of the second preset process, such as Figure 5 In State7, the third preset process covariance matrix q r =29.994, (Right now ).

[0099] It should be noted that, as Figure 5 q in state0 to state7 μ The values ​​of q can be the same or different. μ It can be a preset value, or it can be a value calculated based on the road surface adhesion coefficient of the actual road surface on which the vehicle is driving; there are no restrictions here.

[0100] In some feasible implementations, the determination of the first measurement covariance R(k) based on the vehicle's driving state data can be understood as: when the maximum absolute value of the wheel speed difference between different wheels is max|Δω i | Greater than the preset wheel speed difference threshold Δω TH And / or, the minimum absolute value of the wheel speed of different wheels, min|ω i | Less than the preset wheel speed threshold ω TH When the first preset measurement covariance matrix is ​​obtained as the first measurement covariance R(k); when the maximum value of the absolute value of the wheel speed difference between different wheels is max|Δω i | Less than or equal to the preset wheel speed difference threshold Δω TH And the minimum absolute value of the wheel speed of different wheels, min|ω i | Greater than or equal to the preset wheel speed threshold ω TH At that time, the first measurement covariance R(k) is determined based on the road surface adhesion coefficient μ. Here, determining the first measurement covariance R(k) based on the road surface adhesion coefficient μ can be understood as: when the road surface adhesion coefficient μ is greater than or equal to the second preset road surface adhesion coefficient threshold μ′... TH When the second preset measurement covariance matrix is ​​obtained as the first measurement covariance R(k); when the road surface adhesion coefficient μ is less than the second preset road surface adhesion coefficient threshold μ′ TH At that time, the third preset measurement covariance matrix is ​​obtained as the first measurement covariance R(k).

[0101] In one implementation, the first measurement covariance R(k) can be defined as a diagonal matrix R1. Right now:

[0102]

[0103] in, r r These represent the measurement covariance coefficients corresponding to longitudinal acceleration measurement, lateral acceleration measurement, and yaw rate measurement, respectively. These represent the measurement covariance coefficients corresponding to the wheel speed measurements of the front left wheel, front right wheel, rear left wheel, and rear right wheel, respectively.

[0104] Optionally, the first measurement covariance R(k) can also be defined as a diagonal matrix R². Right now:

[0105]

[0106] in, r rThese represent the measurement covariance coefficients corresponding to longitudinal acceleration, lateral acceleration, and yaw rate, respectively. These represent the measurement covariance coefficients corresponding to the wheel speeds of the front left wheel, front right wheel, rear left wheel, and rear right wheel, respectively. These represent the measurement covariance coefficients corresponding to the vehicle center of gravity sideslip angle obtained based on kinematic estimation, the measurement covariance coefficients corresponding to the vehicle center of gravity sideslip angle obtained based on neural network estimation, and the measurement covariance coefficients corresponding to the vehicle center of gravity sideslip angle obtained based on visual estimation, respectively.

[0107] For example, please see Figure 6 , Figure 6 This is a schematic diagram illustrating the determination of the first measurement covariance R(k) based on driving state data, provided in an embodiment of this application. Figure 6 As shown, the maximum value of the absolute value of the wheel speed difference between different wheels is max|Δω i | Greater than the given wheel speed difference threshold (i.e., the preset wheel speed threshold ω) TH ), and / or, the minimum absolute value of the wheel speed of different wheels, min|ω i | Less than a given wheel speed threshold (i.e., a preset wheel speed threshold ω) TH When ), the first measurement covariance R(k) is taken as follows: Figure 6 The first preset measurement covariance matrix corresponding to state0 is used; for example, the second preset measurement covariance matrix can be the optimal measurement covariance matrix under wheel lockup or slippage conditions; when the maximum absolute value of the wheel speed difference between different wheels is max|Δω i | Less than or equal to the preset wheel speed threshold ω TH And the minimum absolute value of the wheel speed of different wheels, min|ω i | Greater than or equal to the preset wheel speed threshold ω TH Then, the measured covariance is further adaptively adjusted based on the road adhesion coefficient: when the road adhesion coefficient μ is greater than the second preset road adhesion coefficient threshold μ′ TH When, the first measurement covariance R(k) is taken as follows Figure 6 The second preset measurement covariance matrix corresponding to state1, for example, can be the optimal process covariance matrix under high adhesion conditions; when the road adhesion coefficient μ is less than or equal to the second preset road adhesion coefficient threshold μ′ TH When, the first measurement covariance R(k) is taken as follows Figure 6 The third preset measurement covariance matrix corresponding to state2 can be, for example, the optimal process covariance matrix under low-load conditions.

[0108] S403, Obtain the first measurement value y from the vehicle sensor. h .

[0109] In some feasible implementations, the first measurement value y from the vehicle sensor is obtained. h The first measurement y from the vehicle's sensor. h This can include measurements obtained from the vehicle's actual sensors, for example, the first measurement value y. h This may include longitudinal acceleration a measured by an IMU. x lateral acceleration a y And the yaw rate r, and the wheel speed ω of the front left wheel measured by WSS. FL The wheel speed ω of the front right wheel FR The wheel speed ω of the rear left wheel RL The wheel speed ω of the rear right wheel RR etc., without restriction. That is to say, y h =[a x ,a y ,r,ω FL ,ω RR ,ω RL ,ω RR Accordingly, the first measurement covariance

[0110] Optional, the first measured value y h This can also include measurements obtained from the vehicle's actual sensors and measurements acquired from virtual sensors. The measurements from the virtual sensors can include those estimated via neural networks, and / or those obtained based on kinematic state estimation, and / or those obtained based on visual state estimation, etc., without limitation. For example, the vehicle's sideslip angle β can be obtained based on kinematic state estimation. Kinematic The vehicle's sideslip angle β is obtained based on the state estimation of the neural network. NN And the vehicle's center of gravity sideslip angle β is obtained based on vision-based state estimation. Camera Therefore, the first measured value y h =[a x ,a y ,r,ω FL ,ω FR ,ω RL ,ω RR ,β Kinematic ,β NN ,β Cmmera Accordingly, the first measurement covariance The specific values ​​are determined based on the actual scenario and are not limited here. For ease of understanding, the first measurement value involved in the embodiments of this application is illustrated by taking as an example that it includes the measurement value measured by the vehicle's real sensors and the measurement value obtained based on the vehicle's virtual sensors, and the first measurement covariance includes the measurement covariance coefficient corresponding to the measurement value of the vehicle's real sensors and the measurement covariance coefficient corresponding to the measurement value of the vehicle's virtual sensors.

[0111] It should be noted that this application extends the measured values ​​and measurement covariance of the vehicle's actual sensors, i.e., the first measured value y h In addition to the measurements obtained from the vehicle's real sensors, the first measurement covariance also includes the measurements obtained from the vehicle's virtual sensors. Accordingly, the first measurement covariance includes not only the measurement covariance coefficients corresponding to the measurements from the vehicle's real sensors, but also the measurement covariance coefficients corresponding to the measurements from the vehicle's virtual sensors, which can improve the estimation accuracy of vehicle state parameters.

[0112] S404, Based on the first measured value y h The vehicle state parameters are determined by the first process state x, the first process covariance Q, the second process covariance Q(k), and the first measurement covariance R(k).

[0113] In some feasible implementations, based on the first measured value y h The vehicle's state parameters can be determined by the first process state x, the first process covariance Q, the second process covariance Q(k), and the first measurement covariance R(k). Specifically, the above is based on the first measurement value y. h The vehicle state parameters are determined by the first process state x, the first process covariance Q, the second process covariance Q(k), and the first measurement covariance R(k). This can be understood as: determining the first key point data χ based on the first process state x and the first process covariance Q. i (k+1|k) and the second measured value γ i (k), first key point data χ i (k+1|k) and the second measured value γ i (k) all satisfy a Gaussian distribution; based on the first keypoint data χ i (k+1|k), the second measured value γ i (k), the second process covariance Q(k), the first measurement covariance R(k), and the first measurement value y h Determine the vehicle's status parameters. Specifically, the parameters are based on the first key point data χ. i (k+1|k), the second measured value γ i (k), second process covariance Q(k), first measurement covariance R(k), first measurement value y hThe vehicle state parameters are determined by the control variable state u(k), which can be understood as: based on the first key point data χ i (k+1|k) and the second process covariance Q(k) determine the prior process state. and prior process covariance According to the second measured value γ i The measurement estimate is determined by (k) and the first measurement covariance R(k). And measurement estimate covariance Based on the first key point data χ i (k+1|k), second measured value γ i (k) Prior process state Measurement estimate and measurement estimation of covariance Determine the Kalman feedback gain matrix K(k+1|k); based on the Kalman feedback gain matrix K(k+1|k) and the prior process covariance... Measurement Estimation Covariance Prior process state Measurement estimate and the first measured value y h (k) Determine the state of the posterior process and posterior process covariance Based on the posterior process state Determine the vehicle's status parameters.

[0114] For example, please see Figure 7 , Figure 7 This is a flowchart illustrating another vehicle state parameter estimation method provided in an embodiment of this application. Figure 7 As shown, the method mainly includes the following steps S701 to S705:

[0115] S701. Generate Sigma points based on the state of the first process and the covariance of the first process.

[0116] In one implementation, the first process state x and the first process covariance Q at time k are obtained, and then Sigma points can be generated based on the first process state x and the first process covariance Q. Specifically, this can be done based on the sampled values ​​of the first process state x at time k. Sampled values ​​of the first process covariance Q Generate Sigma points χ at time k i (k). Wherein, the generated Sigma points χ i (k) satisfies:

[0117]

[0118] Where, χ i (k) represents the Sigma points at time k. This represents the sampled value of the first process state x at time k. This represents the sampled value of the first process covariance Q at time k, where n represents the dimension of the first process state x (e.g., n = 8), and λ represents the preset coefficients used to calculate the weights. It should be noted that Sigma points are a series of representative points, including the mean point, extracted from the original Gaussian distribution (i.e., a Gaussian distribution based on the first process state mean and the first process covariance). These points are distributed around the first process state mean and represent the entire Gaussian distribution. Generally speaking, the more points extracted from the original Gaussian distribution, the more accurate the UKF approximation of the nonlinear model.

[0119] S702. Perform an unscented transformation on the generated Sigma points to obtain the first key point data and the second measurement value.

[0120] In one implementation, the first keypoint data χ can be obtained by performing a non-scented transformation on the generated Sigma points. i (k+1|k) and the second measured value γ i (k). Specifically, based on the 7-degree-of-freedom (DOF) nonlinear vehicle dynamics model and measurement model, the generated Sigma points χ can be respectively... i (k) Perform a non-scented transformation to obtain the new Sigma points predicted at time k+1, which are the first keypoint data χ. i (k+1|k), and according to χ i (k) Obtain the second measurement value γ i (k). It should be noted that the new Sigma points (i.e., the first keypoint data χ) obtained after the transformation i (k+1|k)) and the second measurement can also be approximated by a new Gaussian distribution. Specifically, the first keypoint data χ i (k+1|k) and the second measured value γ i (k) respectively satisfy:

[0121] χ i (k+1|k)=f(χ i (k),u(k))+w

[0122] γ i (k)=h(χ i (k),u(k))+v

[0123] Where f represents the state transition function, u(k) represents the state of the control variable, w represents the process noise, h represents the measurement function, and v represents the measurement noise.

[0124] S703. Determine the prior process state and prior process covariance based on the first key point data and the second process covariance; determine the measurement estimate and measurement estimate covariance based on the second measurement value and the first measurement covariance; and determine the Kalman feedback gain matrix based on the first key point data, the second measurement value, the prior process state, the measurement estimate, and the measurement estimate covariance.

[0125] In one implementation, firstly, the data can be based on the first key point data χ. i (k+1|k) and the second process covariance Q(k) determine the prior process state. and prior process covariance That is, based on the first key point data χ i (k+1|k) and the second process covariance Q(k) are used to perform prior estimation of the process state and process covariance, thus obtaining the prior process state. and prior process covariance Specifically, the state of the prior process and prior process covariance They respectively satisfy:

[0126]

[0127]

[0128]

[0129] in, This represents the prior process state, that is, the process state at time k+1 predicted at time k. W represents the prior process covariance, that is, the process covariance at time k+1 predicted at time k. i (m) This represents the weighting coefficient.

[0130] Then, based on the second measured value γ i The measurement estimate is determined by (k) and the first measurement covariance R(k). And measurement estimate covariance That is, based on the second measured value γ i (k) and the first measurement covariance R(k), for the measurement estimate and measurement estimation of covariance To make an estimate, specifically, to measure the estimated value. and measurement estimation of covariance They respectively satisfy:

[0131]

[0132]

[0133] Furthermore, according to γ i (k) and measurement estimates First key point data χ i (k+1|k) and prior process state Calculate the cross covariance matrix Specifically, the cross covariance matrix satisfy:

[0134]

[0135] Furthermore, the covariance is estimated based on the measurement. and cross covariance matrix The Kalman feedback gain matrix K(k+1|k) is calculated. Specifically, the Kalman feedback gain matrix K(k+1|k) satisfies:

[0136]

[0137] S704. Determine the posterior process state and posterior process covariance based on the Kalman feedback gain matrix, prior process covariance, measurement estimate covariance, prior process state, measurement estimate, and first measurement value.

[0138] In one implementation, the Kalman feedback gain matrix K(k+1|k) and the prior process covariance can be used as a basis. Measurement Estimation Covariance Prior process state Measurement estimate and the first measured value y h (k) Determine the state of the posterior process and posterior process covariance That is, based on the Kalman feedback gain matrix K(k+1|k) and the prior process covariance Measurement Estimation Covariance Prior process state Measurement estimate and the first measured value y h (k) performs posterior estimation of the process state and process covariance to obtain the posterior process state. and posterior process covariance Specifically, posterior process covariance and posterior process state They respectively satisfy:

[0139]

[0140]

[0141] It should be noted that the posterior process covariance That is, the predicted process covariance at time k+1, and the posterior process state. That is, the predicted process state at time k+1.

[0142] S705. Determine the vehicle state parameters based on the post-process state.

[0143] In one implementation, based on the posterior process state... Determine the vehicle state parameters at time k+1 Specifically, it can be based on the state of the posterior process. Using the control variable state u(k), a comprehensive estimate of the vehicle state parameters at time k+1 is performed:

[0144]

[0145] in, Let g represent the vehicle state parameters at time k+1, and g represent the output function.

[0146] It should be noted that the posterior process covariance and posterior process state It can be used as input for estimating the vehicle state parameters at time k+2, thus allowing the posterior process covariance to be calculated. As a sampled value of the new process covariance, the posterior process state The sampled values ​​of the new process state are used for vehicle state estimation at the next time step from the current time step. Alternatively, this embodiment of the application can be understood as estimating the vehicle state parameters at the current time step by using the vehicle state at the previous time step. In other words, the aforementioned steps S701 to S704 can be performed iteratively based on the new process state and process covariance generated by the posterior estimation.

[0147] In the estimation of vehicle state parameters, this application adaptively adjusts the process covariance and measurement covariance based on driving state data, and then uses the adjusted process covariance and measurement covariance for vehicle state estimation, thereby improving the estimation accuracy of vehicle state parameters. Furthermore, this application introduces the results obtained from other estimation methods (such as kinematic-based state estimation, neural network-based state estimation, and vision-based state estimation) as measurements from a "virtual sensor," expanding the measurements obtained from real sensors and the calculated measurement covariance. Using the expanded measurements and measurement covariance for vehicle state estimation further improves the estimation accuracy of vehicle state, which is beneficial for enhancing the performance of the vehicle dynamics control algorithm.

[0148] The following will combine Figures 8-10 The vehicle state parameter estimation device provided in this application is described in detail.

[0149] Please see Figure 8 , Figure 8 This is a schematic diagram of a vehicle state parameter estimation device provided in an embodiment of this application. Figure 8 The vehicle state parameter estimation device shown can be used to perform the above. Figures 4-7 The described method embodiments include some or all of the functions. The device can be a terminal, such as an in-vehicle terminal, a device within a terminal, or a device compatible with a terminal. The vehicle state parameter estimation device can also be a chip system. Figure 8 The vehicle state parameter estimation device shown may include a transceiver unit 801 and a processing unit 802. The processing unit 802 is used for data processing. The transceiver unit 801 integrates a receiving unit and a transmitting unit. The transceiver unit 801 can also be called a communication unit. Alternatively, the transceiver unit 801 can be split into a receiving unit and a transmitting unit. The processing unit 802 and the transceiver unit 801 described below are similar and will not be repeated here. Wherein:

[0150] The transceiver unit 801 is used to acquire the vehicle's driving status data, the first process state x, and the first process covariance Q;

[0151] Processing unit 802 is used to determine the second process covariance Q(k) and the first measurement covariance R(k) based on the driving state data of the vehicle;

[0152] The transceiver unit 801 is used to acquire the first measurement value y from the vehicle sensor. h ;

[0153] The processing unit 802 is configured to process the first measured value y. h The vehicle state parameters of the vehicle are determined by the first process state x, the first process covariance Q, the second process covariance Q(k), and the first measurement covariance R(k).

[0154] In one possible implementation, the driving state data includes one or more of the following: lateral acceleration change rate. Steering wheel speed lateral acceleration a y Road surface adhesion coefficient μ, and wheel speed ω of the front left wheel FL The wheel speed ω of the front right wheel FR The wheel speed ω of the rear left wheel RL Or the wheel speed ω of the rear right wheel RR .

[0155] In one possible implementation, the processing unit 802 is used for:

[0156] When the road surface adhesion coefficient μ is greater than or equal to the first preset road surface adhesion coefficient threshold μ TH And the absolute value of the rate of change of the lateral acceleration Greater than the preset lateral acceleration change rate threshold And / or, the absolute value of the steering wheel rotation speed. Greater than the preset steering wheel speed threshold When, obtain the first preset process covariance matrix as the second process covariance Q(k); or

[0157] When the road surface adhesion coefficient μ is greater than or equal to the first preset road surface adhesion coefficient threshold μ TH And the absolute value of the rate of change of the lateral acceleration Less than or equal to the preset lateral acceleration change rate threshold and the absolute value of the steering wheel rotation speed. Less than or equal to the preset steering wheel speed threshold At that time, according to the lateral acceleration a y Determine the covariance Q(k) of the second process; or

[0158] When the road surface adhesion coefficient μ is less than the first preset road surface adhesion coefficient threshold μ TH And the absolute value of the rate of change of the lateral acceleration Greater than the preset lateral acceleration change rate threshold And / or, the absolute value of the steering wheel rotation speed. Greater than the preset steering wheel speed threshold When, obtain the second preset process covariance matrix as the second process covariance Q(k); or

[0159] When the road surface adhesion coefficient μ is less than the first preset road surface adhesion coefficient threshold μ TH And the absolute value of the rate of change of the lateral acceleration Less than or equal to the preset lateral acceleration change rate threshold and the absolute value of the steering wheel rotation speed. Less than or equal to the preset steering wheel speed threshold At that time, the third preset process covariance matrix is ​​obtained as the second process covariance Q(k).

[0160] In one possible implementation, the processing unit 802 is used for:

[0161] When the maximum absolute value of the wheel speed difference between different wheels is max|Δω i| Greater than the preset wheel speed difference threshold Δω TH And / or, the minimum absolute value of the wheel speed of different wheels, min|ω i | Less than the preset wheel speed threshold ω TH When the first preset measurement covariance matrix is ​​obtained, it is used as the first measurement covariance R(k);

[0162] When the maximum absolute value of the wheel speed difference between different wheels is max|Δω i | Less than or equal to the preset wheel speed difference threshold Δω TH And the minimum absolute value of the wheel speed of different wheels, min|ω i | Greater than or equal to the preset wheel speed threshold ω TH At that time, the first measurement covariance R(k) is determined based on the road surface adhesion coefficient μ.

[0163] In one possible implementation, the processing unit 802 is used for:

[0164] When the road surface adhesion coefficient μ is greater than or equal to the second preset road surface adhesion coefficient threshold μ′ TH When the time is right, the second preset measurement covariance matrix is ​​obtained as the first measurement covariance R(k);

[0165] When the road surface adhesion coefficient μ is less than the second preset road surface adhesion coefficient threshold μ′ TH When the time comes, the third preset measurement covariance matrix is ​​obtained as the first measurement covariance R(k).

[0166] In one possible implementation, the first measurement value y of the vehicle sensor... h It includes measurements obtained from the vehicle's actual sensors and measurements acquired from virtual vehicle sensors, which are obtained through a neural network.

[0167] In one possible implementation, the first process state x includes one or more of the following:

[0168] Longitudinal velocity v x Lateral velocity v y Yaw rate r, front left wheel speed ω FL The wheel speed ω of the front right wheel FR The wheel speed ω of the rear left wheel RL The wheel speed ω of the rear right wheel RR Or the road surface adhesion coefficient μ.

[0169] In one possible implementation, the first measured value y h Includes one or more of the following:

[0170] Longitudinal acceleration a xlateral acceleration a y Yaw rate r, front left wheel speed ω FL The wheel speed ω of the front right wheel FR The wheel speed ω of the rear left wheel RL The wheel speed ω of the rear right wheel RR Or the centroid side deflection angle β.

[0171] In one possible implementation, the vehicle state parameters include one or more of the following:

[0172] Longitudinal velocity v x Lateral velocity v y Yaw rate r, front left wheel speed ω FL The wheel speed ω of the front right wheel FR The wheel speed ω of the rear left wheel RL The wheel speed ω of the rear right wheel RR Road surface adhesion coefficient μ, longitudinal acceleration a x lateral acceleration a y , centroid side deflection angle β.

[0173] Please see Figure 9 , Figure 9 This is a schematic diagram of another vehicle state parameter estimation device provided in an embodiment of this application. Figure 9 As shown, the vehicle state parameter estimation device includes a processor 901, a communication interface 902, and a memory 903. The processor 901, communication interface 902, and memory 903 are coupled via a bus 904.

[0174] Processor 901 can be one or more central processing units (CPUs). If processor 901 is a CPU, the CPU can be a single-core CPU or a multi-core CPU.

[0175] The processor 901 is used to read the program stored in the memory and, in cooperation with the communication interface 902, execute some or all of the steps of the method performed by the vehicle state parameter estimation device in the above embodiments of this application.

[0176] The memory 903 includes, but is not limited to, random access memory (RAM), erasable programmable read-only memory (EPROM), read-only memory (ROM), or compact disc read-only memory (CD-ROM), etc. The memory 903 is used to store programs, and the processor 901 can read the programs stored in the memory 903 to execute the programs described in the above embodiments of this application. Figures 4-7 The steps in the method shown will not be repeated here.

[0177] Please see Figure 10 , Figure 10 This is a schematic diagram of the structure of a chip provided in an embodiment of this application. Figure 10 As shown, chip 100 may include: a processor 1001, and one or more communication interfaces 1002 coupled to the processor 1001. Wherein:

[0178] Processor 1001 can be used to read and execute computer-readable instructions. In its implementation, processor 1001 mainly includes a controller, an arithmetic logic unit (ALU), and registers. The controller is primarily responsible for instruction decoding and issuing control signals for the operations corresponding to the instructions. The ALU is primarily responsible for performing fixed-point or floating-point arithmetic operations, shift operations, and logical operations, and can also perform address operations and translations. Registers are primarily responsible for storing register operands and intermediate operation results temporarily stored during instruction execution. In its implementation, the hardware architecture of processor 1001 can be an application-specific integrated circuit (ASIC) architecture, MIPS architecture, ARM architecture, or NP architecture, etc. Processor 1001 can be single-core or multi-core.

[0179] The communication interface 1002 can be used to input data to be processed into the processor 1001 and can output the processing results of the processor 1001. For example, the communication interface 1002 can be a general purpose input / output (GPIO) interface, which can be connected to multiple peripheral devices (such as displays (LCDs), cameras, radio frequency (RF) modules, etc.). The communication interface 1002 is connected to the processor 1001 via the bus 1003.

[0180] In this application, processor 1001 can be used to call the implementation program of the vehicle state parameter estimation method provided in one or more embodiments of this application from memory, and execute the instructions contained in the program. Communication interface 1002 can be used to output the execution result of processor 1001. Specifically, in this application, communication interface 1002 can be used to output the vehicle state parameter estimation result of processor 1001. For more information on the vehicle state parameter estimation method provided in one or more embodiments of this application, please refer to the foregoing. Figures 4-7 The various embodiments shown are not described in detail here.

[0181] It should be noted that the functions of the processor 1001 and the communication interface 1002 can be implemented through hardware design, software design, or a combination of hardware and software; no restrictions are imposed here.

[0182] Based on the same inventive concept, the principle and beneficial effects of the vehicle state parameter estimation device provided in the embodiments of this application are similar to the principle and beneficial effects of the vehicle state parameter estimation method in the embodiments of this application. Please refer to the principle and beneficial effects of the method implementation. Furthermore, the relationship between the various steps performed by each related module can also be referred to the description of the relevant content in the foregoing embodiments. For the sake of brevity, it will not be repeated here.

[0183] This application also provides a computer storage medium that can be used to store... Figures 4-7 The computer software instructions used by the vehicle state parameter estimation device in the illustrated embodiment include programs designed for executing the vehicle state parameter estimation device described above. The storage medium includes, but is not limited to, flash memory, hard disk, and solid-state drive.

[0184] This application also provides a computer program product that, when run by a vehicle state parameter estimation device, can execute the above-mentioned... Figures 4-7 The embodiment shown illustrates a vehicle state parameter estimation method designed for a vehicle state parameter estimation device.

[0185] This application also provides a sensor system for providing vehicle state parameter estimation functionality. It includes at least one vehicle state parameter estimation device mentioned in the above embodiments of this application, and at least one other sensor such as a camera or radar. The at least one sensor device within the system can be integrated into a single unit or device, or it can be independently configured as a component or device.

[0186] This application also provides a system for use in autonomous driving or intelligent driving, which includes at least one of the vehicle state parameter estimation device, camera, radar and other sensors mentioned in the above embodiments of this application. At least one device in the system can be integrated into a whole machine or device, or at least one device in the system can be set as an independent component or device.

[0187] Furthermore, any of the above systems can interact with the vehicle's central controller to provide information such as vehicle status parameters for vehicle driving decisions or control.

[0188] This application also provides a terminal, which includes at least one vehicle state parameter estimation device or any of the systems mentioned in the above embodiments of this application. For example, the terminal may include a vehicle, a camera, or a drone, etc., and is not limited thereto.

[0189] Understandably, the steps in the method embodiments of this application may be adjusted, combined, or deleted in order according to actual needs.

[0190] The modules in the device embodiments of this application can be merged, divided, and deleted according to actual needs.

[0191] Those skilled in the art will understand that, in the various embodiments of this application, the sequence number of each process does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0192] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A vehicle state parameter estimation method characterized by, The method includes: Acquire vehicle driving status data and first-process status. Covariance of the first process ; The second process covariance is determined based on the vehicle's driving status data. and the first measurement covariance ; Acquire the first measurement value from the vehicle's sensors. ; According to the first measurement value The first process state The first process covariance The second process covariance and the first measurement covariance Determine the vehicle status parameters of the vehicle; The driving state data includes one or more of the following: lateral acceleration change rate. Steering wheel speed lateral acceleration Road surface adhesion coefficient Wheel speed of the front left wheel Wheel speed of the front right wheel Wheel speed of the rear left wheel Or the speed of the rear right wheel The second process covariance is determined based on the vehicle's driving status data. ,include: When the road surface adhesion coefficient Greater than or equal to the first preset road surface adhesion coefficient threshold And the absolute value of the rate of change of the lateral acceleration Greater than the preset lateral acceleration change rate threshold And / or, the absolute value of the steering wheel rotation speed. Greater than the preset steering wheel speed threshold When the first preset process covariance matrix is ​​obtained, it is used as the second process covariance. ;or When the road surface adhesion coefficient Greater than or equal to the first preset road surface adhesion coefficient threshold And the absolute value of the rate of change of the lateral acceleration Less than or equal to the preset lateral acceleration change rate threshold and the absolute value of the steering wheel rotation speed. Less than or equal to the preset steering wheel speed threshold At that time, according to the lateral acceleration Determine the covariance of the second process ;or When the road surface adhesion coefficient Less than the first preset road surface adhesion coefficient threshold And the absolute value of the rate of change of the lateral acceleration Greater than the preset lateral acceleration change rate threshold And / or, the absolute value of the steering wheel rotation speed. Greater than the preset steering wheel speed threshold At that time, the second preset process covariance matrix is ​​obtained as the second process covariance. ;or When the road surface adhesion coefficient Less than the first preset road surface adhesion coefficient threshold And the absolute value of the rate of change of the lateral acceleration Less than or equal to the preset lateral acceleration change rate threshold and the absolute value of the steering wheel rotation speed. Less than or equal to the preset steering wheel speed threshold At that time, the third preset process covariance matrix is ​​obtained as the second process covariance. .

2. The method according to claim 1, characterized in that, The first measurement covariance is determined based on the vehicle's driving state data. ,include: When the maximum absolute value of the wheel speed difference between different wheels Greater than the preset wheel speed difference threshold And / or, the minimum absolute value of the wheel speeds of different wheels. Less than the preset wheel speed threshold When the first preset measurement covariance matrix is ​​obtained, it is used as the first measurement covariance. ; When the maximum absolute value of the wheel speed difference between different wheels Less than or equal to the preset wheel speed difference threshold And the minimum absolute value of the wheel speed of different wheels. Greater than or equal to the preset wheel speed threshold At that time, based on the road surface adhesion coefficient Determine the first measurement covariance .

3. The method according to claim 2, characterized in that, According to the road surface adhesion coefficient Determine the first measurement covariance ,include: When the road surface adhesion coefficient Greater than or equal to the second preset road surface adhesion coefficient threshold When the first measurement covariance is obtained, a second preset measurement covariance matrix is ​​acquired as the first measurement covariance matrix. ; When the road surface adhesion coefficient Less than the second preset road surface adhesion coefficient threshold At that time, the third preset measurement covariance matrix is ​​obtained as the first measurement covariance. .

4. The method according to any one of claims 1-3, characterized in that, The first measurement value of the vehicle sensor It includes measurements obtained from the vehicle's actual sensors and measurements acquired from virtual vehicle sensors, which are obtained through a neural network.

5. The method according to any one of claims 1-3, characterized in that, First process state Includes one or more of the following: Longitudinal velocity lateral velocity yaw rate Wheel speed of the front left wheel Wheel speed of the front right wheel Wheel speed of the rear left wheel Wheel speed of the rear right wheel or road surface adhesion coefficient .

6. The method according to any one of claims 1-3, characterized in that, First measurement value Includes one or more of the following: longitudinal acceleration lateral acceleration yaw rate Wheel speed of the front left wheel Wheel speed of the front right wheel Wheel speed of the rear left wheel Wheel speed of the rear right wheel or centroid side deflection angle .

7. The method according to any one of claims 1-3, characterized in that, The vehicle status parameters include one or more of the following: Longitudinal velocity lateral velocity yaw rate Wheel speed of the front left wheel Wheel speed of the front right wheel Wheel speed of the rear left wheel Wheel speed of the rear right wheel Road surface adhesion coefficient Longitudinal acceleration lateral acceleration , centroid side slip angle .

8. A vehicle state parameter estimation device, characterized in that, The device includes: The transceiver unit is used to acquire vehicle driving status data and first-process status. Covariance of the first process ; The processing unit is used to determine the second process covariance based on the vehicle's driving state data. and the first measurement covariance ; The transceiver unit is used to acquire the first measurement value from the vehicle sensor. ; The processing unit is configured to process the first measured value. The first process state The first process covariance The second process covariance and the first measurement covariance Determine the vehicle status parameters of the vehicle; The driving state data includes one or more of the following: lateral acceleration change rate. Steering wheel speed lateral acceleration Road surface adhesion coefficient Wheel speed of the front left wheel Wheel speed of the front right wheel Wheel speed of the rear left wheel Or the speed of the rear right wheel The second process covariance is determined based on the vehicle's driving status data. At that time, the processing unit is used for: When the road surface adhesion coefficient Greater than or equal to the first preset road surface adhesion coefficient threshold And the absolute value of the rate of change of the lateral acceleration Greater than the preset lateral acceleration change rate threshold And / or, the absolute value of the steering wheel rotation speed. Greater than the preset steering wheel speed threshold When the first preset process covariance matrix is ​​obtained, it is used as the second process covariance. ;or When the road surface adhesion coefficient Greater than or equal to the first preset road surface adhesion coefficient threshold And the absolute value of the rate of change of the lateral acceleration Less than or equal to the preset lateral acceleration change rate threshold and the absolute value of the steering wheel rotation speed. Less than or equal to the preset steering wheel speed threshold At that time, according to the lateral acceleration Determine the covariance of the second process ;or When the road surface adhesion coefficient Less than the first preset road surface adhesion coefficient threshold And the absolute value of the rate of change of the lateral acceleration Greater than the preset lateral acceleration change rate threshold And / or, the absolute value of the steering wheel rotation speed. Greater than the preset steering wheel speed threshold At that time, the second preset process covariance matrix is ​​obtained as the second process covariance. ;or When the road surface adhesion coefficient Less than the first preset road surface adhesion coefficient threshold And the absolute value of the rate of change of the lateral acceleration Less than or equal to the preset lateral acceleration change rate threshold and the absolute value of the steering wheel rotation speed. Less than or equal to the preset steering wheel speed threshold At that time, the third preset process covariance matrix is ​​obtained as the second process covariance. .

9. The apparatus according to claim 8, characterized in that, The processing unit is used for: When the maximum absolute value of the wheel speed difference between different wheels Greater than the preset wheel speed difference threshold And / or, the minimum absolute value of the wheel speeds of different wheels. Less than the preset wheel speed threshold When the first preset measurement covariance matrix is ​​obtained, it is used as the first measurement covariance. ; When the maximum absolute value of the wheel speed difference between different wheels Less than or equal to the preset wheel speed difference threshold And the minimum absolute value of the wheel speed of different wheels. Greater than or equal to the preset wheel speed threshold At that time, based on the road surface adhesion coefficient Determine the first measurement covariance .

10. The apparatus according to claim 9, characterized in that, The processing unit is used for: When the road surface adhesion coefficient Greater than or equal to the second preset road surface adhesion coefficient threshold When the first measurement covariance is obtained, a second preset measurement covariance matrix is ​​acquired as the first measurement covariance matrix. ; When the road surface adhesion coefficient Less than the second preset road surface adhesion coefficient threshold At that time, the third preset measurement covariance matrix is ​​obtained as the first measurement covariance. .

11. The apparatus according to any one of claims 8-10, characterized in that, The first measurement value of the vehicle sensor It includes measurements obtained from the vehicle's actual sensors and measurements acquired from virtual vehicle sensors, which are obtained through a neural network.

12. The apparatus according to any one of claims 8-10, characterized in that, First process state Includes one or more of the following: Longitudinal velocity lateral velocity yaw rate Wheel speed of the front left wheel Wheel speed of the front right wheel Wheel speed of the rear left wheel Wheel speed of the rear right wheel or road surface adhesion coefficient .

13. The apparatus according to any one of claims 8-10, characterized in that, First measurement value Includes one or more of the following: longitudinal acceleration lateral acceleration yaw rate Wheel speed of the front left wheel Wheel speed of the front right wheel Wheel speed of the rear left wheel Wheel speed of the right wheel or centroid side deflection angle .

14. The apparatus according to any one of claims 8-10, characterized in that, The vehicle status parameters include one or more of the following: Longitudinal velocity lateral velocity yaw rate Wheel speed of the front left wheel Wheel speed of the front right wheel Wheel speed of the rear left wheel Wheel speed of the rear right wheel Road surface adhesion coefficient Longitudinal acceleration lateral acceleration , centroid side slip angle .

15. A vehicle state parameter estimation device, characterized in that, include: Processor and memory; The memory is used to store one or more programs, the one or more programs including computer-executable instructions, wherein when the device is running, the processor executes the one or more programs stored in the memory to cause the device to perform the method as described in any one of claims 1-7.

16. A computer storage medium, characterized in that, The computer storage medium stores a computer program that, when run on the computer, causes the computer to perform the method as described in any one of claims 1-7.

17. A chip, characterized in that, The chip includes: A processor and a communication interface, the processor being configured to invoke and execute instructions from the communication interface, wherein when the processor executes the instructions, it implements the method as described in any one of claims 1-7.

18. A terminal comprising the means as described in any one of claims 8-14.

Images