A longitudinal vehicle speed control method, device, equipment and storage medium

Abstract

The application discloses a longitudinal vehicle speed control method, device, equipment and storage medium. The method comprises the following steps: determining a target slope angle of a road where a target vehicle is located according to the target vehicle's empty load, three-axis acceleration, three-axis angular velocity and actual vehicle speed; determining a target vehicle weight based on the target slope angle by using a preset longitudinal dynamics model and a preset recursive least square algorithm; the preset recursive least square algorithm contains a forgetting factor; determining an expected acceleration of the target vehicle by using a model predictive control algorithm based on a preset expected vehicle speed, the actual vehicle speed and actual acceleration; and determining a torque instruction based on the expected acceleration, the target vehicle weight and the target slope angle, so that a driving motor of the target vehicle controls the longitudinal vehicle speed according to the torque instruction, improves the accuracy of the vehicle longitudinal vehicle speed control and guarantees the comfort of passengers.

Description

Technical Field

[0001] This application relates to the field of vehicle control technology, and in particular to a longitudinal vehicle speed control method, device, equipment and storage medium. Background Technology

[0002] With the rapid development of autonomous driving technology for electric vehicles, longitudinal speed control, as a core component of the autonomous driving system, directly affects the safety, stability, and passenger comfort of the vehicle.

[0003] In existing technologies, the driving and braking torque of a vehicle are usually adjusted based on the difference between the desired speed and the actual speed. This not only results in low control precision and poor control effect, but also fails to guarantee passenger comfort. Summary of the Invention

[0004] In view of the above problems, this application provides a longitudinal speed control method, device, equipment and storage medium, with the aim of improving the accuracy of longitudinal speed control of vehicles and ensuring passenger comfort.

[0005] The embodiments of this application disclose the following technical solutions:

[0006] In a first aspect, this application provides a longitudinal vehicle speed control method, comprising:

[0007] Based on the target vehicle's unloaded mass, three-axis acceleration, three-axis angular velocity, and actual vehicle speed, the target slope angle of the road where the target vehicle is located is determined.

[0008] Based on the target slope angle, the target vehicle weight is determined using a preset longitudinal dynamics model and a preset recursive least squares algorithm; the preset recursive least squares algorithm includes a forgetting factor.

[0009] Based on the preset desired vehicle speed, the actual vehicle speed, and the actual acceleration, the desired acceleration of the target vehicle is determined using a model predictive control algorithm.

[0010] Based on the desired acceleration, the target vehicle weight, and the target slope angle, a torque command is determined so that the drive motor of the target vehicle controls the longitudinal speed according to the torque command.

[0011] Optionally, in the method described above, determining the target slope angle of the road where the target vehicle is located based on the target vehicle's unloaded mass, three-axis acceleration, three-axis angular velocity, and actual vehicle speed includes:

[0012] Based on the unloaded mass, triaxial acceleration, triaxial angular velocity and actual vehicle speed, the slope is calculated using a preset vehicle longitudinal dynamics model to obtain the first slope angle;

[0013] Based on the triaxial acceleration, triaxial angular velocity and actual vehicle speed, the pitch angle of the target vehicle is calculated using unit quaternions and unscented Kalman filtering algorithm.

[0014] Based on the pitch angle of the target vehicle, the second slope angle is determined using the relationship between the slope angle and the pitch angle;

[0015] The first slope angle and the second slope angle are fused using the Kalman filter method to determine the target slope angle.

[0016] Optionally, in the method described above, the step of calculating the pitch angle of the target vehicle based on the triaxial acceleration, triaxial angular velocity, and actual vehicle speed using a unit quaternion and unscented Kalman filter algorithm includes:

[0017] Based on the three-axis angular velocities of the target vehicle, the first-order Runge-Kutta method is used to solve the preset vehicle attitude change equation to obtain the real-time unit quaternion at each sampling time. The vehicle attitude change equation is used to characterize the relationship between the unit quaternion and the three-axis angular velocities, and the unit quaternion is used to represent the real-time attitude of the vehicle.

[0018] Based on the actual vehicle speed and the real-time unit quaternion at each sampling time, the predicted value of the three-axis acceleration is determined using a preset vehicle attitude estimation output equation; the vehicle attitude estimation output equation is used to characterize the relationship between the unit quaternion and the three-axis acceleration.

[0019] Based on the predicted triaxial acceleration and the triaxial acceleration, the attitude estimate of the target vehicle at the current moment is determined using an unscented Kalman filter algorithm.

[0020] The pitch angle of the target vehicle is determined based on the estimated attitude of the target vehicle at the current moment.

[0021] Optionally, in the method described above, the step of using Kalman filtering to fuse the first slope angle and the second slope angle to determine the target slope angle includes:

[0022] The initial parameters of the preset Kalman filter algorithm are predicted to determine the prior estimation result at the current time; the prior estimation result includes the prior slope angle and the prior error covariance matrix at the current time.

[0023] Using the first slope angle as the first observation, the prior estimation result is updated for the first time to obtain the first posterior estimation result;

[0024] The second slope angle is used as the second observation. The first posterior estimation result is updated by a second observation to obtain the second posterior estimation result. The slope angle value in the second posterior estimation result is used as the target slope angle.

[0025] Optionally, the method described above further includes:

[0026] Based on the target vehicle weight, three-axis acceleration, three-axis angular velocity, and actual vehicle speed, the slope is calculated using the vehicle longitudinal dynamics model to obtain the third slope angle; the third slope angle is the slope angle optimized from the first slope angle, and is used to perform the vehicle weight calculation step at the next moment.

[0027] Optionally, in the method described above, determining the desired acceleration of the target vehicle using a model predictive control algorithm based on a preset desired vehicle speed, the actual vehicle speed, and the actual acceleration includes:

[0028] Based on the preset expected vehicle speed, actual vehicle speed, and actual acceleration, the vehicle speed prediction value and acceleration prediction value at each sampling time in the prediction time domain are determined using a preset discrete state space model, and the predicted vehicle speed sequence and predicted acceleration sequence are obtained respectively; the discrete state space model is the discrete state space expression for longitudinal vehicle speed control.

[0029] For each sampling time, when it is determined that the predicted acceleration value at the sampling time in the predicted acceleration sequence is less than a first threshold and the impact is less than a second threshold, the predicted vehicle speed value corresponding to the sampling time in the predicted vehicle speed sequence is optimized, the expected vehicle acceleration increment at the sampling time is calculated, and the expected vehicle acceleration increment sequence is obtained; the impact is the rate of change of the predicted acceleration value per unit time.

[0030] The optimal solution for the desired acceleration increment sequence of the vehicle is obtained by using a preset cost function;

[0031] The expected acceleration of the target vehicle is determined based on the first expected acceleration increment in the expected acceleration increment sequence of the target vehicle; the first expected acceleration increment is the first item in the expected acceleration increment sequence of the target vehicle.

[0032] Secondly, this application provides a longitudinal vehicle speed control device, comprising:

[0033] The slope calculation module is used to determine the target slope angle of the road where the target vehicle is located based on the target vehicle's unloaded mass, three-axis acceleration, three-axis angular velocity and actual vehicle speed.

[0034] The vehicle weight calculation module is used to determine the target vehicle weight based on the target slope angle, using a preset longitudinal dynamics model and a preset recursive least squares algorithm; the preset recursive least squares algorithm includes a forgetting factor.

[0035] The desired acceleration determination module is used to determine the desired acceleration of the target vehicle based on a preset desired vehicle speed, the actual vehicle speed, and the actual acceleration, using a model predictive control algorithm.

[0036] The longitudinal speed control module is used to determine a torque command based on the desired acceleration, the target vehicle weight, and the target slope angle, so that the drive motor of the target vehicle controls the longitudinal speed according to the torque command.

[0037] Optionally, in the apparatus described above, the slope calculation module includes:

[0038] The first slope calculation unit is used to calculate the slope based on the unloaded mass, triaxial acceleration, triaxial angular velocity and actual vehicle speed using a preset vehicle longitudinal dynamics model, and obtain the first slope angle.

[0039] The second slope calculation unit is used to calculate the pitch angle of the target vehicle based on the triaxial acceleration, triaxial angular velocity and actual vehicle speed, using unit quaternions and unscented Kalman filtering algorithm; and to determine the second slope angle based on the pitch angle of the target vehicle using the relationship between the slope angle and the pitch angle.

[0040] The slope fusion unit is used to fuse the first slope angle and the second slope angle using the Kalman filter method to determine the target slope angle.

[0041] Thirdly, this application provides an electronic device, the device including: a processor, and a memory communicatively connected to the processor;

[0042] The memory stores the instructions that the computer executes;

[0043] The processor executes computer execution instructions stored in the memory to implement the longitudinal vehicle speed control method described in any of the above embodiments.

[0044] Fourthly, this application provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the longitudinal vehicle speed control method described in any of the above embodiments.

[0045] Compared with the prior art, this application has the following beneficial effects:

[0046] The method of this application determines the target slope angle of the road where the target vehicle is located based on the target vehicle's unloaded weight, three-axis acceleration, three-axis angular velocity, and actual vehicle speed, avoiding the limitations of traditional preset slope values ​​and adapting to changes in road terrain in real time. Then, based on the target slope angle, the target vehicle weight is determined using a preset longitudinal dynamics model and a preset recursive least squares algorithm. The preset recursive least squares algorithm includes a forgetting factor, which ensures the responsiveness of vehicle weight estimation to real-time driving conditions. The forgetting factor reduces interference from old data and improves the accuracy of in-vehicle estimation. Based on a preset expected vehicle speed, the actual vehicle speed, and... The actual acceleration is used to determine the desired acceleration of the target vehicle using a model predictive control algorithm. Based on the desired acceleration, the target vehicle weight, and the target slope angle, a torque command is determined so that the drive motor of the target vehicle can control the longitudinal speed according to the torque command. This enables the vehicle to track the desired speed while effectively constraining acceleration and rate of change, avoiding discomfort caused by sudden acceleration and deceleration. Furthermore, by fusing the desired acceleration, target vehicle weight, and slope angle to generate the torque command, the drive motor control is made more in line with the actual dynamic characteristics of the vehicle, significantly improving speed tracking accuracy, adaptability to complex road conditions, and passenger comfort. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 A flowchart illustrating a longitudinal vehicle speed control method provided in an embodiment of this application;

[0049] Figure 2 A schematic diagram illustrating the principle of the method for calculating the pitch angle of a target vehicle provided in this application embodiment;

[0050] Figure 3 A schematic diagram illustrating the principle of the target slope angle determination method provided in the embodiments of this application;

[0051] Figure 4 A flowchart illustrating the algorithm for determining the target vehicle weight provided in this application embodiment;

[0052] Figure 5 A flowchart of an algorithm for determining the desired acceleration using a model predictive control algorithm, provided in an embodiment of this application;

[0053] Figure 6 A schematic diagram of a longitudinal vehicle speed control device provided in an embodiment of this application;

[0054] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with specific embodiments and accompanying drawings. It should be particularly noted that the embodiments described in this application are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0056] It should be noted that, unless otherwise defined, the technical or scientific terms used in the embodiments of this application should have the ordinary meaning understood by one of ordinary skill in the art to which this application pertains. The terms "first," "second," and similar terms used in the embodiments of this application do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are only used to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0057] As described earlier, longitudinal speed control is crucial for the autonomous driving functions of electric vehicles. For example, when an electric vehicle in autonomous driving mode performs maneuvers such as overtaking, lane changing, cruise control, and deceleration to a stop, precise control of longitudinal speed is clearly necessary to ensure that the actual vehicle speed effectively tracks the desired speed. Furthermore, to guarantee passenger comfort, the vehicle's longitudinal acceleration and longitudinal impact must be within certain threshold ranges during autonomous driving, further increasing the requirements for longitudinal speed control capabilities.

[0058] In actual driving, vehicle weight and road gradient are key parameters affecting longitudinal dynamic characteristics. If the real-time changes of these two factors are ignored during control, and the driving torque is adjusted only based on the difference between the desired and actual vehicle speed, the speed tracking accuracy will decrease, and it will be difficult to constrain longitudinal acceleration and impact within a comfortable range. In existing technologies, some solutions do not accurately calculate vehicle weight, but only roughly assess the loading status, or rely on preset gradient values ​​instead of real-time estimation. The traditional control algorithms used also struggle to balance control accuracy and comfort requirements. Therefore, there is an urgent need for a method that can combine real-time estimation of vehicle weight and gradient and achieve precise and comfortable longitudinal speed control through advanced control strategies.

[0059] Through research, the inventors have proposed a longitudinal vehicle speed control method, device, equipment, and storage medium to achieve more precise control of longitudinal vehicle speed, improve control accuracy, and ensure passenger comfort.

[0060] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0061] See Figure 1 This figure is a schematic flowchart of a longitudinal vehicle speed control method provided in an embodiment of this application. The method includes:

[0062] S101: Determine the target slope angle of the road where the target vehicle is located based on the target vehicle's unloaded mass, three-axis acceleration, three-axis angular velocity, and actual vehicle speed.

[0063] In this embodiment, the triaxial acceleration and triaxial angular velocity of the target vehicle can be measured using an inertial measurement unit (IMU). The triaxial angular velocity can be measured separately using... , as well as Indicates; triaxial acceleration can be expressed as... , as well as The actual speed of the target vehicle can be measured using a wheel speed meter, and the actual speed can be represented by V; the empty weight of the target vehicle is the pre-set total vehicle weight, which can be represented by V. This indicates that, as an achievable method, the specific implementation steps for determining the target slope angle of the road where the target vehicle is located, based on the target vehicle's unloaded mass, three-axis acceleration, three-axis angular velocity, and actual vehicle speed, include S1011~S1014. The specific implementation steps of S1011~S1014 include:

[0064] S1011: Based on the unloaded mass, three-axis acceleration, three-axis angular velocity and actual vehicle speed, the slope is calculated using a preset vehicle longitudinal dynamics model to obtain the first slope angle.

[0065] In this embodiment, specifically, based on the unloaded mass Triaxial acceleration , as well as Triaxial angular velocity , as well as And the actual vehicle speed V, when calculating the first slope angle using the preset vehicle longitudinal dynamics model, first the unloaded mass... As the initial vehicle mass Combining the actual vehicle speed V and the longitudinal acceleration obtained by differentiating the actual vehicle speed V and the output torque of the target vehicle's drive motor. The data is substituted into the preset vehicle longitudinal dynamics model to obtain the first slope angle.

[0066] For the pre-defined longitudinal dynamics model of the vehicle, assuming that the target vehicle is driving under windless conditions, with no tire slippage, no transmission in the power system, and that all driving and braking torques are provided by the drive motor, and ignoring the vehicle's rotational mass, the longitudinal dynamics model of the target vehicle can be established as follows:

[0067] (1-1)

[0068] in, The output torque of the drive motor, The main reducer reduction ratio is given by η, the power system transmission efficiency is given by r, and r is the wheel radius. Let g be the vehicle mass, α be the road slope angle, and μ be the rolling friction coefficient. Let ρ be the longitudinal acceleration of the vehicle, and ρ be the air density. Where A is the air resistance coefficient, V is the frontal area, and V is the actual vehicle speed.

[0069] Based on equation (1-1), we can obtain:

[0070] (1-2)

[0071] Based on equation (1-2), and considering that the vehicle mass remains essentially constant during driving, we can initially assume that the vehicle mass is equal to the unloaded mass. Then, based on the collected vehicle status information of the target vehicle, we can complete a preliminary estimation of the road slope and obtain the first slope angle. The vehicle state information of the target vehicle includes multiple parameters used in the longitudinal dynamics model to calculate the slope angle.

[0072] S1012: Based on triaxial acceleration, triaxial angular velocity and actual vehicle speed, the pitch angle of the target vehicle is calculated using unit quaternions and unscented Kalman filtering algorithm.

[0073] In this embodiment, a vehicle attitude model is constructed using triaxial acceleration and triaxial angular velocity through unit quaternions. The attitude change equation is solved using the first-order Runge-Kutta method to obtain the real-time unit quaternions, and then the pitch angle of the target vehicle is derived.

[0074] As a feasible approach, based on triaxial acceleration, triaxial angular velocity, and actual vehicle speed, the specific steps for calculating the pitch angle of the target vehicle using unit quaternions and an unscented Kalman filter algorithm include:

[0075] Based on the three-axis angular velocities of the target vehicle, a first-order Runge-Kutta method is used to solve the preset vehicle attitude change equations, obtaining the real-time unit quaternions at each sampling time. The vehicle attitude change equations characterize the relationship between the unit quaternions and the three-axis angular velocities, with the unit quaternions representing the vehicle's real-time attitude. Then, based on the actual vehicle speed and the real-time unit quaternions at each sampling time, the predicted three-axis acceleration values ​​are determined using a preset vehicle attitude estimation output equation. This equation characterizes the relationship between the unit quaternions and the three-axis acceleration. Simultaneously, based on the predicted and actual three-axis acceleration values, an unscented Kalman filter algorithm is used to determine the target vehicle's attitude estimate at the current moment. Finally, based on the target vehicle's attitude estimate at the current moment, the target vehicle's pitch angle is determined.

[0076] In this embodiment, let the unit quaternion be... If used To represent the real-time attitude of the vehicle, we have:

[0077] (1-3)

[0078] In equation (1-3), ψ is the yaw angle, θ is the pitch angle, and ϕ is the roll angle. These three angles are the rotation angles from the vehicle coordinate system to the navigation coordinate system.

[0079] If used To describe the changes in vehicle attitude, we can obtain the vehicle attitude change equation:

[0080] (1-4)

[0081] In equation (1-4), The real-time attitude of the target vehicle; Let ω be the quaternion representation of the three-axle angular velocities of the target vehicle. .

[0082] To improve algorithm efficiency, a first-order Runge-Kutta method is used instead of a higher-order one to solve equation (1-4), which yields the real-time unit quaternion at each sampling time. :

[0083] (1-5)

[0084] In equation (1-5), k is the sampling time, and T is the sampling period. It is the identity matrix.

[0085] Triaxial acceleration measured by IMU , as well as Let the three-axis accelerations of the target vehicle in the vehicle coordinate system be denoted as . , as well as Let be the three-axis accelerations of the target vehicle in the navigation coordinate system. If only the longitudinal motion of the vehicle is considered, then:

[0086] (1-6)

[0087] Furthermore, based on the working principle of acceleration calculation in an IMU, we can know that:

[0088] (1-7)

[0089] In equation (1-7), The longitudinal acceleration of the target vehicle can be obtained by differentiating the actual vehicle speed V. Then, by simultaneously solving equations (1-6) and (1-7), the vehicle attitude estimation output equation can be obtained:

[0090] (1-8)

[0091] Furthermore, based on the actual vehicle speed V and the real-time unit quaternion at each sampling time... The system uses a pre-defined vehicle attitude estimation output equation to determine the predicted triaxial acceleration value corresponding to sampling time k. Then, based on the predicted triaxial acceleration and the triaxial acceleration, an unscented Kalman filter algorithm is used to determine the target vehicle's attitude estimate at the current time. Finally, based on the target vehicle's attitude estimate at the current time, the pitch angle of the target vehicle is determined. Specifically, if we let... This means that the unit quaternion representing the real-time attitude of the vehicle will be used. Let z be a state variable. This means that the three-axis acceleration of the target vehicle is used as the observation, and assuming... Initial error covariance matrix Process noise covariance Observation noise covariance ,in, and All are real numbers. The principle behind the pitch angle calculation method can be found in [reference needed]. Figure 2 , Figure 2 A schematic diagram illustrating the principle of the pitch angle calculation method for the target vehicle provided in this application embodiment. Figure 2 As shown, firstly, the triaxial angular velocities measured by the IMU are... , as well as go through After a delay, the vehicle attitude estimation state equation is input, and prior predictions are made on the current state variables in unit quaternion form to obtain the predicted state variable values. With error covariance matrix If sample points are generated, the state variables of the sample points can be expressed as follows: Simultaneously, the real-time vehicle speed V measured by the wheel speedometer is differentiated to obtain the longitudinal acceleration of the target vehicle. Input the vehicle attitude estimation output equation, make prior predictions on the current observations, and obtain the predicted values ​​of the observations. Subsequently, combined with triaxial acceleration , as well as The state posterior estimation is performed using Kalman filtering to obtain the estimated values ​​of the state variables. With error covariance matrix Regenerate sample points, and then... After normalization, the vehicle pitch angle θ is finally calculated.

[0092] In this embodiment, based on the three-axis angular velocities of the target vehicle, the first-order Runge-Kutta method is used to solve the vehicle attitude change equation to obtain the real-time unit quaternion. This method has high computational efficiency, can iteratively calculate the unit quaternion in a short time, and quickly track vehicle attitude changes, providing relatively real-time basic data for subsequent attitude estimation, meeting the need for rapid updates of attitude information during vehicle movement. Combining the actual vehicle speed and the real-time unit quaternion, the predicted three-axis acceleration value is determined using a preset vehicle attitude estimation output equation. This comprehensively considers the vehicle's motion state and attitude information, making the acceleration prediction value more consistent with the actual vehicle motion. Compared with relying solely on sensor measurements of acceleration, this increases the reliability and accuracy of the data. Then, using an unscented Kalman filter algorithm, the attitude estimate of the target vehicle at the current moment is determined based on the predicted three-axis acceleration value and the actual measured three-axis acceleration. The unscented Kalman filter algorithm can effectively handle noise interference in the system and has a good suppression effect on uncertainties such as sensor measurement errors, thereby improving the accuracy and stability of the attitude estimation value and reducing the error fluctuation of the attitude estimation. Finally, the pitch angle of the target vehicle is determined based on the attitude estimation value. Accurate pitch angle information can provide key road condition information for vehicle longitudinal speed control, helping the vehicle to better adapt to changes in road gradient and improve the safety, stability and passenger comfort of vehicle driving.

[0093] S1013: Based on the pitch angle of the target vehicle, determine the second slope angle using the relationship between the slope angle and the pitch angle.

[0094] In this embodiment, if the vehicle is considered a rigid body, then the road slope angle α is equal in magnitude to the defined vehicle pitch angle θ, but in opposite directions of rotation. Therefore:

[0095] (1-9)

[0096] Equation (1-9) is the relationship between the vehicle pitch angle and the road slope angle. Based on this equation, the second slope angle can be determined according to the calculated θ. .

[0097] S1014: Using the Kalman filter method, the first slope angle and the second slope angle are fused to determine the target slope angle.

[0098] Both vehicle longitudinal dynamics models and vehicle pitch attitude can be used to estimate road slope, but each has its own advantages and disadvantages. Specifically, the slope estimation method based on the vehicle longitudinal dynamics model has the advantages of simple modeling and fewer required parameters, but the disadvantages are limited estimation accuracy and the need for accurate vehicle weight values. The slope estimation method based on vehicle pitch attitude has the advantages of higher estimation accuracy, but the disadvantages are complex modeling, limited algorithm robustness, and susceptibility to external interference. Therefore, in this embodiment, the Kalman filter method is used to estimate the first slope angle. With the second slope angle The target slope angle is determined by merging the data.

[0099] As an feasible approach, the specific steps for determining the target slope angle by fusing the first and second slope angles using the Kalman filter method include: predicting the initial parameters of the preset Kalman filter algorithm to determine the prior estimation result at the current time; the prior estimation result includes the prior slope angle and the prior error covariance matrix at the current time; using the first slope angle as the first observation, performing a first observation update on the prior estimation result to obtain a first posterior estimation result; using the second slope angle as the second observation, performing a second observation update on the first posterior estimation result to obtain a second posterior estimation result, and using the slope angle value in the second posterior estimation result as the target slope angle.

[0100] Specifically, the principle of the target slope angle determination method can be found in [reference needed]. Figure 3 , Figure 3 This is a schematic diagram illustrating the principle of the target slope angle determination method for a target vehicle provided in an embodiment of this application. Figure 3 As shown, if the road slope estimation results based on the vehicle's longitudinal dynamics model and vehicle pitch attitude are respectively... as well as ,make Let the road slope angle α after fusion be used as a state variable, and let , That is, by taking the estimated first slope angle and the second slope angle as observations respectively, the target slope angle of the target vehicle is determined using a Kalman filter fusion dual-observation method, which includes: firstly, initialization is performed, setting the initial state variables. Initial error covariance matrix And the process noise covariance Q, and the covariance of the two observation noises and Entering the prediction phase, prior state estimation. Estimate from the posterior state of the previous time step We obtain the prior error covariance matrix. From the posterior error covariance matrix of the previous time step Add this to Q to obtain the result. Then, consider the first slope angle. Perform observation updates and calculate Kalman gain. Using observations Compared with prior state estimation The difference, combined with The first posterior state estimate is obtained. And update the error covariance matrix to Then, regarding the second slope angle... Perform observation updates, based on Calculate Kalman gain Using observations and The difference and This yields the final posterior state estimate. And error covariance matrix The fusion result is then output. Afterwards, the system waits for the next data point and repeats the "prediction-dual observation update" process to continuously determine the target slope angle.

[0101] In this embodiment, a priori estimation result is obtained by predicting the initial parameters, providing an initial state and error range reference for subsequent observation updates. This makes the entire filtering process more targeted and initially locks in the approximate range of the slope angle. Secondly, the first slope angle is used as the first observation for the first observation update. The information from this observation is used to correct the prior estimate, resulting in a first posterior estimation result that is closer to the true value. This effectively integrates the measurement information related to the first slope angle, improving the accuracy of the slope angle estimation. Finally, a second slope angle is introduced as the second observation for the second observation update. This further combines measurement data from another dimension to optimize the first posterior estimation result. The final slope angle value in the second posterior estimation result, i.e., the target slope angle, fully integrates the information from both observations, greatly improving the accuracy and reliability of the slope angle estimation. This provides accurate basic data for subsequent operations such as vehicle longitudinal speed control based on the slope angle, helping to improve the safety, stability, and comfort of vehicle driving.

[0102] In this embodiment, based on the target vehicle weight, three-axis acceleration, three-axis angular velocity, and actual vehicle speed, the slope is calculated using the vehicle's longitudinal dynamics model to obtain a third slope angle. This third slope angle is an optimized version of the first slope angle and is used to perform the vehicle weight calculation step at the next moment. Through this cyclical optimization mechanism of "slope calculation - vehicle weight calculation," vehicle weight calculation can be based on more accurate slope information. Subsequent slope calculations can then rely on a more accurate vehicle weight, continuously improving the estimation accuracy of these two key parameters: vehicle weight and slope. This provides more reliable data support for subsequent stages such as vehicle longitudinal speed control, ultimately contributing to improved vehicle driving safety, stability, and control precision.

[0103] In this embodiment, based on the unloaded mass, triaxial acceleration, triaxial angular velocity, and actual vehicle speed, a pre-set vehicle longitudinal dynamics model is used to calculate the slope, obtaining the first slope angle. This allows for an initial determination of the slope range based on the vehicle's basic parameters and dynamic characteristics, providing an initial basis for subsequent optimization that aligns with the vehicle's physical properties. Then, based on the triaxial acceleration, triaxial angular velocity, and actual vehicle speed, a unit quaternion and unscented Kalman filter algorithm are used to calculate the target vehicle's pitch angle. Based on the target vehicle's pitch angle, the relationship between the slope angle and the pitch angle is used to determine the second slope angle, enabling precise capture of vehicle attitude changes. This method effectively avoids errors caused by simplification of the dynamic model and supplements the shortcomings of pure dynamic calculation. Finally, the Kalman filter method is used to fuse the first slope angle and the second slope angle to determine the target slope angle. This method can integrate the physical rationality of the first slope angle and the attitude sensitivity of the second slope angle, and can also suppress noise interference in the two types of data through the filtering algorithm. The final output target slope angle takes into account the advantages of different calculation logics, with higher accuracy and stronger stability. It can provide accurate road condition data support for subsequent vehicle weight estimation, longitudinal vehicle speed control and other links, and help vehicles achieve safer and more stable driving control in complex slope scenarios.

[0104] S102: Based on the target slope angle, the target vehicle weight is determined using a preset longitudinal dynamics model and a preset recursive least squares algorithm.

[0105] The preset recursive least squares algorithm includes a forgetting factor.

[0106] In this embodiment, see Figure 4 , Figure 4 This is a schematic diagram illustrating the target vehicle weight determination method provided in an embodiment of this application. Figure 4 As shown, based on the longitudinal dynamics model of the target vehicle as shown in Equation (1-1), a real-time vehicle weight estimation algorithm can be completed by combining it with a recursive least squares algorithm with a forgetting factor. Specifically, from Equation (1-1), we can obtain:

[0107] (1-10)

[0108] make , to represent system output, Let represent the system input, then the system input and output at time k can be expressed as follows: as well as Assume the total vehicle weight Let the parameters to be estimated be defined as follows: the real-time estimated values ​​of the vehicle weight at time k and time k-1 are respectively... and Then we can get:

[0109] (1-11)

[0110] Equation (1-11) is the specific calculation formula for the real-time vehicle weight estimation algorithm. In the formula, This is the forgetting factor, which can be either a constant or a time-varying parameter, and its value ranges from 0 to 1. Since the vehicle's weight can be considered constant during the period from starting to stopping, therefore... The value should be close to 1; in this embodiment, it is taken as 0.99. Let be the Kalman gain, used to represent the state estimate at time k with respect to the new measurement y(k). The degree of correction; The error covariance matrix describes the state estimate at time k. The degree of uncertainty. Meanwhile, to reduce the computational burden, after the vehicle starts, the real-time estimate of the vehicle weight can be obtained as shown in the following formula. By sampling and analysis, we can obtain:

[0111] (1-12)

[0112] In equation (1-12), The number of sampling points is the preset number, where i represents the i-th sampling point. The preset real-time estimation result of vehicle weight The period for performing sampling analysis, T is Real-time computing cycle For the most recent indivual The average value of the sampling results, where RMSE is the root mean square error.

[0113] Let σ be the convergence threshold. If RMSE > σ, then the real-time vehicle weight estimation algorithm has not yet converged, and vehicle weight estimation needs to continue. In this case, to reduce the output value of the vehicle weight estimator... The deviation from the true value is set as follows: the real-time estimated vehicle weight at time k. If greater than the vehicle's unloaded weight Then let the vehicle weight estimator output value equal ;otherwise, It should be the vehicle's unloaded weight. .

[0114] Conversely, if RMSE ≤ σ, then the real-time vehicle weight estimation algorithm can be considered to have converged, and no further iterative calculations are needed. The estimated vehicle weight output by the estimator is... Should equal to And maintain that value until the car stops.

[0115] Furthermore, based on the above embodiments, the method may further include:

[0116] Based on the target vehicle weight, three-axis acceleration, three-axis angular velocity, and actual vehicle speed, the slope is calculated using the vehicle's longitudinal dynamics model to obtain the third slope angle. The third slope angle is the slope angle optimized from the first slope angle and is used to perform the vehicle weight calculation step at the next moment.

[0117] In this embodiment, when S102 outputs the actual vehicle weight Then, it can be input back into S101 to use the actual vehicle weight output by S102, and then combined with the three-axis acceleration, three-axis angular velocity and actual vehicle speed, to calculate the slope using the vehicle longitudinal dynamics model as shown in Equation (1-1) to obtain the third slope angle; then the target slope angle is recalculated using the third slope angle, and then the vehicle weight calculation step at the next moment is executed.

[0118] S103: Based on the preset desired vehicle speed, actual vehicle speed, and actual acceleration, the desired acceleration of the target vehicle is determined using a model predictive control algorithm.

[0119] In this embodiment, based on the expected vehicle speed at the next moment Actual vehicle speed V, longitudinal acceleration The desired acceleration of the target vehicle is determined using the Model Predictive Control (MPC) algorithm. This allows the actual vehicle speed V to smoothly and quickly follow the desired vehicle speed. Among them, the desired vehicle speed This decision was made by the planning and decision-making level for the target vehicle.

[0120] As an feasible approach, the specific steps for determining the desired acceleration of a target vehicle using a model predictive control algorithm, based on preset desired vehicle speed, actual vehicle speed, and actual acceleration, include: Based on the preset desired vehicle speed, actual vehicle speed, and actual acceleration, using a preset discrete state-space model, determining the predicted vehicle speed and acceleration values ​​at each sampling time point within the prediction time domain, and obtaining the predicted vehicle speed sequence and predicted acceleration sequence respectively; wherein, the discrete state-space model is the discrete state-space expression for longitudinal vehicle speed control. Then, for each sampling time point, when the predicted acceleration value at the sampling time point in the predicted acceleration sequence is less than a first threshold and the impact is less than a second threshold, optimizing the predicted vehicle speed value corresponding to the sampling time point in the predicted vehicle speed sequence, calculating the vehicle's desired acceleration increment at the sampling time point, and obtaining the vehicle's desired acceleration increment sequence; wherein, the impact is the rate of change of the predicted acceleration value per unit time. Finally, using a preset cost function, solving for the optimal solution of the vehicle's desired acceleration increment sequence, obtaining the target vehicle's desired acceleration increment sequence. Furthermore, the expected acceleration of the target vehicle is determined based on the first expected acceleration increment in the expected acceleration increment sequence of the target vehicle; wherein, the first expected acceleration increment is the first term in the expected acceleration increment sequence of the target vehicle.

[0121] In this embodiment, see Figure 5 This figure is a flowchart of the algorithm for determining the desired acceleration using a model predictive control algorithm, as provided in an embodiment of this application. Figure 5 As shown, the desired acceleration is calculated using the MPC algorithm. The specific principle is as follows:

[0122] The longitudinal speed control process of a vehicle can be described with reference to a first-order inertial system, and can be expressed by the following formula:

[0123] (1-13)

[0124] In equation (1-13) above, K=1 is the system gain, and τ is the system time constant. Based on this equation, the state-space expression for longitudinal vehicle speed control can be derived as follows:

[0125] (1-14)

[0126] In the above equation (1-14), , , , , , .

[0127] Combining this equation with the forward Euler method, the discrete state-space expression for longitudinal vehicle speed control can be obtained as follows:

[0128] (1-15)

[0129] In the above equation (1-15), , , , , , T is the sampling period of the MPC algorithm, and k is the current sampling time. This is the desired acceleration of the vehicle set by the system at time k.

[0130] To illustrate the effect of changes in vehicle acceleration on vehicle speed, based on the above equation (1-15), a new state-space expression can be constructed as follows:

[0131] (1-16)

[0132] in, , , , , , , This is the desired acceleration increment of the vehicle set by the system at time k.

[0133] If we use the above equation (1-16) to analyze the system from... arrive The state variables are derived. This can be called the prediction step size, and then we can obtain:

[0134] (1-17)

[0135] Similarly, the system can be analyzed using equation (1-16) above. arrive By deriving the output, we can obtain:

[0136] (1-18)

[0137] Based on the above equation (1-18), if we assume that the system is known from arrive All inputs, of which To control the step size, and 1≤ ≤ If the system from arrive By predicting the output, we can obtain:

[0138] (1-19)

[0139] In the above equation (1-19), It can be represented as:

[0140] (1-20)

[0141] In the above equation (1-19), It can be represented as:

[0142] (1-21)

[0143] In the above equation (1-19), It can be represented as:

[0144] (1-22)

[0145] In the above equation (1-19), It can be represented as:

[0146] (1-23)

[0147] In the above equation (1-19), It can be represented as:

[0148] (1-24)

[0149] From equation (1-19), it can be seen that, given the current actual vehicle speed V and longitudinal acceleration of the system... Under the premise of assuming the target vehicle's expected acceleration increment in the control time domain, It can predict the speed of the target vehicle within the prediction time domain. The duration of the control time domain is determined by... (Decision), the duration of the predicted time domain is determined by Decision. Clearly, to ensure the effectiveness of longitudinal speed control and passenger comfort, this predicted speed value should be as close as possible to the desired speed value, and the predicted vehicle acceleration and impact must be kept within the specified threshold range and minimized. Based on the above conditions, the predicted speed sequence... By performing optimized management, the optimal sequence of expected vehicle acceleration increments can be calculated. Specifically, the constructed cost function can be represented by the following expression:

[0150] (1-25)

[0151] In the above equation (1-25), and Let be the expected and predicted vehicle speeds at time k+i, respectively. The formula (1-18) can be used for derivation and calculation. Q represents the weight of the difference between the expected vehicle speed and the predicted vehicle speed. and Let R and S be the expected acceleration increment and expected acceleration of the vehicle at time k+i, respectively, and R and S be the weight values ​​of the two, respectively. and The relationship can be represented as:

[0152] (1-26)

[0153] also, and The following constraints must be met:

[0154] (1-27)

[0155] In the above equation (1-27), , , , The desired acceleration of the vehicle and the upper and lower limits of the desired acceleration increment are set to ensure passenger comfort. In this embodiment, The absolute value is 1.25 m / s 2 within, The absolute value is 17.64 m / s 3 Within.

[0156] For longitudinal vehicle speed control, each element in the optimal vehicle desired acceleration increment sequence must minimize the cost function value shown in equation (1-25) while satisfying the constraints shown in equation (1-27). This problem is transformed into a quadratic programming problem with inequality constraints, and the optimal vehicle desired acceleration increment sequence can then be solved. .and then, The specific value can be expressed as:

[0157] (1-28)

[0158] Will The first element Substituting into equation (1-26), the expected acceleration of the vehicle at time k can be calculated. When we reach time k+1, the expected vehicle speed at time k+1 is... Actual vehicle speed V, actual longitudinal acceleration Based on vehicle information, future vehicle speeds are re-predicted, and a new sequence of expected vehicle acceleration increments is obtained through optimized calculations. By repeating this process, the desired acceleration can be achieved. Real-time solution.

[0159] In this embodiment, based on a preset discrete state-space model, and combining the desired vehicle speed, actual vehicle speed, and actual acceleration, a predicted vehicle speed sequence and a predicted acceleration sequence are generated in the prediction time domain. This allows for the prediction of the vehicle's future motion trend, providing a forward-looking basis for subsequent control decisions and avoiding speed tracking deviations caused by real-time response lags. Furthermore, by setting acceleration and impact constraints, the predicted acceleration and impact at each sampling moment are double-checked. Only when comfort requirements are met are the predicted vehicle speed values ​​optimized and the desired acceleration increment calculated. This avoids situations that can easily cause passenger jerking, such as rapid acceleration and sudden changes in acceleration, significantly improving ride comfort. Finally, the current desired acceleration is determined based on the optimized desired acceleration increment sequence. This ensures that the vehicle speed converges smoothly to the desired speed, and the sequential incremental control ensures the smoothness of acceleration changes. This provides a precise and gentle control target for subsequent drive motor torque adjustment, ultimately helping the vehicle achieve precise speed control in complex driving scenarios while ensuring passenger comfort.

[0160] S104: Based on the desired acceleration, target vehicle weight, and target slope angle, determine the torque command so that the drive motor of the target vehicle controls the longitudinal speed according to the torque command.

[0161] In this embodiment, based on the desired acceleration Actual vehicle speed V and previously estimated target vehicle weight Based on data such as the target slope angle α, the torque command is determined. Output torque of the drive motor of the target vehicle Adjustments are made to achieve the desired acceleration. It was achieved.

[0162] In this embodiment, the target slope angle of the road where the target vehicle is located is determined based on the target vehicle's unloaded mass, three-axis acceleration, three-axis angular velocity, and actual vehicle speed. This avoids the limitations of traditional preset slope values ​​and can adapt to changes in road terrain in real time. Then, based on the target slope angle, the target vehicle weight is determined using a preset longitudinal dynamics model and a preset recursive least squares algorithm. The preset recursive least squares algorithm includes a forgetting factor, which ensures the responsiveness of vehicle weight estimation to real-time driving conditions. The forgetting factor reduces interference from old data and improves the accuracy of in-vehicle estimation. Based on a preset expected vehicle speed and the actual vehicle speed... The desired acceleration of the target vehicle is determined using a model predictive control algorithm based on the actual acceleration, the target vehicle weight, and the target slope angle. A torque command is then determined so that the drive motor of the target vehicle can control the longitudinal speed according to the torque command. This allows for effective constraint of acceleration and rate of change while tracking the desired speed, avoiding discomfort caused by sudden acceleration or deceleration. Furthermore, by fusing the desired acceleration, target vehicle weight, and slope angle to generate the torque command, the drive motor control is made more closely aligned with the actual dynamic characteristics of the vehicle, significantly improving speed tracking accuracy, adaptability to complex road conditions, and passenger comfort.

[0163] Figure 6 This is a schematic diagram of a longitudinal vehicle speed control device provided in an embodiment of this application. Figure 6 As shown, the device 60 includes a slope calculation module 61, a vehicle weight calculation module 62, a desired acceleration determination module 63, and a longitudinal vehicle speed control module 64.

[0164] The slope calculation module 61 determines the target slope angle of the road where the target vehicle is located based on the target vehicle's unloaded mass, three-axis acceleration, three-axis angular velocity, and actual vehicle speed. The vehicle weight calculation module 62 determines the target vehicle weight based on the target slope angle, using a preset longitudinal dynamics model and a preset recursive least squares algorithm; the preset recursive least squares algorithm includes a forgetting factor. The desired acceleration determination module 63 determines the desired acceleration of the target vehicle based on a preset desired vehicle speed, actual vehicle speed, and actual acceleration, using a model predictive control algorithm. The longitudinal speed control module 64 determines the torque command based on the desired acceleration, target vehicle weight, and target slope angle, so that the target vehicle's drive motor controls the longitudinal speed according to the torque command.

[0165] The longitudinal vehicle speed control device provided in this application embodiment can execute the technical solution shown in the above method embodiment. Its implementation principle and beneficial effects are similar, and will not be described again here.

[0166] Furthermore, based on the above embodiments, the slope calculation module 61 includes a first slope calculation unit, a second slope calculation unit, and a slope fusion unit.

[0167] The system comprises the following components: a first slope calculation unit, a second slope calculation unit, and a third slope fusion unit. The first slope calculation unit calculates the target vehicle's pitch angle using a pre-defined longitudinal dynamics model based on the unloaded mass, three-axis acceleration, three-axis angular velocity, and actual vehicle speed. The second slope calculation unit calculates the target vehicle's pitch angle using unit quaternions and an unscented Kalman filter algorithm, based on the three-axis acceleration, three-axis angular velocity, and actual vehicle speed. The second slope angle is then determined based on the relationship between the slope angle and the pitch angle. The third slope fusion unit uses a Kalman filter to fuse the first and second slope angles to determine the target slope angle.

[0168] The longitudinal vehicle speed control device provided in this application embodiment can execute the technical solution shown in the above method embodiment. Its implementation principle and beneficial effects are similar, and will not be described again here.

[0169] Furthermore, based on the above embodiments, when the pitch angle of the target vehicle is calculated using unit quaternions and an unscented Kalman filter algorithm based on the triaxial acceleration, triaxial angular velocity, and actual vehicle speed, the second slope calculation unit is specifically used to solve the preset vehicle attitude change equation using the first-order Runge-Kutta method based on the triaxial angular velocity of the target vehicle, obtaining the real-time unit quaternion at each sampling moment; the vehicle attitude change equation is used to characterize the relationship between the unit quaternion and the triaxial angular velocity, and the unit quaternion is used to represent the real-time attitude of the vehicle; based on the actual vehicle speed and the real-time unit quaternion at each sampling moment, the predicted value of the triaxial acceleration is determined using the preset vehicle attitude estimation output equation; the vehicle attitude estimation output equation is used to characterize the relationship between the unit quaternion and the triaxial acceleration; based on the predicted value of the triaxial acceleration and the triaxial acceleration, the attitude estimate of the target vehicle at the current moment is determined using the unscented Kalman filter algorithm; based on the attitude estimate of the target vehicle at the current moment, the pitch angle of the target vehicle is determined.

[0170] The longitudinal vehicle speed control device provided in this application embodiment can execute the technical solution shown in the above method embodiment. Its implementation principle and beneficial effects are similar, and will not be described again here.

[0171] Furthermore, based on the above embodiments, the slope fusion unit is specifically used to predict the initial parameters of the preset Kalman filter algorithm and determine the prior estimation result at the current time. The prior estimation result includes the prior slope angle and the prior error covariance matrix at the current time. The first slope angle is used as the first observation to update the prior estimation result for the first time, and the first posterior estimation result is obtained. The second slope angle is used as the second observation to update the first posterior estimation result for the second time, and the slope angle value in the second posterior estimation result is used as the target slope angle.

[0172] The longitudinal vehicle speed control device provided in this application embodiment can execute the technical solution shown in the above method embodiment. Its implementation principle and beneficial effects are similar, and will not be described again here.

[0173] Furthermore, based on the above embodiments, the second slope calculation unit can also be used to calculate the slope using the vehicle longitudinal dynamics model based on the target vehicle weight, three-axis acceleration, three-axis angular velocity and actual vehicle speed, to obtain the third slope angle; the third slope angle is the slope angle after optimizing the first slope angle, and is used to execute the vehicle weight calculation step at the next moment.

[0174] The longitudinal vehicle speed control device provided in this application embodiment can execute the technical solution shown in the above method embodiment. Its implementation principle and beneficial effects are similar, and will not be described again here.

[0175] Further, based on the above embodiments, the expected acceleration determination module 63 is specifically used to determine the predicted vehicle speed and acceleration values ​​at each sampling time in the prediction time domain based on the preset expected vehicle speed, actual vehicle speed, and actual acceleration, using a preset discrete state-space model, and obtain the predicted vehicle speed sequence and the predicted acceleration sequence respectively; the discrete state-space model is the discrete state-space expression for longitudinal vehicle speed control; for each sampling time, when the predicted acceleration value at the sampling time in the predicted acceleration sequence is less than a first threshold and the impact is less than a second threshold, the predicted vehicle speed value corresponding to the sampling time in the predicted vehicle speed sequence is optimized, the vehicle expected acceleration increment at the sampling time is calculated, and the vehicle expected acceleration increment sequence is obtained; the impact is the rate of change of the acceleration prediction value per unit time; the vehicle expected acceleration increment sequence is solved optimally using a preset cost function to obtain the target vehicle expected acceleration increment sequence; based on the first expected acceleration increment in the target vehicle expected acceleration increment sequence, the expected acceleration of the target vehicle is determined; the first expected acceleration increment is the first term in the target vehicle expected acceleration increment sequence.

[0176] The longitudinal vehicle speed control device provided in this application embodiment can execute the technical solution shown in the above method embodiment. Its implementation principle and beneficial effects are similar, and will not be described again here.

[0177] See Figure 7 The figure is a schematic diagram of the structure of an electronic device provided in an embodiment of this application, including:

[0178] Memory 11 is used to store computer programs;

[0179] The processor 12 is used to implement the steps of the longitudinal vehicle speed control method described in any of the above method embodiments when executing the computer program.

[0180] In this embodiment, the device can be an in-vehicle computer, a PC (Personal Computer), or a terminal device such as a smartphone, tablet computer, handheld computer, or portable computer.

[0181] The device may include a memory 11, a processor 12, and a bus 13.

[0182] The memory 11 includes at least one type of readable storage medium, such as flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 11 may be an internal storage unit of the device, such as the hard disk of the device. In other embodiments, the memory 11 may be an external storage device of the device, such as a plug-in hard disk, SmartMedia Card (SMC), Secure Digital (SD) card, Flash Card, etc. Furthermore, the memory 11 may include both internal and external storage units of the device. The memory 11 can be used not only to store application software and various types of data installed on the device, such as program code for executing longitudinal vehicle speed control methods, but also to temporarily store data that has been output or will be output. In some embodiments, the processor 12 may be a central processing unit (CPU).

[0183] In some embodiments, processor 12 may be a central processing unit (CPU), controller, microcontroller, microprocessor or other data processing chip, used to run program code stored in memory 11 or process data, such as program code for executing a longitudinal vehicle speed control method.

[0184] This bus 13 can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 7 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0185] Furthermore, the device may also include a network interface 14, which may optionally include a wired interface and / or a wireless interface (such as a Wi-Fi interface, a Bluetooth interface, etc.), typically used to establish communication connections between the device and other electronic devices.

[0186] Optionally, the device may further include a user interface 15, which may include a display, an input unit such as a keyboard, and optionally, a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen, etc. The display may also be appropriately referred to as a screen or display unit, used to display information processed in the device and to display a visual user interface.

[0187] Figure 7 Only devices with components 11-15 are shown; those skilled in the art will understand that... Figure 7 The structure shown does not constitute a limitation on the device and may include fewer or more components than shown, or combine certain components, or have different component arrangements.

[0188] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this application also provides a computer-readable storage medium storing computer instructions for causing the computer to perform the methods described in any of the above embodiments.

[0189] The computer-readable media in this application embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0190] The computer instructions stored in the storage medium of the above embodiments are used to cause the computer to perform the methods described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0191] It should be noted that the various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for methods, apparatuses, electronic devices, and media, since they are basically similar to the method embodiments, the descriptions are relatively simple, and relevant parts can be referred to the descriptions of the method embodiments. The methods, apparatuses, electronic devices, and media described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components indicated as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of the solution in this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0192] The above description is merely one specific embodiment of this application, but the scope of protection of this application 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 this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A longitudinal vehicle speed control method, characterized in that, include: Based on the unloaded mass, triaxial acceleration, triaxial angular velocity and actual vehicle speed, the slope is calculated using a preset vehicle longitudinal dynamics model to obtain the first slope angle; Based on the triaxial acceleration, triaxial angular velocity and actual vehicle speed, the pitch angle of the target vehicle is calculated using unit quaternions and unscented Kalman filtering algorithm. Based on the pitch angle of the target vehicle, the second slope angle is determined using the relationship between the slope angle and the pitch angle; The first slope angle and the second slope angle are fused using the Kalman filter method to determine the target slope angle; Based on the target slope angle, the target vehicle weight is determined using a preset longitudinal dynamics model and a preset recursive least squares algorithm; the preset recursive least squares algorithm includes a forgetting factor. Based on the preset expected vehicle speed, actual vehicle speed, and actual acceleration, the vehicle speed prediction value and acceleration prediction value at each sampling time in the prediction time domain are determined using a preset discrete state space model, and the predicted vehicle speed sequence and predicted acceleration sequence are obtained respectively; the discrete state space model is the discrete state space expression for longitudinal vehicle speed control. For each sampling time, when it is determined that the predicted acceleration value at the sampling time in the predicted acceleration sequence is less than a first threshold and the impact is less than a second threshold, the predicted vehicle speed value corresponding to the sampling time in the predicted vehicle speed sequence is optimized, the expected vehicle acceleration increment at the sampling time is calculated, and the expected vehicle acceleration increment sequence is obtained; the impact is the rate of change of the predicted acceleration value per unit time. The optimal solution for the desired acceleration increment sequence of the vehicle is obtained by using a preset cost function; Based on the first expected acceleration increment in the expected acceleration increment sequence of the target vehicle, the expected acceleration of the target vehicle is determined; the first expected acceleration increment is the first item in the expected acceleration increment sequence of the target vehicle; based on the expected acceleration, the target vehicle weight, and the target slope angle, a torque command is determined so that the drive motor of the target vehicle controls the longitudinal speed according to the torque command.

2. The method according to claim 1, characterized in that, Based on the triaxial acceleration, triaxial angular velocity, and actual vehicle speed, the pitch angle of the target vehicle is calculated using a unit quaternion and unscented Kalman filter algorithm, including: Based on the three-axis angular velocities of the target vehicle, the first-order Runge-Kutta method is used to solve the preset vehicle attitude change equation to obtain the real-time unit quaternion at each sampling time. The vehicle attitude change equation is used to characterize the relationship between the unit quaternion and the three-axis angular velocities, and the unit quaternion is used to represent the real-time attitude of the vehicle. Based on the actual vehicle speed and the real-time unit quaternion at each sampling time, the predicted value of the three-axis acceleration is determined using a preset vehicle attitude estimation output equation; the vehicle attitude estimation output equation is used to characterize the relationship between the unit quaternion and the three-axis acceleration. Based on the predicted triaxial acceleration and the triaxial acceleration, the attitude estimate of the target vehicle at the current moment is determined using an unscented Kalman filter algorithm. The pitch angle of the target vehicle is determined based on the estimated attitude of the target vehicle at the current moment.

3. The method according to claim 1, characterized in that, The method of using Kalman filtering to fuse the first slope angle and the second slope angle to determine the target slope angle includes: The initial parameters of the preset Kalman filter algorithm are predicted to determine the prior estimation result at the current time; the prior estimation result includes the prior slope angle and the prior error covariance matrix at the current time. Using the first slope angle as the first observation, the prior estimation result is updated for the first time to obtain the first posterior estimation result; The second slope angle is used as the second observation. The first posterior estimation result is updated by a second observation to obtain the second posterior estimation result. The slope angle value in the second posterior estimation result is used as the target slope angle.

4. The method according to claim 1, characterized in that, The method further includes: Based on the target vehicle weight, three-axis acceleration, three-axis angular velocity, and actual vehicle speed, the slope is calculated using the vehicle longitudinal dynamics model to obtain the third slope angle; the third slope angle is the slope angle optimized from the first slope angle, and is used to perform the vehicle weight calculation step at the next moment.

5. A longitudinal vehicle speed control device, characterized in that, include: The first slope calculation unit is used to calculate the slope based on the unloaded mass, triaxial acceleration, triaxial angular velocity and actual vehicle speed using a preset vehicle longitudinal dynamics model, and obtain the first slope angle. The second slope calculation unit is used to calculate the pitch angle of the target vehicle based on the triaxial acceleration, triaxial angular velocity and actual vehicle speed, using unit quaternions and unscented Kalman filtering algorithm; and to determine the second slope angle based on the pitch angle of the target vehicle using the relationship between the slope angle and the pitch angle. The slope fusion unit is used to fuse the first slope angle and the second slope angle using the Kalman filter method to determine the target slope angle; The vehicle weight calculation module is used to determine the target vehicle weight based on the target slope angle, using a preset longitudinal dynamics model and a preset recursive least squares algorithm; the preset recursive least squares algorithm includes a forgetting factor. The desired acceleration determination module is used to determine the predicted vehicle speed and acceleration values ​​at each sampling time in the prediction time domain based on the preset desired vehicle speed, actual vehicle speed, and actual acceleration, using a preset discrete state-space model, and to obtain the predicted vehicle speed sequence and the predicted acceleration sequence respectively; the discrete state-space model is the discrete state-space expression for longitudinal vehicle speed control. The expected acceleration determination module is further configured to, for each sampling time, optimize the vehicle speed prediction value corresponding to the sampling time in the predicted vehicle speed sequence when the predicted acceleration prediction value at the sampling time in the predicted acceleration sequence is less than a first threshold and the impact is less than a second threshold, calculate the vehicle expected acceleration increment at the sampling time, and obtain the vehicle expected acceleration increment sequence; the impact is the rate of change of the acceleration prediction value per unit time. The desired acceleration determination module is further configured to use a preset cost function to solve for the optimal solution of the vehicle desired acceleration increment sequence, thereby obtaining the target vehicle desired acceleration increment sequence. The desired acceleration determination module is further configured to determine the desired acceleration of the target vehicle based on the first desired acceleration increment in the desired acceleration increment sequence of the target vehicle; the first desired acceleration increment is the first item in the desired acceleration increment sequence of the target vehicle. The longitudinal speed control module is used to determine a torque command based on the desired acceleration, the target vehicle weight, and the target slope angle, so that the drive motor of the target vehicle controls the longitudinal speed according to the torque command.

6. An electronic device, characterized in that, The device includes: a processor, and a memory communicatively connected to the processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory to implement the method as described in any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1 to 4.

Images