Vehicle control method, vehicle and computer readable storage medium

Abstract

The embodiment of the invention provides a vehicle control method, a vehicle and a computer readable storage medium, and the method comprises the steps that multi-source driving data of the vehicle are acquired, and the multi-source driving data are used for representing dynamic state parameters of the vehicle under different working conditions; the multi-source driving data is input into a dynamic model of the vehicle for prediction, the initial acceleration of the vehicle is obtained, the dynamic model is obtained by training a multi-source driving data sample of the vehicle sample and an initial acceleration sample corresponding to the multi-source driving data sample, and the initial acceleration of the vehicle is obtained; the multi-source driving data sample is used for representing dynamic state parameters of the vehicle sample under different working conditions; the initial acceleration is compared with a target acceleration, a comparison result is obtained, and the target acceleration is determined by a simulation actuator of the vehicle; and determining a control strategy of the vehicle based on the comparison result, and controlling the vehicle to run according to the control strategy. The technical problem that the vehicle cannot be effectively controlled is solved.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the technical field of vehicle control, in particular, to a vehicle control method, a vehicle and a computer readable storage medium. BACKGROUND

[0002] Currently, before the vehicle is controlled, the parameters of the vehicle can be calibrated. The calibration method for the parameters of the vehicle in the related art often focuses on data acquisition and modeling under a single working condition, which limits the generalization ability of the calibration model in actual application, and it is difficult to cover a wide range of actual driving scenarios. Therefore, there is still a technical problem that the vehicle cannot be effectively controlled.

[0003] At present, there is no good solution to the above problems. SUMMARY

[0004] Embodiments of the present application provide a vehicle control method, a vehicle and a computer readable storage medium to at least solve the technical problem that the vehicle cannot be effectively controlled.

[0005] According to an aspect of embodiments of the present application, a vehicle control method is provided. The method can include: obtaining multi-source driving data of a vehicle, wherein the multi-source driving data is used to represent dynamic state parameters of the vehicle under different working conditions; inputting the multi-source driving data into a dynamic model of the vehicle for prediction to obtain an initial acceleration of the vehicle, wherein the dynamic model is trained using multi-source driving data samples of a vehicle sample and initial acceleration samples corresponding to the multi-source driving data samples, and the multi-source driving data samples are used to represent dynamic state parameters of the vehicle sample under different working conditions; comparing the initial acceleration with a target acceleration to obtain a comparison result, wherein the target acceleration is determined by a simulation executor of the vehicle, and the comparison result is used to represent a degree of deviation of the initial acceleration from the target acceleration; determining a control strategy of the vehicle based on the comparison result, and controlling the vehicle to travel according to the control strategy.

[0006] Further, the dynamic state parameters include pedal opening, speed and acceleration, the different working conditions include load working conditions and road surface working conditions, and inputting the multi-source driving data into the dynamic model of the vehicle for prediction to obtain the initial acceleration of the vehicle includes: inputting the multi-source driving data into the dynamic model for prediction to obtain a pedal power conversion coefficient of the vehicle and a friction force of the vehicle; generating a two-dimensional interpolation table based on the pedal power conversion coefficient and the friction force, wherein the two-dimensional interpolation table is used to represent a corresponding relationship between the pedal opening, the speed, the load working condition, the road surface working condition and the acceleration; determining the initial acceleration based on the two-dimensional interpolation table.

[0007] Further, the method further comprises: in response to the comparison result being that the initial acceleration deviates from the target acceleration by a degree greater than a degree threshold, adjusting the pedal power conversion coefficient and the friction force to obtain an adjustment result; updating the two-dimensional interpolation table based on the adjustment result; and determining the initial acceleration based on the two-dimensional interpolation table, including: determining the initial acceleration based on the updated two-dimensional interpolation table.

[0008] Further, the method further comprises: updating the dynamic model based on the adjustment result, and determining the initial acceleration according to the updated dynamic model.

[0009] Further, determining the control strategy of the vehicle based on the comparison result, including: in response to the comparison result being that the initial acceleration deviates from the target acceleration by a degree less than or equal to the degree threshold, determining the control strategy based on the pedal power conversion coefficient and the friction force.

[0010] Further, obtaining the multi-source driving data of the vehicle, including: obtaining a load level corresponding to a load condition, a road adhesion coefficient corresponding to a road condition, and a pedal opening degree; and in response to the vehicle being in a starting state and the pedal opening degree being fixed, obtaining a speed in a target time period at the fixed pedal opening degree and an acceleration in the target time period at the pedal opening degree.

[0011] Further, in response to the vehicle being in a starting state and the pedal opening degree being fixed, obtaining a speed in a target time period at the fixed pedal opening degree and an acceleration in the target time period at the pedal opening degree, including: in response to the target time period being greater than or equal to a time period threshold, or the speed being greater than or equal to a speed threshold, determining a first size relationship between the pedal opening degree and a pedal opening degree threshold; in response to the first size relationship being that the pedal opening degree is greater than or equal to the pedal opening degree threshold, determining a second size relationship between the road adhesion coefficient and a road adhesion coefficient threshold; and in response to the second size relationship being that the road adhesion coefficient is greater than or equal to the road adhesion coefficient threshold and the load level is greater than or equal to a load level threshold, obtaining the speed and the acceleration.

[0012] Further, comparing the initial acceleration and the target acceleration to obtain a comparison result, including: determining a driving scene of the vehicle; and in response to the driving scene being a target driving scene, comparing the initial acceleration and the target acceleration to obtain the comparison result, wherein the target driving scene is used to represent a driving scene set in the simulation actuator.

[0013] According to another aspect of the embodiments of this application, a vehicle control device is also provided, comprising: an acquisition unit for acquiring multi-source driving data of the vehicle, wherein the multi-source driving data is used to represent the dynamic state parameters of the vehicle under different operating conditions; a prediction unit for inputting the multi-source driving data into a dynamic model of the vehicle for prediction to obtain the initial acceleration of the vehicle, wherein the dynamic model is trained using multi-source driving data samples of the vehicle sample and initial acceleration samples corresponding to the multi-source driving data samples, wherein the multi-source driving data samples are used to represent the dynamic state parameters of the vehicle sample under different operating conditions; a comparison unit for comparing the initial acceleration with a target acceleration to obtain a comparison result, wherein the target acceleration is determined by the vehicle's simulator, and the comparison result is used to indicate the degree to which the initial acceleration deviates from the target acceleration; and a control unit for determining a control strategy for the vehicle based on the comparison result, and controlling the vehicle to drive according to the control strategy.

[0014] According to another aspect of the present invention, a processor is also provided. The processor is used to run a program, wherein the program, when run by the processor, performs the methods described in the embodiments of the present invention.

[0015] According to another aspect of the embodiments of this application, a vehicle is also provided, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods in various embodiments of this application when it runs.

[0016] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is running, it controls the device where the computer-readable storage medium is located to perform the methods of various embodiments of this application.

[0017] According to another aspect of the embodiments of this application, a computer program product is also provided, including a computer program that, when executed by a processor, implements the methods of various embodiments of this application.

[0018] According to another aspect of the embodiments of this application, a computer program product is also provided, including a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, implements the methods in various embodiments of this application.

[0019] According to another aspect of the embodiments of this application, a computer program is also provided, which, when executed by a processor, implements the methods of the various embodiments of this application.

[0020] In this embodiment, by acquiring multi-source driving data of the vehicle, which includes dynamic state parameters under different operating conditions, the initial acceleration of the vehicle can be obtained by predicting the multi-source driving data using the vehicle's dynamic model. Then, the initial acceleration is compared with the target acceleration determined by the vehicle's simulator to determine the degree of deviation between the initial acceleration and the target acceleration. Based on this deviation, a control strategy is determined to control the vehicle's movement. In other words, this embodiment achieves precise vehicle control by training a dynamic model using multi-dimensional data samples and combining real-time deviation feedback. Especially under complex and variable operating conditions, it significantly improves the reliability of vehicle simulation testing, reduces the cost and time of subsequent real-vehicle debugging, and overcomes the limitations of related technologies where inaccurate vehicle control is caused by a lack of multi-condition adaptability, coarse nonlinear relationship modeling, and a lack of real-time deviation feedback mechanisms during simulation testing. This solves the technical problem of ineffective vehicle control and achieves the technical effect of effective vehicle control. Attached Figure Description

[0021] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0022] Figure 1 This is a flowchart of a vehicle control method according to an embodiment of this application;

[0023] Figure 2 This is a flowchart of a real vehicle data acquisition method according to an embodiment of this application;

[0024] Figure 3 This is a flowchart illustrating a method for calculating the accelerator / brake pedal power conversion coefficient and the current vehicle rolling friction force according to an embodiment of this application.

[0025] Figure 4 This is a schematic diagram of a vehicle control device according to an embodiment of this application. Detailed Implementation

[0026] 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 should fall within the scope of protection of the present application.

[0027] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0028] According to an embodiment of this application, a vehicle control method embodiment is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0029] This embodiment provides a vehicle control method. Figure 1 This is a flowchart of a vehicle control method according to an embodiment of this application, such as... Figure 1 As shown, the process may include the following steps.

[0030] Step S102: Obtain multi-source driving data of the vehicle.

[0031] In the technical solution provided in step S102 of this application, multi-source driving data can be used to represent the dynamic state parameters of the vehicle under different operating conditions. Multi-source driving data may include accelerator / brake pedal opening, vehicle speed, and vehicle acceleration, which are only examples here. The different operating conditions mentioned above can be different load conditions or different road surface conditions. Specifically, accelerator / brake pedal opening can be simply referred to as pedal opening, and can be represented by B; vehicle speed can be simply referred to as speed or vehicle velocity, and can be represented by v; vehicle acceleration can be simply referred to as acceleration, and can be represented by a.

[0032] In this embodiment, the acquired multi-source driving data can be used to construct and verify the vehicle's dynamics model, ensuring that the dynamics model can accurately reflect the vehicle's behavior under different operating conditions.

[0033] Optionally, the accelerator / brake pedal opening can record data on the vehicle's pedals at various percentage opening states, which can be used to understand the relationship between driver intent and vehicle response.

[0034] Optionally, vehicle speed is a measurement of the speed of a vehicle over time at different pedal openings, which can be used to analyze the vehicle's dynamic response characteristics.

[0035] Optionally, acceleration records changes in vehicle acceleration. For example, when the pedal opening changes, acceleration can be used to measure the vehicle's power performance and control effectiveness.

[0036] Optionally, multi-source driving data can be acquired through sensors, such as acceleration sensors, speed sensors, accelerator / brake pedal position sensors, load level sensors, weighing sensors or pressure sensors, and road condition sensors, etc. These are just examples and no specific limitations are made here.

[0037] Optionally, acceleration sensors can be used to measure the vehicle's lateral, longitudinal, and vertical acceleration. Speed ​​sensors can be used to monitor the vehicle's real-time speed and can be integrated into the vehicle's wheels or drivetrain to provide vehicle speed information. Accelerator / brake pedal position sensors can be used to record the position of the accelerator or brake pedal, i.e., the pedal opening. Load level sensors can directly measure load levels, while weighing or pressure sensors can be used to indirectly measure load levels to evaluate the vehicle's dynamic performance under different load levels. Road condition sensors can include sensors for road conditions such as humidity, temperature, and roughness to estimate the road adhesion coefficient, thereby determining the friction between the vehicle and the ground.

[0038] It should be noted that the types and acquisition methods of multi-source driving data are only illustrative examples and are not subject to specific restrictions.

[0039] In this embodiment of the application, by collecting the accelerator / brake pedal opening, speed and acceleration data of the vehicle under the above different working conditions, a multi-source driving database containing a variety of dynamic state parameters can be formed, thereby providing the necessary data support for the subsequent establishment of the dynamic model, ensuring the accuracy and applicability of the dynamic model, and enabling the dynamic model to effectively handle the longitudinal dynamic control problem of the vehicle under complex and varied driving conditions.

[0040] Step S104: Input the multi-source driving data into the vehicle's dynamic model for prediction to obtain the vehicle's initial acceleration.

[0041] In the technical solution provided in step S104 of this application, the dynamic model can be trained using multi-source driving data samples of the vehicle sample and the corresponding initial acceleration samples. The multi-source driving data samples can be used to represent the dynamic state parameters of the vehicle sample under different operating conditions. The initial acceleration can be a predicted acceleration, which can be obtained through... To express.

[0042] In this embodiment, after acquiring multi-source driving data of the vehicle, the multi-source driving data can be input into the vehicle's dynamics model for prediction to obtain the vehicle's initial acceleration. The dynamics model is obtained by training with a large number of multi-source driving data samples. These multi-source driving data samples not only include parameters such as accelerator / brake pedal opening, vehicle speed, and vehicle acceleration, but also cover corresponding data under different operating conditions (e.g., load conditions and road conditions). Based on these multi-source driving data samples, the dynamics model can learn the nonlinear relationship between accelerator / brake pedal opening and other dynamic parameters (e.g., acceleration), and how to adjust these nonlinear relationships according to the load level and road conditions.

[0043] Optionally, after the dynamics model is trained, new multi-source driving data can be fed into the dynamics model as input. This data can include the current pedal opening, vehicle speed, load level, road conditions, etc. Then, based on the input multi-source driving data and the learned dynamic behavior patterns, the dynamics model can predict the vehicle's initial acceleration (…). This refers to the expected change in acceleration of the vehicle under the current operating conditions.

[0044] Optionally, through the above method, the dynamic model can predict how a specific pedal opening will affect the vehicle's acceleration under various given operating conditions, thus providing a basis for achieving precise longitudinal vehicle control. Predicting acceleration. It serves as an important basis for formulating subsequent control strategies, and can predict acceleration. The acceleration is compared with the target acceleration to evaluate the accuracy of the dynamic model prediction and the actual effect of vehicle control. The control parameters are then adjusted based on the comparison results to optimize the vehicle's performance in the loop simulation test.

[0045] In this embodiment, a pre-trained dynamics model is used to predict and generate the vehicle's expected initial acceleration under the current operating conditions, providing a quantitative reference for subsequent control strategy optimization and verification. This method based on dynamics model prediction can more accurately simulate and control vehicle dynamics, improving the accuracy and reliability of in-loop simulation testing of autonomous driving algorithms.

[0046] Step S106: Compare the initial acceleration with the target acceleration to obtain the comparison result.

[0047] In the technical solution provided in step S106 of this application, the target acceleration can be determined by the vehicle's simulation actuator, and the comparison result can be used to indicate the degree to which the initial acceleration deviates from the target acceleration. The target acceleration can be the actual output acceleration of the simulation actuator. .

[0048] In this embodiment, after inputting multi-source driving data into the vehicle's dynamics model for prediction and obtaining the vehicle's initial acceleration, the initial acceleration can be compared with the target acceleration to calculate the degree of deviation between the two, thus obtaining a comparison result. This comparison result describes the acceleration difference between the acceleration predicted by the dynamics model and the actual output acceleration of the simulated actuator. This acceleration difference reflects the accuracy of the dynamics model's prediction and the precision of vehicle control.

[0049] Optionally, the target acceleration is the acceleration value actually output by the simulator under the current operating conditions according to the instructions of the autonomous driving algorithm, representing the acceleration target that the vehicle is expected to achieve in the in-loop simulation test.

[0050] Optionally, the initial acceleration is the acceleration value predicted by the trained dynamic model based on the input multi-source driving data. It is based on historical data and current operating conditions, reflecting the dynamic model's understanding and prediction of the vehicle's longitudinal dynamic behavior.

[0051] Optionally, the absolute difference between the predicted acceleration and the actual output acceleration of the simulated actuator can be calculated. This allows for a comparison, quantifying the degree of deviation between the initial acceleration predicted by the dynamic model and the target acceleration. The smaller the deviation, the more accurate the dynamic model prediction and the more precise the vehicle control.

[0052] In this embodiment, the degree of deviation between the initial acceleration predicted by the dynamic model and the target acceleration expected to be output by the simulated actuator is quantified. This comparison result can evaluate the prediction accuracy of the dynamic model and the accuracy of the vehicle's longitudinal control, and also serves as the basis for subsequent self-correction and control strategy optimization. Through continuous monitoring and adjustment, it can be ensured that the longitudinal dynamic behavior of the vehicle under various operating conditions conforms to the predetermined control objectives, providing a solid foundation for the development and verification of autonomous driving algorithms.

[0053] Step S108: Based on the comparison results, determine the vehicle control strategy, and control the vehicle to drive according to the control strategy.

[0054] In the technical solution provided by step S108 of this application, after comparing the initial acceleration with the target acceleration and obtaining the comparison result, the vehicle control strategy can be determined based on the comparison result, and the vehicle can be controlled to drive according to the control strategy.

[0055] In this embodiment, by evaluating the deviation between the vehicle's initial acceleration and the target acceleration in real time, and then adjusting the vehicle's longitudinal dynamics control parameters, it can be ensured that the vehicle travels in the expected manner.

[0056] Optionally, the deviation Δa between the initial acceleration and the target acceleration can be calculated in real time. If Δa exceeds a set threshold (e.g., 5%), the control strategy needs to be adjusted. The direction of this deviation can then be determined, i.e., whether the initial acceleration is too small or too large compared to the target acceleration, thus guiding the adjustment of the control strategy. Based on the direction and magnitude of the deviation, the values ​​of the pedal power conversion coefficient K and the friction coefficient C can also be adjusted to reduce the deviation. The adjustment range of the pedal power conversion coefficient K and the friction coefficient C can be dynamically determined according to the severity of the deviation; for example, a larger adjustment range is required when the deviation is large.

[0057] Optionally, applying the adjusted parameters K and C to the dynamic model can re-predict the initial acceleration and generate control commands (such as throttle or brake pedal opening commands) that are more closely matched to the target acceleration.

[0058] Optionally, the generated control commands can be sent to the vehicle's powertrain (e.g., engine and braking system) to control the vehicle to perform acceleration or deceleration, adjusting the vehicle's initial acceleration to near the target value. While executing the control commands, the difference between the vehicle's initial acceleration and the target acceleration can be continuously monitored. If the deviation exceeds a threshold again, the above process is repeated until the deviation is controlled within the threshold range. This process forms a closed-loop control system, where real-time evaluation of the deviation and dynamic adjustment of parameters constitute the key parts of the closed-loop control. Through continuous feedback and adjustment mechanisms, closed-loop control ensures that the vehicle's longitudinal dynamics performance conforms to the set control objectives.

[0059] In this embodiment, the predicted results of the dynamic model (e.g., initial acceleration) are compared with the target acceleration, and the vehicle's control strategy and parameters are adjusted based on the comparison results. This improves the longitudinal dynamic control accuracy of the vehicle under various operating conditions. Through closed-loop control and real-time parameter adjustment, the vehicle's autonomous driving system can respond more flexibly and accurately to the needs of different driving scenarios, ensuring safe, comfortable, and energy-efficient driving performance in the testing and use of functions such as Adaptive Cruise Control (ACC) and Automatic Emergency Braking (AEB). This optimization process not only reduces the cost and time of subsequent real-vehicle debugging but also provides efficient data support and assurance for the iterative upgrade of the vehicle's intelligent driving system.

[0060] It should be noted that in the vehicle control method of this embodiment, the vehicle can also interact with roadside equipment and terminal equipment. Optionally, the vehicle can send an information subscription request to the roadside equipment. This message subscription request can include specific types of information that the vehicle needs to receive, such as road conditions, traffic signal status, and obstacle warnings ahead. In response to the information subscription request, the roadside equipment can send roadside perception information to the vehicle. For example, the roadside equipment will filter out roadside perception information that meets the vehicle's needs based on its own perception capabilities and stored information, and send it to the vehicle at a certain frequency. In addition to communicating with the roadside equipment, the vehicle can also receive driving scenario switching instructions transmitted by the terminal equipment through the network. For example, the driving scenario switching instructions can be used to switch the vehicle to energy-saving mode, sport mode, autonomous driving mode, etc., so that the vehicle can adapt to new driving scenarios.

[0061] Through steps S102 to S108, multi-source driving data of the vehicle is acquired. Since the multi-source driving data contains dynamic state parameters under different operating conditions, the initial acceleration of the vehicle can be obtained by predicting the multi-source driving data using the vehicle's dynamic model. Then, the initial acceleration is compared with the target acceleration determined by the vehicle's simulation actuator to determine the degree of deviation between the initial acceleration and the target acceleration. Based on this deviation, a control strategy is determined to control the vehicle's movement. In other words, this embodiment of the application achieves precise vehicle control by training a dynamic model using multi-dimensional data samples and combining real-time deviation feedback. Especially under complex and variable operating conditions, it significantly improves the reliability of vehicle simulation testing, reduces the cost and time of subsequent real-vehicle debugging, and overcomes the limitations of related technologies in simulation testing caused by a lack of multi-condition adaptability, coarse nonlinear relationship modeling, and a lack of real-time deviation feedback mechanisms, resulting in inaccurate vehicle control. This solves the technical problem of ineffective vehicle control and achieves the technical effect of effective vehicle control.

[0062] The above-mentioned method of this application will be further described below.

[0063] As an optional implementation, the dynamic state parameters include pedal opening, speed, and acceleration. Different operating conditions include load conditions and road surface conditions. Step S104 involves inputting multi-source driving data into the vehicle's dynamic model for prediction to obtain the vehicle's initial acceleration. This includes: inputting multi-source driving data into the dynamic model for prediction to obtain the vehicle's pedal power conversion coefficient and friction force; generating a two-dimensional interpolation table based on the pedal power conversion coefficient and friction force, wherein the two-dimensional interpolation table is used to represent the correspondence between pedal opening, speed, load conditions, road surface conditions, and acceleration; and determining the initial acceleration based on the two-dimensional interpolation table.

[0064] In this embodiment, the load condition can include three states: no load, half load, and full load. Each state can correspond to a different vehicle load level (also called load rating). The load level can be determined by M. k The values ​​are represented as k=1, 2, 3. Data collected under different load conditions can be used to establish a dynamic model capable of handling load variations. Road conditions can take into account various road conditions encountered by the vehicle during operation (e.g., road adhesion coefficient u), such as dry road surface (u1=0.8), wet and slippery road surface (u2=0.4), and icy and snowy road surface (u3=0.2), etc. These different road conditions can affect the vehicle's dynamic behavior.

[0065] Optionally, the pedal power conversion coefficient can be the accelerator / brake pedal power conversion coefficient, which can be represented by K. The vehicle's friction force can be the current rolling friction force, which can be represented by C. A two-dimensional interpolation table can be used to represent the correspondence between pedal opening, speed, load condition, road condition, and acceleration. For example, each "load level - road condition" combination can correspond to one two-dimensional interpolation table, enabling real-time querying of acceleration for any combination of pedal opening B and speed v.

[0066] Optionally, the accelerator / brake pedal power conversion factor describes the conversion efficiency between pedal opening and output power. The accelerator / brake pedal power conversion factor varies under different load and road conditions. Therefore, the dynamic model needs to be able to predict appropriate accelerator / brake pedal power conversion factor values ​​based on different operating conditions. The rolling friction of a vehicle is the frictional force during vehicle operation, which is affected by load and road conditions, reflecting the rolling resistance of the vehicle under specific conditions.

[0067] Optionally, multi-source driving data, including pedal opening, speed, load conditions, and road conditions, can be input into the dynamic model. The dynamic model can predict two parameters of the vehicle: the pedal power conversion coefficient and the vehicle's friction force, based on the training data.

[0068] Optionally, after obtaining the vehicle's pedal power conversion coefficient and friction force, a two-dimensional interpolation table can be constructed. This table is used to quickly find the initial acceleration (i.e., predict the acceleration) at a given pedal opening and speed. The construction of this two-dimensional interpolation table can be based on multiple data points under different load conditions and road surface conditions. Each data point can contain the corresponding vehicle's pedal power conversion coefficient and friction force.

[0069] Optionally, under different load and road conditions, multiple data points can be collected through real vehicle testing. Each data point can record the values ​​of pedal opening, speed, pedal power conversion coefficient, and friction. Using these data points, a two-dimensional interpolation table can be established for each "load level-road condition" combination, where the horizontal axis can be the pedal opening, the vertical axis can be the speed, and the values ​​in the two-dimensional interpolation table can be the corresponding accelerations.

[0070] Optionally, in simulation testing or real-time control, a two-dimensional interpolation table can be used to quickly find the predicted acceleration values ​​for each combination of B and v, thereby enabling real-time prediction and control of the vehicle's longitudinal dynamics.

[0071] Optionally, in actual operation, based on a two-dimensional interpolation table, the initial acceleration under each combination of pedal opening B and velocity v can be determined through interpolation calculation. This process does not require recalculating the values ​​of K and C, thereby greatly improving the speed of data query and processing.

[0072] In this embodiment, the above steps, by introducing the pedal power conversion coefficient K and friction force C, utilize a two-dimensional interpolation table to achieve rapid prediction of the vehicle's longitudinal dynamics. This overcomes the delay and inaccuracy problems of acceleration prediction in traditional methods, especially when facing complex and variable operating conditions, providing more accurate prediction results. This effectively supports the optimization of vehicle control strategies and the in-loop simulation testing of autonomous driving algorithms. Furthermore, the two-dimensional interpolation table allows for real-time responses to the vehicle's longitudinal dynamics under various operating conditions, ensuring continuity and accuracy in testing and actual operation.

[0073] As an optional implementation, the method further includes: adjusting the pedal power conversion coefficient and friction force in response to a comparison result that the initial acceleration deviates from the target acceleration by a greater than a degree threshold, to obtain an adjustment result; updating the two-dimensional interpolation table based on the adjustment result; and determining the initial acceleration based on the two-dimensional interpolation table, including: determining the initial acceleration based on the updated two-dimensional interpolation table.

[0074] In this embodiment, if the comparison result is the degree to which the initial acceleration deviates from the target acceleration... When the initial acceleration predicted by the dynamic model deviates significantly from the target acceleration of the simulated actuator, the correction mechanism is activated, which can adjust the pedal power conversion coefficient and friction force.

[0075] For example, when When the percentage exceeds 5%, a correction mechanism is triggered. (If the target acceleration is too large), then the current K value can be decreased by 2% and the C value increased by 1%; if (If the target acceleration is too small), the current K value can be increased by 2% and the C value decreased by 1%; after correction, the two-dimensional interpolation table can be updated to ensure dynamic matching between the dynamic model and the simulation actuator.

[0076] Optionally, after each correction mechanism operation, the adjustment results of K and C can be used to update the previously established two-dimensional interpolation table. That is, as the correction mechanism continues to run, the dynamic model can be continuously adjusted according to the latest feedback information to better adapt to the current working conditions and improve the consistency between the initial acceleration and the target acceleration.

[0077] Optionally, the updated two-dimensional interpolation table can more accurately reflect the actual dynamic behavior of the vehicle under different load conditions and road adhesion coefficients. By dynamically adjusting the pedal power conversion coefficient and friction, the dynamic model can be corrected in real time when facing changes in operating conditions, ensuring that the predicted results of the dynamic model keep dynamically matched with the actual output of the simulation actuator, thereby improving the accuracy and reliability of the entire simulation testing process.

[0078] Optionally, the key to the above-mentioned dynamic correction mechanism lies in its ability to detect and respond to the deviation between the prediction of the dynamic model and the actual execution in real time. By fine-tuning key dynamic parameters, this deviation can be reduced, ultimately improving the accuracy and stability of the vehicle's longitudinal dynamic control. This not only enhances the adaptability of the dynamic model but also optimizes the simulation test results and reduces the burden of subsequent debugging work.

[0079] In this embodiment, a dynamic correction mechanism based on comparison result feedback is introduced to adjust the pedal power conversion coefficient K and friction force C in real time to ensure that the initial acceleration predicted by the dynamic model is consistent with the target acceleration actually output by the simulated actuator.

[0080] As an optional implementation, the method further includes: updating the dynamic model based on the adjustment results, and determining the initial acceleration according to the updated dynamic model.

[0081] In this embodiment, the dynamic model is updated based on the adjustment results, and the updated dynamic model is used to redetermine the initial acceleration. This iterative mechanism can continuously optimize the initial acceleration, ensuring that the acceleration predicted by the dynamic model, that is, the initial acceleration, is closer to the target acceleration, thereby improving the longitudinal control accuracy of the vehicle in the in-loop simulation scenario.

[0082] Optionally, if a significant deviation is found between the initial acceleration and the target acceleration during the comparison process (e.g., Δa > 5%), a correction mechanism can be activated to adjust the pedal power conversion coefficient and friction. If the target acceleration is too large, the K value is decreased (by 2%) and the C value is increased (by 1%); conversely, if the target acceleration is too small, the K value is increased (by 2%) and the C value is decreased (by 1%). These adjustments can be used to update the dynamic model, enabling it to more accurately map the relationship between pedal opening, speed, load conditions, road surface conditions, and acceleration.

[0083] Optionally, after the dynamics model is adjusted and updated based on the comparison results, multi-source driving data can be used again as input to make predictions based on the updated dynamics model, resulting in a new initial acceleration. This process can be viewed as a reassessment of the vehicle's longitudinal dynamics behavior, used to verify the effectiveness of the dynamics model adjustment and check whether the requirement for a closer match to the target acceleration has been achieved.

[0084] Optionally, to ensure continuous improvement and optimization of the dynamic model, the aforementioned process of cyclical verification and adjustment is iterative. After each prediction, the difference between the initial acceleration and the target acceleration is compared. If it still exceeds a preset threshold, the dynamic model parameters are adjusted, and the above process is repeated until the required accuracy is met, thereby achieving accurate calibration of the longitudinal motion control parameters of the vehicle in the in-loop simulation scenario.

[0085] In this embodiment, by introducing a real-time deviation feedback and parameter self-correction mechanism, the dynamic model can be continuously optimized to ensure that the predicted results are highly consistent with actual needs. This process not only improves the accuracy and reliability of the dynamic model but also provides strong support for the development of autonomous driving algorithms, helping to reduce the cost and cycle of later real-vehicle debugging, accelerate algorithm iteration, and promote the development of intelligent driving technology. Through continuous iterative verification and adjustment, it can adapt to changes in various operating conditions, providing more accurate parameter calibration for vehicle longitudinal control, thereby obtaining more reliable and stable results in in-loop simulation testing.

[0086] As an optional implementation, step S108, based on the comparison results, determines the vehicle control strategy, including: in response to the comparison results indicating that the degree of deviation of the initial acceleration from the target acceleration is less than or equal to a degree threshold, determining the control strategy based on the pedal power conversion coefficient and friction.

[0087] In this embodiment, when the comparison results show that the deviation between the initial acceleration and the target acceleration is less than or equal to a predetermined threshold, it means that the vehicle's current dynamic model parameters (e.g., pedal power conversion coefficient and friction) are sufficiently close to the ideal state to meet the requirements of in-loop simulation testing. In this case, determining the vehicle's control strategy focuses more on maintaining the existing parameters unchanged while ensuring that the vehicle's dynamic performance can stably achieve the expected target.

[0088] Optionally, when the deviation is acceptable—that is, when the initial acceleration deviates from the target acceleration by less than or equal to a threshold—the vehicle's load level and road adhesion coefficient can be monitored in real time. This is because even if the deviation is within the allowable range, changes in load level and road conditions may gradually alter the applicability of the dynamic model. Parameter adjustments can be omitted, as frequent adjustments may introduce unnecessary fluctuations, thus affecting control stability.

[0089] Optionally, based on the current pedal power conversion coefficient and friction, combined with the vehicle's real-time status (e.g., speed, acceleration, load level, road conditions), control commands that match the target acceleration can be generated. These control commands can include fine-tuning of the accelerator or brake pedal opening to ensure the vehicle smoothly reaches the target state. Even with small deviations, the vehicle's dynamic state can be continuously monitored, and future vehicle behavior can be predicted based on the dynamic model, ensuring the vehicle can smoothly cope with upcoming changes in driving scenarios.

[0090] Optionally, the threshold can be set according to the vehicle type, driving scenario, and the requirements of various systems within the vehicle to ensure high energy efficiency while meeting safety and comfort requirements. To avoid sudden changes in control commands, filtering or smoothing techniques can be employed to ensure continuity and stability during vehicle control.

[0091] In this embodiment, when the initial acceleration deviates from the target acceleration by less than or equal to a threshold value, the control strategy based on the current pedal power conversion coefficient and friction should remain stable to avoid unnecessary parameter adjustments. This not only ensures that the vehicle's dynamic performance remains consistent with the expected target but also reduces uncertainties in the control process, improving the overall efficiency and accuracy of in-loop simulation testing. Furthermore, by continuously monitoring the vehicle's status and making timely predictive adjustments, stable and efficient vehicle operation can be achieved in various driving scenarios.

[0092] As an optional implementation, step S102 involves acquiring multi-source driving data of the vehicle, including: acquiring the load level corresponding to the load condition, the road adhesion coefficient corresponding to the road condition, and the pedal opening; in response to the vehicle being in a starting state and the pedal opening being fixed, acquiring the speed for a target time period under a fixed pedal opening, and the acceleration for a target time period under a fixed pedal opening.

[0093] In this embodiment, acquiring multi-source driving data of the vehicle can comprehensively capture the dynamic behavior characteristics of the vehicle under various operating conditions, especially for the three key variables: load condition, road surface condition, and pedal opening.

[0094] Optionally, load conditions (e.g., no load, half load, or full load) and road conditions (e.g., dry, wet, or icy / snowy roads) correspond to different load classes (M). k (k=1,2,3) and road surface adhesion coefficient (u) l (l=1,2,3). This is to account for the influence of vehicle mass and road conditions on vehicle dynamics in subsequent data collection and analysis. Afterwards, the pedal opening (B) of the vehicle's accelerator or brake pedal during the target time period can be collected. i Pedal opening is a crucial bridge connecting driver intent and vehicle dynamic response; therefore, meticulously recording subtle changes in pedal opening is essential for constructing accurate dynamic models.

[0095] Optionally, the vehicle's starting status can be determined based on the vehicle's engine status, vehicle speed, or other starting signals. When the vehicle is starting and the pedal opening is set to a fixed value, recording the vehicle's speed and acceleration at that specific pedal opening can begin immediately. To cover the entire range of the vehicle accelerating from a standstill to its maximum speed, the target time period (e.g., data collection duration) can be set to a sufficiently long duration, such as 30 seconds, and the sampling frequency can be sufficiently high, such as once every 0.1 seconds. This is merely an example and no specific limitations are imposed here.

[0096] Optionally, the collected data on load level, road surface adhesion coefficient, pedal opening, speed, and acceleration can be organized into a multidimensional data matrix. This multidimensional data matrix not only includes data from a single test but also summarizes the results of multiple tests, facilitating the analysis of vehicle dynamic behavior patterns under different operating conditions. Using this multidimensional data matrix, a vehicle dynamics model can be trained, focusing on the relationship between pedal opening and vehicle acceleration, and how this relationship is affected by load level and road surface adhesion coefficient. Through training the dynamics model, the acceleration trend of the vehicle under given operating conditions and pedal opening can be predicted.

[0097] For example, suppose within a time interval Within this context, K and C remain constant. This can be used to represent sampling time. The coefficients K and C for this time period are calculated based on data collected from the actual vehicle, for example, using the following formula. and :

[0098]

[0099] We can obtain the set of K coefficients for each piecewise function:

[0100]

[0101] Alternatively, C can be used for subsequent piecewise interpolation modeling.

[0102] Optionally, when the brake / accelerator pedal opening is fixed, based on the previously derived simplified dynamics model, within the speed range... acceleration .

[0103] Optionally, the above relationship of fixed pedal opening can be extended to a two-dimensional combination scenario of pedal opening range and velocity range. In this case, the acceleration formula is generalized to: .

[0104] Optionally, based on the above steps, a two-dimensional interpolation table for multiple working conditions can be generated (one interpolation table for each "load level-road condition" combination), enabling real-time query of acceleration for any combination of B and v.

[0105] Optionally, by providing real-time deviation feedback, the acceleration predicted by the interpolation table can be compared in real time during the algorithm-in-the-loop simulation. The actual output acceleration of the simulated actuator .

[0106] In this embodiment, the detailed data collection and processing procedures described above enable the construction of a comprehensive vehicle dynamics behavior database, providing a solid data foundation for the establishment and optimization of dynamic models. This not only helps improve the accuracy and reliability of vehicle-in-the-loop simulation testing but also provides crucial support for the development and optimization of autonomous driving algorithms.

[0107] As an optional implementation, in response to the vehicle being in a starting state and the pedal opening being fixed, acquiring the speed and acceleration for a target time period under the fixed pedal opening includes: in response to the target time period being greater than or equal to a time period threshold, or the speed being greater than or equal to a speed threshold, determining a first magnitude relationship between the pedal opening and a pedal opening threshold; in response to the first magnitude relationship being that the pedal opening is greater than or equal to the pedal opening threshold, determining a second magnitude relationship between the road surface adhesion coefficient and a road surface adhesion coefficient threshold; in response to the second magnitude relationship being that the road surface adhesion coefficient is greater than or equal to the road surface adhesion coefficient threshold and the load level is greater than or equal to a load level threshold, acquiring the speed and acceleration.

[0108] In this embodiment, when the vehicle is in the starting state and the pedal opening is fixed, a series of threshold judgment mechanisms can be used to ensure that speed and acceleration are accurately acquired and used under certain operating conditions.

[0109] Optionally, if the target time period is greater than or equal to a time period threshold, or the speed is greater than or equal to a speed threshold, a first relationship between the pedal opening and the pedal opening threshold can be determined. If the first relationship is that the pedal opening is greater than or equal to the pedal opening threshold, a second relationship between the road surface adhesion coefficient and the road surface adhesion coefficient threshold can be further determined. If the second relationship is that the road surface adhesion coefficient is greater than or equal to the road surface adhesion coefficient threshold, and the load level is greater than or equal to the load level threshold, then the current working condition is suitable for data acquisition and dynamic prediction, and speed and acceleration can be obtained.

[0110] In this embodiment, by introducing a judgment mechanism for pedal opening threshold, road surface adhesion coefficient threshold, and load level threshold, data acquisition and dynamic prediction can be ensured under sufficiently stable operating conditions. This avoids invalid or inaccurate predictions when the pedal opening is too small, the road surface adhesion coefficient is too low, or the load level is insufficient. Through the above steps, it is possible to identify which data points should be used to update the dynamic model, thereby optimizing the longitudinal motion control of the vehicle in the in-loop simulation scenario, ensuring good consistency between algorithm commands and actual power output, and improving the overall control accuracy and reliability of the vehicle.

[0111] As an optional implementation, step S106 compares the initial acceleration with the target acceleration to obtain a comparison result, including: determining the driving scenario of the vehicle; in response to the driving scenario being the target driving scenario, comparing the initial acceleration with the target acceleration to obtain a comparison result, wherein the target driving scenario is used to represent the driving scenario set in the simulation actuator.

[0112] In this embodiment, the comparison between the initial acceleration and the target acceleration depends not only on the numerical difference between the two, but also on the driving scenario in which the vehicle is operating. Therefore, the driving scenario can be determined first. The target driving scenario can be a typical ACC scenario.

[0113] Optionally, driving scenario recognition can be based on data from vehicle sensors, such as radar, cameras, Global Positioning System (GPS), and in-vehicle status sensors. For ACC systems, typical ACC scenarios can include following other vehicles, acceleration / deceleration, and stop-and-go driving.

[0114] Optionally, following the vehicle in front means the vehicle automatically adjusts its speed to maintain a safe distance. Acceleration / deceleration means the vehicle autonomously accelerates or decelerates to adapt to road speed limits or traffic conditions. Start-stop means the vehicle automatically restarts and moves forward after stopping at traffic lights or in congested situations.

[0115] Optionally, when the vehicle is identified as being in a target driving scenario, the system can begin to compare the predicted initial acceleration with the target acceleration. The target acceleration is the acceleration that the ACC system expects to output. It can be set according to the current driving scenario (e.g., distance to the vehicle in front, road conditions, etc.) to ensure optimal driving performance, such as maintaining a safe distance and adapting to changes in traffic flow.

[0116] Optionally, by calculating the absolute difference between the initial acceleration and the target acceleration. The comparison results quantify the difference between the initial acceleration predicted by the dynamic model and the target acceleration set by the ACC system.

[0117] In this embodiment, by comparing the initial acceleration with the target acceleration under specific driving scenarios, the predictive performance of the dynamics model can be evaluated in real time. Based on the comparison results, key parameters (such as the pedal power conversion coefficient and friction) can be adjusted to optimize the longitudinal control accuracy of the vehicle in loop simulation testing. This process not only improves the efficiency and accuracy of ACC system loop simulation testing but also reduces the cost of subsequent real-vehicle debugging and correction, providing strong support for the development of intelligent driving systems. Through continuous iteration and optimization, the optimal vehicle longitudinal dynamics control strategy can be gradually approximated, ensuring safe and efficient driving in various driving scenarios.

[0118] The vehicle control method of this invention acquires multi-source driving data of the vehicle. Since the multi-source driving data contains dynamic state parameters under different operating conditions, the initial acceleration of the vehicle can be obtained by predicting the multi-source driving data using the vehicle's dynamic model. Then, the initial acceleration is compared with the target acceleration determined by the vehicle's simulation actuator to determine the degree of deviation between the initial acceleration and the target acceleration. Based on this deviation, a control strategy is determined to control the vehicle's movement. In other words, this embodiment of the application achieves precise vehicle control by training a dynamic model using multi-dimensional data samples and combining it with real-time deviation feedback. Especially under complex and variable operating conditions, it significantly improves the reliability of vehicle simulation testing, reduces the cost and time of subsequent real-vehicle debugging, and overcomes the limitations of related technologies in simulation testing caused by a lack of multi-condition adaptability, coarse nonlinear relationship modeling, and a lack of real-time deviation feedback mechanisms, resulting in inaccurate vehicle control. Therefore, it solves the technical problem of ineffective vehicle control and achieves the technical effect of effective vehicle control.

[0119] The above technical solutions of the present application embodiments will be further illustrated below with reference to preferred embodiments of the present invention.

[0120] Currently, adaptive cruise control (ACC) and automatic emergency braking (AEB) are core functions with high practical value in the Advanced Driver Assistance Systems (ADAS) system. Their longitudinal control performance directly determines the safety, comfort, and energy economy of vehicle driving.

[0121] In related technologies, intelligent driving simulation testing serves as a crucial verification method in the early stages of algorithm development. The reliability of the test results depends entirely on the accuracy of parameter calibration between the throttle, brake actuators, and vehicle acceleration. Current traditional calibration methods suffer from the following technical shortcomings: limited data acquisition dimensions and lack of adaptability to multiple operating conditions; coarse modeling of nonlinear relationships and poor control continuity; lack of real-time deviation feedback and lag in parameter correction; and the absence of a comparative verification system between simulation data and real vehicle data, relying solely on internal simulation verification for calibration parameters. This frequently leads to situations where simulation results are met but real-vehicle failures occur, requiring the deployment of multiple test vehicles for real-vehicle debugging, increasing manpower and site costs by over 60%, and extending the debugging cycle by 2-3 times.

[0122] In order to accurately simulate the real power output characteristics of the test vehicle, this application provides a method for calibrating the longitudinal motion control parameters of the vehicle in an algorithm-in-the-loop simulation scenario.

[0123] Specifically, this application overcomes the limitations of traditional single-condition data acquisition by acquiring real-vehicle data, enabling parameter calibration under multiple loads and road surface adhesion coefficients, thus improving the adaptability of the calibration results to different operating conditions. By establishing a fusion dynamic model combining segmented modeling and interpolation smoothing, the relationship between vehicle acceleration and velocity is analyzed when the accelerator and brake pedal openings are fixed, resolving the problem of abrupt acceleration changes at segment transitions and ensuring continuous control. A real-time deviation feedback and parameter self-correction mechanism is designed to dynamically compensate for model deviations caused by changes in operating conditions, avoiding repetitive manual calibration. A two-dimensional interpolation table is generated and fitted to obtain the nonlinear relationship between brake / accelerator, velocity, and acceleration, providing a continuous and accurate longitudinal dynamic model for the algorithm and ensuring the smoothness of vehicle control. By constructing a simulation-real-vehicle data closed-loop verification system, the disconnect between simulation and real-vehicle data is reduced, lowering subsequent debugging costs.

[0124] Figure 2 This is a flowchart of a real vehicle data acquisition method according to an embodiment of this application, such as... Figure 2 As shown, the method may include the following steps.

[0125] Step S201: Perform data initialization and set the full load level M. k Road surface adhesion coefficient u l and brake / accelerator pedal opening B i .

[0126] In this embodiment, data initialization can be performed, and the full load level M can be set. k Road surface adhesion coefficient u l and brake / accelerator pedal opening B i .

[0127] Step S202, start the actual vehicle and input B i .

[0128] In this embodiment, the actual vehicle can be started, and the brake / accelerator pedal opening B can be input. i . It can be used to represent the set of brake / accelerator pedal openings.

[0129] Step S203, sample t=j Sampling time, j=1.

[0130] In this embodiment, t=j can be sampled. , j=1. Where, It can be used to represent sampling time. t can be used to represent the duration of brake / accelerator pedal action.

[0131] Step S204: Collect the current vehicle speed and current acceleration.

[0132] In this embodiment, the current vehicle speed can be collected. and current acceleration . It can be used to indicate the opening of the brake / accelerator pedal. At the moment of action Current acceleration. It can be used to indicate the opening of the brake / accelerator pedal. At the moment of action Current vehicle speed.

[0133] Step S205, j The sampling time is less than 30 seconds and the current vehicle speed is less than the vehicle's maximum speed.

[0134] In this embodiment, it can be determined whether j Sampling time < 30s and current vehicle speed < maximum vehicle speed If yes, proceed to step S206; otherwise, proceed to step S207.

[0135] Step S206, j = j + 1, return to C.

[0136] In this embodiment, if j If the sampling time is less than 30 seconds and the current vehicle speed is less than the maximum vehicle speed, then let j = j + 1 and return C, where C represents a constant.

[0137] Step S207: Save the data matrix.

[0138] In this embodiment, if j If the sampling time is ≥30s, or the current vehicle speed is ≥ the vehicle's maximum speed, then save the data matrix M. k -u l -B i . For full load levels (k=1,2,3), u l The coefficient of adhesion is l = 1, 2, 3.

[0139] Step S208, B i <B n .

[0140] In this embodiment, it can be determined whether B i <B n If yes, proceed to step S209; otherwise, proceed to step S210.

[0141] Step S209, B i =B i +5%, return B.

[0142] In this embodiment, if B i <B n Then let Bi =B i +5%, returns B, where B represents a constant.

[0143] Step S210, u l <u3.

[0144] In this embodiment, if B i ≥B n Then it can be determined whether u l <u3. If yes, proceed to step S211; otherwise, proceed to step S212.

[0145] Step S211, u l =u l +1, return A.

[0146] In this embodiment, if u l If u < u3, then let u l =u l +1, returns A, where A represents a constant.

[0147] Step S212, M k <M3.

[0148] In this embodiment, if u l If ≥u3, then it can be determined whether M is present. k <M3. If yes, proceed to step S213; otherwise, proceed to step S214.

[0149] Step S213, M k =M k +1, return A.

[0150] In this embodiment, if M k <M3, then let M k =M k +1, returns A, where A represents a constant.

[0151] Step S214: Data acquisition completed.

[0152] In this embodiment, if M k If the value is ≥M3, then data collection is complete.

[0153] In this embodiment of the invention, the planning and execution of the entire real vehicle data acquisition process described above can ensure the accuracy and integrity of the data, which is important for establishing a reliable vehicle dynamics model and improving the test effect of in-loop simulation.

[0154] Figure 3 This is a flowchart illustrating a method for calculating the accelerator / brake pedal power conversion coefficient and the current vehicle rolling friction force according to an embodiment of this application. Figure 3As shown, the method may include the following steps.

[0155] Step S301, i=1, B=B i .

[0156] In this embodiment, we can let i=1 and B=B i .

[0157] Step S302, B≤B n .

[0158] In this embodiment, it can be determined whether B≤B n If yes, proceed to step S304; otherwise, proceed to step S303.

[0159] Step S303, End.

[0160] In this embodiment, if B > B n If so, execution will end.

[0161] Step S304, j=1.

[0162] In this embodiment, if B≤B n If so, let j=1.

[0163] Step S305, .

[0164] In this embodiment, it can be determined by the above formula. and .

[0165] Step S306, .

[0166] In this embodiment, when determining and After that, it can be determined whether If yes, proceed to step S307; otherwise, proceed to step S308.

[0167] Step S307, i = i + 1.

[0168] In this embodiment, if Let i = i + 1, and then check again whether B ≤ B. n .

[0169] Step S308, j = j + 1.

[0170] In this embodiment, if Then let j = j + 1, and again determine the formula in step S305. and .

[0171] In this embodiment, by collecting real vehicle data and solving for K (accelerator / brake pedal power conversion coefficient) and C (current vehicle rolling friction), and considering the nonlinear relationships under multiple operating conditions, a more refined dynamic model can be established. This allows the dynamic model to better reflect the vehicle's true dynamic performance under different loads and road surface adhesion coefficients. The solved K and C values ​​are not only used for initial dynamic model building but also enable real-time deviation feedback and parameter self-correction mechanisms. This means that if a deviation is found between the predicted acceleration and the actual acceleration during simulation testing or actual vehicle driving, these deviations can be dynamically compensated by adjusting the K and C values, ensuring the accuracy and effectiveness of the dynamic model and avoiding the tedious manual calibration and error accumulation problems required by traditional methods.

[0172] Optionally, by using K and C values ​​based on actual data, the accuracy of simulation testing can be significantly improved. This is because the dynamic model more closely resembles the real physical characteristics of a vehicle, especially under multi-condition scenarios, such as different loads and road conditions. The dynamic model can better predict the vehicle's acceleration, reducing the difference between simulation and real-vehicle results and improving the reliability and credibility of the test.

[0173] Alternatively, by utilizing a closed-loop verification system combining simulation and real-vehicle data in the early stages, the cost and time required for later real-vehicle debugging can be significantly reduced. This is because most parameter optimization and verification can be completed in a simulation environment, avoiding extensive real-vehicle testing, reducing resource consumption, and shortening the product development cycle.

[0174] Optionally, because the solution process for K and C considers the influence of various working conditions, the dynamic model has strong adaptability. This means that whether in in-loop simulation or in real vehicle testing, the dynamic model can quickly adjust to changes in road conditions or in-vehicle loads, providing acceleration predictions that are closer to reality, thereby ensuring the stability and accuracy of the algorithm.

[0175] Optionally, the improved accuracy of the dynamic model and the existence of a dynamic self-correction mechanism can accelerate the algorithm iteration process. Feedback and improvement points can be obtained promptly after each simulation test, allowing for faster algorithm optimization and improved overall performance of the vehicle's intelligent driving system.

[0176] In summary, the iterative solution process of K and C can construct a more realistic and adaptable dynamic model, thereby improving the accuracy and efficiency of simulation testing, reducing the need for subsequent real-vehicle debugging, and promoting rapid iteration and optimization of the algorithm.

[0177] The following example, using a specific scenario, will provide further explanation.

[0178] Table 1 is a data and working condition setting table according to an embodiment of this application. As shown in Table 1, it describes the corresponding data and working condition settings of accelerator / brake pedal opening and vehicle speed and acceleration under different loads and different road surface adhesion coefficients.

[0179] Table 1 Data and Operating Condition Settings

[0180]

[0181] Optionally, for each "load-road surface-aperture" combination, record the current vehicle speed at time t=j×τ (j=1,2,...,300). and This forms a multi-dimensional data matrix. A multi-dimensional data matrix can be described by the following formula:

[0182]

[0183] Optionally, based on the principle of vehicle dynamic balance, a basic relationship can be established between acceleration, throttle / brake opening, and vehicle speed. and achievable .

[0184] Where P represents power, F represents driving force, M represents total vehicle mass, and f represents rolling friction. After coefficient transformation, the following is obtained: .

[0185] Where K can be used to represent the accelerator / brake pedal power conversion coefficient, and C can be used to represent the current rolling friction force of the vehicle.

[0186] Optionally, the above formula can be transformed using equivalent coefficients. Factors related to the power output of the accelerator and brake pedals can be combined into an accelerator / brake pedal power conversion coefficient K (which is strongly correlated with pedal opening, as it is determined by the pedal opening). Factors related to rolling friction can be combined into the current vehicle rolling friction force C (which comprehensively reflects the hindering effect of rolling friction on acceleration). Finally, a simplified acceleration formula is obtained. .

[0187] Alternatively, assume that within the time interval Within this time period, K and C remain constant. The coefficients K and C for this time period are calculated based on data collected from actual vehicles using the following formula:

[0188]

[0189] We can obtain the set of K coefficients for each piecewise function:

[0190]

[0191] Alternatively, C can be used for subsequent piecewise interpolation modeling.

[0192] Optionally, when the brake / accelerator pedal opening is fixed, based on the previously derived simplified dynamics model, within the speed range... acceleration .

[0193] Optionally, the above relationship of fixed pedal opening can be extended to a two-dimensional combination scenario of pedal opening range and velocity range. In this case, the acceleration formula is generalized to: .

[0194] Optionally, based on the above steps, a two-dimensional interpolation table for multiple working conditions can be generated (one interpolation table for each "load-road surface" combination), enabling real-time query of acceleration for any B and v combination.

[0195] Optionally, by providing real-time deviation feedback, the acceleration predicted by the interpolation table can be compared in real-time during the algorithm-in-the-loop simulation. The actual output acceleration of the simulated actuator Calculate the deviation .

[0196] Optionally, when When the percentage exceeds 5%, a correction mechanism is triggered. (If the target acceleration is too large), then the current K value can be decreased by 2% and the C value increased by 1%; if (If the target acceleration is too small), the current K value can be increased by 2% and the C value decreased by 1%; after correction, the two-dimensional interpolation table can be updated to ensure dynamic matching between the dynamic model and the simulation actuator.

[0197] Optionally, through closed-loop data verification using simulation and real vehicles, typical ACC scenarios are selected, such as following and stopping, and cutting in and out of the vehicle in front. In the simulation test, a calibrated interpolation table is used to record simulation data (vehicle speed, acceleration, pedal opening); the same scenario is reproduced in a proportional test track to collect real vehicle data. The deviation of key indicators between simulation and real vehicle data is calculated. If the deviation exceeds the standard, the process returns to the step of collecting supplementary operating condition data and iteratively optimizing the model.

[0198] In this embodiment, by acquiring multi-source driving data of the vehicle, which includes dynamic state parameters under different operating conditions, the initial acceleration of the vehicle can be obtained by predicting the multi-source driving data using the vehicle's dynamic model. Then, the initial acceleration is compared with the target acceleration determined by the vehicle's simulator to determine the degree of deviation between the initial acceleration and the target acceleration. Based on this deviation, a control strategy is determined to control the vehicle's movement. In other words, this embodiment achieves precise vehicle control by training a dynamic model using multi-dimensional data samples and combining real-time deviation feedback. Especially under complex and variable operating conditions, it significantly improves the reliability of vehicle simulation testing, reduces the cost and time of subsequent real-vehicle debugging, and overcomes the limitations of related technologies where inaccurate vehicle control is caused by a lack of multi-condition adaptability, coarse nonlinear relationship modeling, and a lack of real-time deviation feedback mechanisms during simulation testing. This solves the technical problem of ineffective vehicle control and achieves the technical effect of effective vehicle control.

[0199] According to an embodiment of this application, a vehicle control device is provided. It should be noted that the vehicle control device can be used to execute the above-described vehicle control method.

[0200] Figure 4 This is a schematic diagram of a vehicle control device according to an embodiment of this application, such as... Figure 4 As shown, the vehicle control device 400 may include: an acquisition unit 402, a prediction unit 404, a comparison unit 406, and a control unit 408.

[0201] The acquisition unit 402 is used to acquire multi-source driving data of the vehicle, wherein the multi-source driving data is used to represent the dynamic state parameters of the vehicle under different operating conditions.

[0202] The prediction unit 404 is used to input multi-source driving data into the vehicle's dynamic model for prediction to obtain the vehicle's initial acceleration. The dynamic model is trained using multi-source driving data samples of the vehicle sample and the initial acceleration samples corresponding to the multi-source driving data samples. The multi-source driving data samples are used to represent the dynamic state parameters of the vehicle sample under different operating conditions.

[0203] The comparison unit 406 is used to compare the initial acceleration with the target acceleration to obtain a comparison result. The target acceleration is determined by the vehicle's simulation actuator, and the comparison result is used to indicate the degree to which the initial acceleration deviates from the target acceleration.

[0204] The control unit 408 is used to determine the vehicle control strategy based on the comparison results, and to control the vehicle driving according to the control strategy.

[0205] Optionally, the dynamic state parameters include pedal opening, speed, and acceleration, and different operating conditions include load conditions and road surface conditions. The prediction unit 404 includes: a prediction subunit, used to input multi-source driving data into the dynamic model for prediction to obtain the vehicle's pedal power conversion coefficient and vehicle friction force; a generation subunit, used to generate a two-dimensional interpolation table based on the pedal power conversion coefficient and friction force, wherein the two-dimensional interpolation table is used to represent the correspondence between pedal opening, speed, load conditions, road surface conditions, and acceleration; and a first determination subunit, used to determine the initial acceleration based on the two-dimensional interpolation table.

[0206] Optionally, the vehicle control device 400 further includes: an adjustment subunit, configured to adjust the pedal power conversion coefficient and friction force in response to a comparison result that the initial acceleration deviates from the target acceleration by a degree greater than a degree threshold, thereby obtaining an adjustment result; an update subunit, configured to update the two-dimensional interpolation table based on the adjustment result; and a first determination subunit including: a second determination subunit, configured to determine the initial acceleration based on the updated two-dimensional interpolation table.

[0207] Optionally, the vehicle control unit 400 further includes an update unit for updating the dynamic model based on the adjustment results, and determining the initial acceleration according to the updated dynamic model.

[0208] Optionally, the control unit 408 includes a third determining subunit, used to determine a control strategy based on the pedal power conversion coefficient and friction force in response to a comparison result indicating that the degree of deviation of the initial acceleration from the target acceleration is less than or equal to a degree threshold.

[0209] Optionally, the acquisition unit 402 includes: a first acquisition subunit, used to acquire the load level corresponding to the load condition, the road adhesion coefficient corresponding to the road condition, and the pedal opening; and a second acquisition subunit, used to acquire the speed and acceleration of the target time period under a fixed pedal opening in response to the vehicle being in a starting state and the pedal opening being fixed.

[0210] Optionally, the second acquisition subunit includes: a fourth determining subunit, configured to determine a first magnitude relationship between the pedal opening and the pedal opening threshold in response to a target time period being greater than or equal to a time period threshold, or a speed being greater than or equal to a speed threshold; a fifth determining subunit, configured to determine a second magnitude relationship between the road surface adhesion coefficient and the road surface adhesion coefficient threshold in response to the first magnitude relationship being that the pedal opening is greater than or equal to the pedal opening threshold; and a third acquisition subunit, configured to acquire speed and acceleration in response to the second magnitude relationship being that the road surface adhesion coefficient is greater than or equal to the road surface adhesion coefficient threshold and the load level is greater than or equal to the load level threshold.

[0211] Optionally, the control unit 408 includes: a fifth determining subunit for determining the driving scenario of the vehicle; and a comparison subunit for comparing the initial acceleration with the target acceleration in response to the driving scenario being the target driving scenario, and obtaining a comparison result, wherein the target driving scenario represents the driving scenario set in the simulation actuator.

[0212] In the vehicle control device of this embodiment, the acquisition unit 402 acquires multi-source driving data of the vehicle, wherein the multi-source driving data is used to represent the dynamic state parameters of the vehicle under different operating conditions; the prediction unit 404 inputs the multi-source driving data into the dynamic model of the vehicle for prediction to obtain the initial acceleration of the vehicle, wherein the dynamic model is trained using multi-source driving data samples of the vehicle sample and the corresponding initial acceleration samples of the multi-source driving data samples, wherein the multi-source driving data samples are used to represent the dynamic state parameters of the vehicle sample under different operating conditions; the comparison unit 406 compares the initial acceleration with the target acceleration to obtain the comparison result, wherein the target acceleration is determined by the vehicle's simulation actuator, and the comparison result is used to indicate the degree to which the initial acceleration deviates from the target acceleration; the control unit 408 determines the vehicle control strategy based on the comparison result, and controls the vehicle to drive according to the control strategy, thereby solving the technical problem of ineffective vehicle control and achieving the technical effect of effective vehicle control.

[0213] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0214] Embodiments of this application also provide a vehicle, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods described in various embodiments of this application when it runs.

[0215] Embodiments of this application also provide a computer-readable storage medium including a stored executable program, wherein, when the executable program is running, it controls the device where the computer-readable storage medium is located to perform the methods of various embodiments of this application.

[0216] Embodiments of this application also provide a computer program product, including a computer program that, when executed by a processor, implements the methods of various embodiments of this application.

[0217] Embodiments of this application also provide a computer program product, including a non-volatile computer-readable storage medium for storing a computer program that, when executed by a processor, implements the methods in various embodiments of this application.

[0218] Embodiments of this application also provide a computer program that, when executed by a processor, implements the methods described in the various embodiments of this application.

[0219] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0220] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0221] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0222] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0223] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0224] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for controlling a vehicle, characterized in that, include: Acquire multi-source driving data of the vehicle, wherein the multi-source driving data is used to represent the dynamic state parameters of the vehicle under different operating conditions; The multi-source driving data is input into the vehicle's dynamic model for prediction to obtain the vehicle's initial acceleration. The dynamic model is trained using multi-source driving data samples of the vehicle sample and the initial acceleration samples corresponding to the multi-source driving data samples. The multi-source driving data samples are used to represent the dynamic state parameters of the vehicle sample under different operating conditions. The initial acceleration is compared with the target acceleration to obtain a comparison result, wherein the target acceleration is determined by the vehicle's simulation actuator, and the comparison result is used to indicate the degree to which the initial acceleration deviates from the target acceleration; Based on the comparison results, a control strategy for the vehicle is determined, and the vehicle is controlled to drive according to the control strategy.

2. The method according to claim 1, characterized in that, The dynamic state parameters include pedal opening, velocity, and acceleration. The different operating conditions include load conditions and road surface conditions. The multi-source driving data is input into the vehicle's dynamic model for prediction to obtain the vehicle's initial acceleration, including: The multi-source driving data is input into the dynamic model for prediction to obtain the vehicle's pedal power conversion coefficient and the vehicle's friction force. Based on the pedal power conversion coefficient and the friction force, a two-dimensional interpolation table is generated, wherein the two-dimensional interpolation table is used to represent the correspondence between the pedal opening, the speed, the load condition, the road surface condition and the acceleration; The initial acceleration is determined based on the two-dimensional interpolation table.

3. The method according to claim 2, characterized in that, The method further includes: In response to the comparison result that the initial acceleration deviates from the target acceleration by a greater than a threshold, the pedal power conversion coefficient and the friction force are adjusted to obtain the adjustment result; Based on the adjustment results, update the two-dimensional interpolation table; Determining the initial acceleration based on the two-dimensional interpolation table includes: The initial acceleration is determined based on the updated two-dimensional interpolation table.

4. The method according to claim 3, characterized in that, The method further includes: Based on the adjustment results, the dynamic model is updated, and the initial acceleration is determined according to the updated dynamic model.

5. The method according to claim 2, characterized in that, Based on the comparison results, the control strategy for the vehicle is determined, including: In response to the comparison result that the degree to which the initial acceleration deviates from the target acceleration is less than or equal to the degree threshold, the control strategy is determined based on the pedal power conversion coefficient and the friction force.

6. The method according to claim 2, characterized in that, Obtaining multi-source driving data of the vehicle, including: Obtain the load level corresponding to the load condition, the road adhesion coefficient corresponding to the road surface condition, and the pedal opening; In response to the vehicle being in a running state and the pedal opening being fixed, the speed for a target time period at the fixed pedal opening and the acceleration for the target time period at the pedal opening are obtained.

7. The method according to claim 6, characterized in that, In response to the vehicle being in a running state and the pedal opening being fixed, the speed for a target time period at the fixed pedal opening, and the acceleration for the target time period at the pedal opening, are obtained, including: In response to the target time period being greater than or equal to a time period threshold, or the speed being greater than or equal to a speed threshold, a first magnitude relationship between the pedal opening and the pedal opening threshold is determined; In response to the first size relationship being that the pedal opening is greater than or equal to the pedal opening threshold, a second size relationship between the road surface adhesion coefficient and the road surface adhesion coefficient threshold is determined; In response to the second magnitude relationship being that the road surface adhesion coefficient is greater than or equal to the road surface adhesion coefficient threshold and the load level is greater than or equal to the load level threshold, the speed and the acceleration are obtained.

8. The method according to any one of claims 1-7, characterized in that, The initial acceleration is compared with the target acceleration to obtain the comparison results, including: Determine the driving scenario of the vehicle; In response to the driving scenario being a target driving scenario, the initial acceleration is compared with the target acceleration to obtain the comparison result, wherein the target driving scenario is used to represent the driving scenario set in the simulation actuator.

9. A vehicle, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program, when running, performs the method according to any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored executable program, wherein, when the executable program is executed, it controls the device on which the storage medium is located to perform the method according to any one of claims 1 to 8.

Images