Vehicle control method and device, electronic equipment and storage medium

Abstract

The application relates to a vehicle control method and device, electronic equipment and a storage medium. In the vehicle driving process, multi-modal perception data of the vehicle is acquired, and panoramic description information describing the current driving scene of the vehicle is generated based on the multi-modal perception data. According to the panoramic description information, a basic chassis control strategy matched with the current driving scene is determined. Based on the basic chassis control strategy, a target chassis control parameter is determined. According to the target chassis control parameter, a corresponding chassis control operation of the chassis actuator of the vehicle is controlled. The multi-modal perception data of the vehicle is acquired to generate the panoramic description information, which breaks the limitation of the sensor perception dimension, and makes the cognition of the system to the driving scene more comprehensive and complete. Meanwhile, the matched basic chassis control strategy is determined according to the panoramic description information, which solves the core pain point that the chassis control strategy does not match the actual driving scene.

Description

Technical Field

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

[0002] In the field of vehicle chassis control technology, active control is gradually emerging. This method uses sensors such as vision and radar to perceive road conditions in advance and implements pre-aiming control based on active suspension or CDC (continuous damping control system).

[0003] However, existing active control technologies still have significant limitations. They lack a comprehensive understanding of driving scenarios and the ability to identify and actively adjust to complex and ever-changing driving scenarios. They are unable to cope with changing road and environmental conditions, which leads to a mismatch between chassis control strategies and actual driving scenarios. They are unable to make accurate real-time adjustments based on comprehensive factors such as road conditions, weather, traffic flow, and driver status. Summary of the Invention

[0004] This application provides a vehicle control method, device, electronic device, and storage medium to solve the technical problem that existing chassis control technologies have a single perception dimension of driving scenarios and lack the ability to comprehensively identify complex driving scenarios, which leads to a rigid chassis control strategy and a lack of active adjustment capabilities.

[0005] In a first aspect, this application provides a vehicle control method, the method comprising: During vehicle operation, multimodal perception data of the vehicle is acquired, and panoramic description information describing the current driving scenario of the vehicle is generated based on the multimodal perception data. Based on the panoramic description information, a basic chassis control strategy matching the current driving scenario is determined; Based on the aforementioned basic chassis control strategy, the target chassis control parameters are determined; Based on the target chassis control parameters, the chassis actuators of the vehicle are controlled to perform corresponding chassis control operations.

[0006] In one possible implementation, based on the basic chassis control strategy, the target chassis control parameters are determined, including: The default chassis control parameters in the basic chassis control strategy are determined as the target chassis control parameters.

[0007] In one possible implementation, based on the basic chassis control strategy, the target chassis control parameters are determined, including: Obtain personalized user information corresponding to the current driving scenario of the vehicle; Based on the user's personalized information, the default chassis control parameters in the basic chassis control strategy are adjusted to generate target chassis control parameters.

[0008] In one possible implementation, obtaining user-personalized information corresponding to the current driving scenario of the vehicle includes: Obtain user commands issued by the user in the current driving scenario; Based on the user's personalized information, the default chassis control parameters of the basic chassis control strategy are adjusted to generate target chassis control parameters, including: Based on the panoramic description information at the time the user command is issued, a basic chassis control strategy matching the driving scenario at the time the user command is issued is determined; According to the user instructions, the default chassis control parameters in the matching basic chassis control strategy are adjusted to generate target chassis control parameters.

[0009] In one possible implementation, obtaining user-personalized information corresponding to the current driving scenario of the vehicle includes: Based on the panoramic description information, retrieve historical preference data that matches the current driving scenario from the user's historical preference database; Based on the user's personalized information, the default chassis control parameters of the basic chassis control strategy are adjusted to generate target chassis control parameters, including: Based on the historical preference data, the default chassis control parameters in the basic chassis control strategy are adjusted to generate the target chassis control parameters.

[0010] In one possible implementation, the historical preference data is obtained in the following way: Obtain historical user commands and panoramic description information of the time when the historical user commands were issued; Based on the historical user instructions, the default chassis control parameters in the basic chassis control strategy at the time the historical user instructions were issued are adjusted to generate historical adjustment results. The panoramic description information at the time the historical user command was issued and the historical adjustment results are associated and stored in the historical preference database.

[0011] In one possible implementation, based on the user personalization information, the default chassis control parameters of the basic chassis control strategy are adjusted to generate target chassis control parameters, including: The user's personalized information and the panoramic description information are input together into the adaptive parameter adjuster; The adaptive parameter adjuster adjusts the default chassis control parameters within the preset parameter range corresponding to the basic chassis control strategy to generate the target chassis control parameters. The preset parameter range represents the safety constraint boundary that cannot be modified by the user.

[0012] Secondly, this application provides a vehicle control device, the device comprising: The panoramic description information generation module is used to acquire the multimodal perception data of the vehicle during vehicle operation, and generate panoramic description information describing the current driving scenario of the vehicle based on the multimodal perception data. The control strategy determination module is used to determine a basic chassis control strategy that matches the current driving scenario based on the panoramic description information. The target chassis control parameter determination module is used to determine the target chassis control parameters based on the basic chassis control strategy. The chassis control execution module is used to control the chassis actuators of the vehicle to perform corresponding chassis control operations according to the target chassis control parameters.

[0013] In one possible implementation, the target chassis control parameter determination module is specifically used for: The default chassis control parameters in the basic chassis control strategy are determined as the target chassis control parameters.

[0014] In one possible implementation, the target chassis control parameter determination module includes: A personalized information acquisition unit is used to acquire user personalized information corresponding to the current driving scenario of the vehicle. The target chassis control parameter generation unit is used to adjust the default chassis control parameters in the basic chassis control strategy based on the user's personalized information, and generate target chassis control parameters.

[0015] In one possible implementation, the personalized information acquisition unit is specifically used for: Obtain user commands issued by the user in the current driving scenario; The target chassis control parameter generation unit is specifically used for: Based on the panoramic description information at the time the user command is issued, a basic chassis control strategy matching the driving scenario at the time the user command is issued is determined; According to the user instructions, the default chassis control parameters in the matching basic chassis control strategy are adjusted to generate target chassis control parameters.

[0016] In one possible implementation, the personalized information acquisition unit is specifically used for: Based on the panoramic description information, retrieve historical preference data that matches the current driving scenario from the user's historical preference database; The target chassis control parameter generation unit is specifically used for: Based on the historical preference data, the default chassis control parameters in the basic chassis control strategy are adjusted to generate the target chassis control parameters.

[0017] In one possible implementation, the historical preference data is obtained in the following way: Obtain historical user commands and panoramic description information of the time when the historical user commands were issued; Based on the historical user instructions, the default chassis control parameters in the basic chassis control strategy at the time the historical user instructions were issued are adjusted to generate historical adjustment results. The panoramic description information at the time the historical user command was issued and the historical adjustment results are associated and stored in the historical preference database.

[0018] In one possible implementation, the target chassis control parameter generation unit is specifically used for: The user's personalized information and the panoramic description information are input together into the adaptive parameter adjuster; The adaptive parameter adjuster adjusts the default chassis control parameters within the preset parameter range corresponding to the basic chassis control strategy to generate the target chassis control parameters. The preset parameter range represents the safety constraint boundary that cannot be modified by the user.

[0019] Thirdly, this application provides an electronic device, including: a processor and a memory, wherein the processor is configured to execute a vehicle control program stored in the memory to implement the vehicle control method described in any one of the first aspects.

[0020] Fourthly, this application provides a storage medium storing one or more programs that can be executed by one or more processors to implement the vehicle control method described in any one aspect.

[0021] Compared with the prior art, the technical solution provided in this application has the following advantages: The method provided in this application acquires multimodal perception data of the vehicle during vehicle operation and generates panoramic description information describing the current driving scenario based on the multimodal perception data; determines a basic chassis control strategy matching the current driving scenario based on the panoramic description information; determines target chassis control parameters based on the basic chassis control strategy; and controls the vehicle's chassis actuators to perform corresponding chassis control operations based on the target chassis control parameters. By acquiring multimodal perception data of the vehicle to generate panoramic description information, the limitations of sensor perception dimensions are broken, allowing the system to have a more comprehensive and complete understanding of the driving scenario; at the same time, determining a matching basic chassis control strategy based on the panoramic description information solves the core pain point of mismatch between chassis control strategy and actual driving scenario; furthermore, by further determining the target chassis control parameters based on the basic chassis control strategy, the chassis control strategy becomes more accurate and fits the actual scenario requirements. Attached Figure Description

[0022] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0023] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] One or more embodiments are illustrated by way of example with reference numerals in the accompanying drawings. These illustrations do not constitute a limitation on the embodiments. Elements with the same reference numerals in the drawings are denoted as similar elements. Unless otherwise stated, the figures in the drawings are not to be limited by scale.

[0025] Figure 1 A flowchart illustrating an embodiment of a vehicle control method provided in this application; Figure 2 A flowchart illustrating another embodiment of the vehicle control method provided in this application; Figure 3 A flowchart illustrating an embodiment of another vehicle control method provided in this application; Figure 4 A block diagram of a vehicle control device provided in an embodiment of this application; Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0027] The following disclosure provides numerous different embodiments or examples for implementing various structures of this application. To simplify the disclosure, specific examples of components and arrangements are described below. These are merely examples and are not intended to limit the scope of this application. Furthermore, reference numerals and / or letters may be repeated in different examples. Such repetition is for simplification and clarity and does not in itself indicate a relationship between the various embodiments and / or arrangements discussed.

[0028] To address the technical problems of existing chassis control technologies, such as limited perception dimensions of driving scenarios and lack of comprehensive recognition capabilities for complex driving scenarios, leading to rigid chassis control strategies and a lack of proactive adjustment capabilities, this application provides a vehicle control method, device, electronic device, and storage medium. By acquiring multimodal perception data of the vehicle to generate panoramic description information, it breaks through the limitations of sensor perception dimensions, enabling the system to have a more comprehensive and complete understanding of driving scenarios. Simultaneously, it determines a matching basic chassis control strategy based on the panoramic description information, solving the core pain point of mismatch between chassis control strategies and actual driving scenarios. Furthermore, based on the basic chassis control strategy, it further determines target chassis control parameters, making the chassis control strategy more accurate and aligned with actual scenario requirements.

[0029] Figure 1 A flowchart illustrating an embodiment of a vehicle control method provided in this application includes the following steps: Step 101: During vehicle operation, acquire multimodal perception data of the vehicle and generate panoramic description information describing the current driving scenario of the vehicle based on the multimodal perception data.

[0030] Multimodal perception data can refer to information such as vision, point cloud, driver status, and vehicle status collected through multiple sensors (such as cameras, LiDAR, etc.).

[0031] Panoramic description information refers to the structured and semantic description of a driving scene generated by fusing and understanding multimodal data through a visual language model. It includes quantitative indicators such as road environment, weather, driver emotions, risk level, and causal relationships between various elements.

[0032] In one embodiment, multimodal perception data, including external visual data, in-vehicle driver data, and vehicle dynamic data, are collected in real time through a front-view camera, LiDAR, driver's camera, and vehicle network. All data is then timestamped via hardware and transmitted to the onboard computing unit via the CAN bus for integration and verification. The multimodal perception data is input into a multimodal perception model to obtain panoramic description information output by the model.

[0033] The aforementioned multimodal language model includes: a feature extraction and alignment unit, used to extract multimodal features from multimodal perception data and align the multimodal features; a scene description generation unit, used to gradually generate a structured scene description of the current driving scene from the aligned multimodal features through a decoder; and to convert the collected point cloud data into road surface roughness levels using a point cloud computing module; and a multimodal fusion unit, which fuses the road surface roughness levels and structured scene descriptions to output panoramic description information.

[0034] Specifically, the point cloud data collected by LiDAR is processed by an independent point cloud computing module to output road surface unevenness. The scene description generation unit (such as a visual language model) can generate a structured scene description based on the multimodal perception data solution and the point cloud data. In addition, the scene description generation unit is pre-trained on internet images and text and fine-tuned with a driving dataset, enabling it to identify scene elements and analyze causal relationships (such as rain causing slippery roads), thereby outputting more accurate scene-specific description information.

[0035] Specifically, external visual data can be acquired by the front-view camera capturing images of the road ahead; LiDAR collects point cloud data at a sampling frequency of 10Hz for road surface unevenness detection. Driver data can be acquired by the driver's camera capturing facial expressions and posture data, as well as voice commands. Vehicle dynamic data can be obtained through the vehicle network, such as data signals indicating the vehicle's current state, including vehicle speed, yaw rate, steering wheel angle, accelerator pedal opening, and brake pedal opening.

[0036] For example, when a vehicle is driving slowly in the city, the camera captures images of the asphalt road surface, rain, and slow-moving traffic; the lidar measures the point cloud height variance at 0.02 (low unevenness); the driver's camera identifies the driver's tense expression and focused gaze; and the vehicle network outputs a vehicle speed of 20 km / h and an accelerator pedal opening of 5%. The collected multimodal perception data is processed by a multimodal language model to generate descriptions: asphalt road surface (98%), rain, slippery rainy road surface, unevenness level 1, slow driving, tense emotions, focused, and risk level 2. The analysis also identifies that rain causes slippery road surfaces and traffic congestion causes driver tension.

[0037] Based on the above embodiments, the limitations of traditional single-sensor, single-type data can be overcome, integrating three types of data: external vehicle data, internal vehicle data, and vehicle data. This enables comprehensive perception of driving scenarios and avoids misjudgments due to missing data. Furthermore, it can analyze the causal relationships between elements, allowing the system to understand the causes of scenarios and improving its ability to recognize complex and multifaceted scenarios. For example, it can distinguish whether slippery road surfaces are caused by rainfall or water spraying, providing a more accurate basis for adjusting control strategies.

[0038] Step 102: Based on the panoramic description information, determine the basic chassis control strategy that matches the current driving scenario.

[0039] The basic chassis control strategy refers to a standardized chassis control scheme that is matched with the current driving scenario based on the panoramic description information. It includes the core control logic of suspension, steering, drive / braking, default parameter range and actuator coordination rules in this scenario. It is the basic framework of chassis control, preset by professional vehicle engineers, and has safety constraints that cannot be modified by the user.

[0040] A finite state machine is a finite state machine written by professional vehicle engineers during the vehicle testing phase. This finite state machine presets a variety of typical driving scenarios (such as city driving, city congestion, rain and snow, off-road extrication, comfortable long-distance driving, continuous bumps, high-speed cornering, etc.) and configures a corresponding basic chassis control strategy for each scenario.

[0041] In one embodiment, the panoramic description information is converted into a panoramic description vector of the driving scene. Combined with the current vehicle state estimation result, a finite state machine is used to match the basic chassis control strategy that matches the panoramic description information. In this embodiment, the finite state machine is no longer triggered by a single road condition, but determines the basic chassis control strategy that best fits the current driving scene through the comprehensive matching of multiple elements in the panoramic description vector (such as road surface slippage, driving risk level, and vehicle speed).

[0042] For example, when the panoramic description information is: road surface slipperiness: icy road surface, driving risk level 4, vehicle speed 60km / h, road surface unevenness level 2, the finite state machine comprehensively matches the rain and snow slippery scenario. The corresponding basic chassis control strategy is to increase the roof damping coefficient of the suspension to suppress sideslip, increase the trajectory tracking weight of the steering system, reduce the torque response sensitivity of the drive system, and shorten the braking response time of the braking system.

[0043] For another example, when the panoramic description information is: severe congestion, vehicle speed 10km / h, driver mood: irritable, road surface type: concrete, the finite state machine is matched to the urban congestion scenario. The corresponding basic chassis control strategy is to use a low damping coefficient for the suspension to ensure comfort, and to reduce the proportional gain / increase the integral gain of the PID (Proportional-Integral-Derivative) in the drive system. That is, to reduce the longitudinal chassis control parameters, so that the acceleration changes smoothly and avoids low-speed jerking.

[0044] Based on a finite state machine-based control strategy for pre-defined typical scenarios, this approach simplifies the real-time computation of the intelligent decision-making module, improves decision-making efficiency, and meets the real-time requirements of vehicle chassis control. It overcomes the limitations of traditional control strategies triggered by single-scenario features (such as vehicle speed alone), and through multi-element comprehensive matching of panoramic description vectors, enables the control strategy to be highly adapted to complex and multifaceted real-world driving scenarios.

[0045] Step 103: Determine the target chassis control parameters based on the basic chassis control strategy.

[0046] Default chassis control parameters refer to the baseline values ​​of chassis control parameters preset for the corresponding scenario in the basic chassis control strategy, which meet general driving needs. They cover longitudinal control parameters, lateral control parameters, vertical control parameters, etc., and are the basis for parameter adjustment.

[0047] The target chassis control parameters refer to the specific control parameter values ​​generated after confirming or adjusting the default chassis control parameters to adapt to the current driving scenario and the user's personalized needs. These values ​​can be directly sent to the chassis actuators and are the final execution basis for chassis control.

[0048] As one possible implementation, determining the target chassis control parameters based on the basic chassis control strategy can include: determining the default chassis control parameters in the basic chassis control strategy as the target chassis control parameters.

[0049] Specifically, after the intelligent decision-making module matches the basic chassis control strategy, if it does not obtain personalized user information (such as no special instructions from the user or the system not learning a specific driving style), it will directly use the default chassis control parameters in the strategy as the target chassis control parameters without additional adjustment, thus simplifying the decision-making process.

[0050] As another possible implementation, determining the target chassis control parameters based on the basic chassis control strategy may include: obtaining user-personalized information corresponding to the current driving scenario of the vehicle; adjusting the default chassis control parameters in the basic chassis control strategy according to the user-personalized information to generate the target chassis control parameters.

[0051] Personalized user information can refer to information that corresponds to the current driving scenario of the vehicle and reflects the user's driving needs and driving style. This includes user-inputted language commands (such as ultimate comfort or sporty handling), driving styles learned by the system from historical driving data (such as aggressive or stable driving), and driver state characteristics (such as emotions and driving habits).

[0052] Furthermore, based on user-personalized information, the default chassis control parameters of the basic chassis control strategy are adjusted to generate target chassis control parameters. This includes: inputting user-personalized information and panoramic description information into the adaptive parameter adjuster; the adaptive parameter adjuster adjusts the default chassis control parameters within the preset parameter range corresponding to the basic chassis control strategy to generate target chassis control parameters; wherein, the preset parameter range represents the safety constraint boundary that cannot be modified by the user.

[0053] The adaptive parameter adjuster can refer to the parameter adjustment module built into the intelligent decision module, which is based on language model and retrieval enhancement generation technology. It can adjust the default chassis control parameters in a personalized manner within the safety constraints according to the user's personalized information and panoramic description information.

[0054] The preset parameter range refers to the adjustable range set for each default chassis control parameter in the basic chassis control strategy. It is a safety constraint boundary that cannot be modified by the user, ensuring that the parameters still meet the vehicle dynamics safety requirements after adjustment, and avoiding driving hazards caused by over-adjustment.

[0055] Specifically, personalized user information is acquired through the in-vehicle voice interaction system. This personalized information, along with panoramic description information, is input into the adaptive parameter adjuster. The adaptive parameter adjuster analyzes the user's needs based on the personalized information and, within the preset parameter range corresponding to the basic chassis control strategy, performs personalized offset adjustments on the default chassis control parameters to generate target chassis control parameters. The adjusted parameters are strictly limited to the preset parameter range to ensure that safety constraints are not exceeded.

[0056] For example, suppose the user is in a continuous bumpy driving scenario and inputs a voice command to pursue ultimate comfort and reduce vehicle vibration. Based on this voice command and panoramic description information, the adaptive parameter adjuster adjusts the default control parameters in the basic control strategy within the preset parameter range, optimizes the suppression of sprung mass displacement and velocity, maximizes the reduction of vehicle vibration, and achieves the ultimate driving comfort requirement.

[0057] For another example, suppose we are currently in a high-speed cornering scenario. The default lateral control parameters are Q:R=7:3, and the preset range of Q:R is (5:5-9:1). Based on the acquired user personalization information, the system learns that the user's current need is an aggressive driving style, pursuing precise steering. Based on this need, the adaptive parameter adjuster increases the proportion of the trajectory weight Q within the preset range, and then generates the corresponding target chassis control parameters Q:R=8:2, improving the trajectory tracking accuracy during cornering and matching the aggressive driving style. This is merely an example, and the embodiments of this application do not impose any limitations.

[0058] The above embodiments balance safety and personalization. By using preset parameter ranges as unmodifiable safety constraints, they ensure that all adjusted target parameters meet vehicle dynamics safety requirements, avoiding driving risks such as rollover and skidding caused by over-adjustment. Within the safe range, personalized adjustment of chassis control parameters is achieved, allowing the chassis response to match different users' driving styles and real-time needs, thus enhancing the human-vehicle integration experience.

[0059] Step 104: Based on the target chassis control parameters, control the vehicle's chassis actuators to perform the corresponding chassis control operations.

[0060] Chassis actuators can refer to the hardware devices in a vehicle chassis that receive control commands and execute specific chassis adjustment actions. They are the physical carriers for implementing chassis control strategies. For example, they can include four main categories: suspension systems (continuously variable damping shock absorbers, active suspension), steering systems, and drive / braking systems.

[0061] Chassis control operation refers to the specific adjustment actions and coordinated control behaviors completed by the chassis actuators according to the target chassis control parameters. It includes suspension damping adjustment, active suspension power output, front wheel steering angle adjustment, drive torque distribution, braking torque adjustment, etc., ultimately achieving dynamic adjustment of vehicle chassis performance.

[0062] In one embodiment, the chassis domain controller receives the target chassis control parameters and adopts a control architecture that is decoupled longitudinally, laterally, and vertically. Based on the dynamic characteristics, the target chassis control parameters are calculated into specific control commands for each chassis actuator, and the chassis actuators are controlled to perform the corresponding chassis control operations.

[0063] Specifically, a longitudinal PID control algorithm can be used to achieve precise tracking control of the vehicle's target acceleration. This can be achieved based on the longitudinal chassis control parameters in the target chassis control parameters and the following formula (I), thus realizing the optimal longitudinal control strategy for different driving scenarios: Formula (1); Where e(t) is the difference between the target acceleration and the current acceleration, and T is the calculated total required torque (driving torque is positive, braking torque is negative). The proportional gain Kp affects the directness and agility of the torque response, the integral gain Ki is used to eliminate steady-state errors, and the derivative gain Kd is used to suppress excessively rapid changes in acceleration. Based on the longitudinal parameters (such as Kp, Ki, and Kd) in the target chassis control parameters, the optimal longitudinal control strategy for different driving scenarios is implemented.

[0064] Precise control of the front wheel steering angle is achieved based on lateral chassis control parameters, enabling the vehicle to travel along the desired trajectory. Optimal trajectory tracking is achieved by minimizing the following objective function, as detailed in Formula (II): Formula (II); in, This is the actual output trajectory. For reference trajectory, To control the increments, Q and R are the trajectory weight matrix and control weight matrix, respectively, and N1 and N2 are the prediction time domain and control time domain, respectively. Based on the lateral chassis control parameters (such as the Q:R ratio) in the target chassis control parameters, both trajectory tracking accuracy and driving stability are considered. J represents the objective function (or cost function) in the Model Predictive Control (MPC) algorithm, which is used to quantify the control performance. The core task of the MPC controller is to find a set of control increments that minimizes the value of J.

[0065] The vertical actuators in the chassis control mechanism include continuously variable damping shock absorbers and active suspension. In typical scenarios, continuous damping control can be performed based on the ceiling damping principle. The implementation principle of the ceiling damping control algorithm is shown in formula (III) below: Formula (III); in, Csky is the target damping force of the continuously variable damping shock absorber, and Csky is the ceiling damping coefficient. The vertical velocity of the sprung mass, based on the vertical chassis control parameters in the target chassis control parameters, such as the ceiling damping coefficient, effectively suppresses the pitch and roll of the vehicle body, improving ride comfort and driving stability.

[0066] For example, suppose the current target chassis control parameters are: longitudinal parameters Kp=3, Ki=3, Kd=2; lateral parameters Q:R=8:2; vertical parameter Csky=1000; then the chassis domain controller obtains the corresponding control actions based on the above target chassis control parameters, and then calls the drive system, braking system, steering system, and continuously variable damping shock absorbers in the chassis actuators to execute the corresponding control actions. Specifically, the drive system reduces torque response sensitivity (Kp decreases) to avoid wheel slippage; the braking system shortens braking response time to improve braking efficiency; the steering system increases trajectory tracking weight (Q:R=8:2) to ensure steering precision; the continuously variable damping shock absorber increases the damping coefficient (Csky=1000) to suppress vehicle sideslip and roll, and the coordinated action of each mechanism improves driving stability on slippery roads.

[0067] The above embodiments achieve unified coordination of various actuators through a chassis domain controller, avoiding chassis performance conflicts caused by independent adjustments of a single mechanism, and improving the overall integrity and stability of chassis control. For example, the coordinated action of steering and suspension during cornering ensures both precise steering and suppression of body roll. Simultaneously, combined with real-time updates of panoramic description information, continuous closed-loop rolling optimization of chassis control operations is performed, enabling rapid response to dynamic changes in driving scenarios. This ensures that chassis control is always highly adapted to the current scenario and user needs, improving driving safety and comfort.

[0068] The method provided in this application acquires multimodal perception data of the vehicle during driving and generates panoramic description information describing the current driving scenario based on the multimodal perception data. Based on the panoramic description information, a basic chassis control strategy matching the current driving scenario is determined. Based on the basic chassis control strategy, target chassis control parameters are determined. Based on the target chassis control parameters, the vehicle's chassis actuators are controlled to perform corresponding chassis control operations. By acquiring multimodal perception data of the vehicle to generate panoramic description information, the limitations of sensor perception dimensions are overcome, allowing the system to have a more comprehensive and complete understanding of the driving scenario. Simultaneously, determining a matching basic chassis control strategy based on the panoramic description information solves the core pain point of mismatch between chassis control strategies and actual driving scenarios. Furthermore, by further determining the target chassis control parameters based on the basic chassis control strategy, the chassis control strategy becomes more accurate and better suited to actual scenario requirements.

[0069] Figure 2 A flowchart illustrating another embodiment of the vehicle control method provided in this application is shown below. Figure 1 Based on the illustrated process, this section mainly describes how to adjust the corresponding basic chassis control strategy based on the personalized commands collected from users in real time, thereby obtaining target chassis control parameters that better match the user commands. This includes the following steps: Step 201: During vehicle operation, acquire multimodal perception data of the vehicle and generate panoramic description information describing the current driving scenario of the vehicle based on the multimodal perception data.

[0070] Step 202: Based on the panoramic description information, determine the basic chassis control strategy that matches the current driving scenario.

[0071] For steps 201-202 above, please refer to the above. Figure 1 The detailed description of the relevant embodiments will not be repeated here.

[0072] Step 203: Obtain user commands issued by the user in the current driving scenario.

[0073] Step 204: Based on the panoramic description information at the time the user command is issued, determine the basic chassis control strategy that matches the driving scenario at the time the user command is issued.

[0074] The following is a unified explanation of steps 203-204 above: User commands can refer to the subjective control intentions expressed by the driver in the current driving scenario through voice, gestures, or the central control interface, such as natural language commands like "I want to be more comfortable" or "Tighten the suspension a bit."

[0075] The panoramic description information at the moment the command is issued can refer to the structured scene description generated in real time by the panoramic cognition module at the moment the user issues the command, which includes elements such as road, weather, and driver status.

[0076] In one embodiment, the system collects driver voice commands via an in-vehicle microphone array, recognizes preset gesture commands via a driver's side camera, or receives command text via touch input on the central control screen. Upon receiving the command text, the system records the timestamp of the command issuance, retrieves the corresponding panoramic description information from the cache based on the timestamp, and obtains the basic chassis control strategy matching the panoramic description information from the finite state machine.

[0077] For example, suppose the vehicle was previously in a "high-speed driving" scenario. At time t2, the user issues the command to "turn on comfort mode". At this time, the vehicle has just entered the ramp. The panoramic description information at time t2 is ramp driving, speed 30km / h, curve, asphalt road surface. The basic strategy of the finite state machine secondary matching is "continuous curves" to replace the previous "high-speed driving" strategy. Subsequently, the comfort parameters are adjusted for the user based on the "continuous curves" strategy.

[0078] In the above embodiments, the system timestamps user commands to accurately match the panoramic description information at the moment the user command is issued, ensuring that the command is highly adapted to the current driving scenario. The accurately matched panoramic description information is then input into a finite state machine for secondary matching and confirmation of the basic chassis control strategy, rather than directly using the previous control strategy. This avoids strategy mismatch due to dynamic changes in the scenario and ensures a high degree of adaptation between the basic chassis control strategy and the current driving scenario.

[0079] Step 205: According to the user's instructions, adjust the default chassis control parameters in the matching basic chassis control strategy to generate the target chassis control parameters.

[0080] In one embodiment, the user command and the panoramic description information at the time of command issuance are input together to the adaptive parameter regulator; within the preset parameter safety range corresponding to the basic chassis control strategy, the adaptive parameter regulator adjusts the default parameters by offset. During the adjustment process, the adjustment experience under similar historical scenarios can be retrieved to improve the rationality and consistency of the adjustment, and the adjusted target chassis control parameters are output and sent to the chassis domain controller.

[0081] For example, in "high-speed cruise mode," the default k1 = 0.8; the user's command is "I want it to be smoother." Considering the "high-speed cruise, clear weather" scenario, the adaptive regulator adjusts k1 to 0.65 within a safe range (0.5, 1.2) (reducing response sensitivity to make the vehicle ride more smoothly), while keeping other default basic chassis control parameters unchanged, thus generating the target chassis control parameters. This is merely an example, and the embodiments in this application do not impose limitations.

[0082] Step 206: Based on the target chassis control parameters, control the vehicle's chassis actuators to perform the corresponding chassis control operations.

[0083] For step 206 above, please refer to the above. Figure 1 The detailed description of the relevant embodiments will not be repeated here.

[0084] Through the above Figure 2 The description of the illustrated embodiment utilizes a timestamp and cache matching mechanism to accurately retrieve panoramic description information at the moment the user command is issued. This ensures that the user command is highly bound to the current real-time driving scenario, avoiding parameter adjustments and scenario conflicts caused by dynamic changes in the scenario (such as entering a ramp at high speed or switching from congestion to free flow). Furthermore, all parameter adjustments are strictly limited to the safety parameters preset by the basic strategy. User commands cannot exceed the safety constraints, satisfying personalized needs while mitigating safety risks from over-adjustment through technical means (such as avoiding excessively soft suspension leading to increased body roll in cornering scenarios), thus ensuring the basic safety baseline for vehicle operation.

[0085] Figure 3 A flowchart illustrating another embodiment of the vehicle control method provided in this application is shown below. Figure 1 Based on the illustrated process, this section mainly describes how to adjust the corresponding basic chassis control strategy based on the user's historical preference data, thereby obtaining target chassis control parameters that better match the user's driving style or driving preferences. This includes the following steps: Step 301: During vehicle operation, acquire multimodal perception data of the vehicle and generate panoramic description information describing the current driving scenario of the vehicle based on the multimodal perception data.

[0086] Step 302: Based on the panoramic description information, determine the basic chassis control strategy that matches the current driving scenario.

[0087] For steps 301-302 above, please refer to the above. Figure 1 The detailed description of the relevant embodiments will not be repeated here.

[0088] Step 303: Based on the panoramic description information, retrieve historical preference data that matches the current driving scenario from the user's historical preference database.

[0089] In one embodiment, historical preference data is obtained by: acquiring historical user commands and panoramic description information at the time the historical user commands were issued; adjusting the default chassis control parameters in the basic chassis control strategy at the time the historical user commands were issued according to the historical user commands, generating historical adjustment results; and storing the panoramic description information at the time the historical user commands were issued and the historical adjustment results in a historical preference database.

[0090] Historical preference data can be used to store users' historical adjustment records in different driving scenarios, including historical panoramic description information, historical user commands, and historical parameter adjustment results. The historical adjustment database can bind the above-mentioned historical panoramic description information, historical user commands, and historical adjustment results to establish a structured historical preference database. This historical preference database supports classification and indexing according to driving scenario characteristics, so as to quickly find historical preference data that matches the current driving scenario.

[0091] Specifically, after generating panoramic description information for the current driving scene, the system uses this information as a query condition to perform a similarity search in the historical preference database. Through vector similarity matching (such as cosine similarity) or rule matching, it identifies the most similar historical scene and its corresponding adjustment result. If multiple similar scenes exist, weighted fusion or the best-matching record can be selected.

[0092] For example, last week, in a driving scenario of "driving at high speed in the rain," a user used the voice command "soften the suspension." The system then reduced the default damping coefficient in the basic strategy "city congestion mode" to 2500 and associated the panoramic description information, driving scenario, and adjustment result into the historical preference database. When the vehicle is in a "driving at high speed in the rain" scenario again, after the panoramic description information is generated, the system retrieves the corresponding historical preference data from the historical preference database based on this information.

[0093] Step 304: Based on historical preference data, adjust the default chassis control parameters in the basic chassis control strategy to generate target chassis control parameters.

[0094] In one embodiment, the retrieved historical preference data (adjustment results) is used as a reference to offset the default parameters in the current basic chassis control strategy and generate target chassis control parameters.

[0095] Specifically, if the historical scenario is highly similar to the current scenario, the historical adjustment result can be directly used as the target parameter (within a safe range). If multiple similar historical scenarios exist, a weighted average of the multiple historical adjustment results can be performed, with the weight determined by the scenario similarity. Furthermore, the adjustment process must ensure that the target parameter always remains within the preset safe range and cannot be modified by the user beyond the limit. The adjusted default chassis control parameters are then determined as the target chassis control parameters and sent to the chassis domain controller for execution.

[0096] For example, the current scenario's panoramic description information is: high-speed driving in rain, speed 100km / h, asphalt road surface, driving risk level 2. The matched basic chassis control strategy has a default damping coefficient of 3000 (safe range: 2000-4000). The system retrieves corresponding historical data from the historical preference database: in the same scenario before, the user adjusted the default damping coefficient to 2500 using the "soften suspension" command. Because the current scenario is highly similar to the historical scenario, and 2500 is within the safe range, this historical adjustment result is directly reused, determining the target chassis control parameter as a damping coefficient of 2500, while keeping other default parameters unchanged, and then issuing it to the chassis domain controller for execution. This is just an example of a direct reuse adjustment method. In addition, when multiple similar historical scenarios are found, a weighted fusion method can be used to determine the final adjustment range, thereby generating the adjusted target chassis control parameters.

[0097] Step 305: Based on the target chassis control parameters, control the vehicle's chassis actuators to perform the corresponding chassis control operations.

[0098] For step 305 above, please refer to the above. Figure 1 The detailed description of the relevant embodiments will not be repeated here.

[0099] Through the above Figure 3 The embodiments described herein demonstrate that the system automatically adjusts parameters by matching historical preference data, eliminating the need for real-time user commands. This avoids users repeatedly expressing the same needs, reducing operational costs, and is particularly suitable for long-distance driving and high-frequency driving scenarios. Furthermore, the adjustments are based on the user's long-term driving preferences, with parameter settings more closely matching the user's fixed driving style (such as stable or aggressive), achieving personalized chassis control tailored to each individual and improving driver-vehicle integration. In addition, the historical preference database is dynamically updated with the user's driving behavior, allowing the system to continuously learn changes in the user's driving style (such as a user gradually shifting from aggressive to stable driving), enabling long-term iterative optimization of the control strategy.

[0100] Figure 4 A block diagram of a vehicle control device provided in this application embodiment, the device comprising: The panoramic description information generation module 41 is used to acquire the multimodal perception data of the vehicle during vehicle driving, and generate panoramic description information describing the current driving scene of the vehicle based on the multimodal perception data. The control strategy determination module 42 is used to determine a basic chassis control strategy that matches the current driving scenario based on the panoramic description information. The target chassis control parameter determination module 43 is used to determine the target chassis control parameters based on the basic chassis control strategy. The chassis control execution module 44 is used to control the chassis actuator of the vehicle to perform corresponding chassis control operations according to the target chassis control parameters.

[0101] In one possible implementation, the target chassis control parameter determination module 43 is specifically used for: The default chassis control parameters in the basic chassis control strategy are determined as the target chassis control parameters.

[0102] In one possible implementation, the target chassis control parameter determination module 43 includes: A personalized information acquisition unit is used to acquire user personalized information corresponding to the current driving scenario of the vehicle. The target chassis control parameter generation unit is used to adjust the default chassis control parameters in the basic chassis control strategy based on the user's personalized information, and generate target chassis control parameters.

[0103] In one possible implementation, the personalized information acquisition unit is specifically used for: Obtain user commands issued by the user in the current driving scenario; The target chassis control parameter generation unit is specifically used for: Based on the panoramic description information at the time the user command is issued, a basic chassis control strategy matching the driving scenario at the time the user command is issued is determined; According to the user instructions, the default chassis control parameters in the matching basic chassis control strategy are adjusted to generate target chassis control parameters.

[0104] In one possible implementation, the personalized information acquisition unit is specifically used for: Based on the panoramic description information, retrieve historical preference data that matches the current driving scenario from the user's historical preference database; The target chassis control parameter generation unit is specifically used for: Based on the historical preference data, the default chassis control parameters in the basic chassis control strategy are adjusted to generate the target chassis control parameters.

[0105] In one possible implementation, the historical preference data is obtained in the following way: Obtain historical user commands and panoramic description information of the time when the historical user commands were issued; Based on the historical user instructions, the default chassis control parameters in the basic chassis control strategy at the time the historical user instructions were issued are adjusted to generate historical adjustment results. The panoramic description information at the time the historical user command was issued and the historical adjustment results are associated and stored in the historical preference database.

[0106] In one possible implementation, the target chassis control parameter generation unit is specifically used for: The user's personalized information and the panoramic description information are input together into the adaptive parameter adjuster; The adaptive parameter adjuster adjusts the default chassis control parameters within the preset parameter range corresponding to the basic chassis control strategy to generate the target chassis control parameters. The preset parameter range represents the safety constraint boundary that cannot be modified by the user.

[0107] like Figure 5 As shown in the figure, this application provides an electronic device, including a processor 111, a communication interface 112, a memory 113, and a communication bus 114, wherein the processor 111, the communication interface 112, and the memory 113 communicate with each other through the communication bus 114. Memory 113 is used to store computer programs; In one embodiment of this application, when the processor 111 executes the program stored in the memory 113, it implements the vehicle control method provided in any of the foregoing method embodiments, including: During vehicle operation, multimodal perception data of the vehicle is acquired, and panoramic description information describing the current driving scenario of the vehicle is generated based on the multimodal perception data. Based on the panoramic description information, a basic chassis control strategy matching the current driving scenario is determined; Based on the aforementioned basic chassis control strategy, the target chassis control parameters are determined; Based on the target chassis control parameters, the chassis actuators of the vehicle are controlled to perform corresponding chassis control operations.

[0108] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the vehicle control method provided in any of the foregoing method embodiments.

[0109] The device embodiments described above are merely illustrative. 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 network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0110] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented using software plus a general-purpose hardware platform, or of course, using hardware. Based on this understanding, the above technical solutions, in essence or the parts that contribute to the related technology, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0111] It should be understood that the terminology used herein is for the purpose of describing particular exemplary embodiments only and is not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms “a,” “an,” and “described” as used herein may also mean including the plural forms. The terms “comprising,” “including,” “containing,” and “having” are inclusive and therefore indicate the presence of the stated features, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, elements, components, and / or combinations thereof. The method steps, processes, and operations described herein are not construed as requiring them to be performed in a particular order described or illustrated unless the order of performance is explicitly indicated. It should also be understood that additional or alternative steps may be used.

[0112] The above description is merely a specific embodiment of this application, enabling those skilled in the art to understand or implement this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A vehicle control method, characterized in that, The method includes: During vehicle operation, multimodal perception data of the vehicle is acquired, and panoramic description information describing the current driving scenario of the vehicle is generated based on the multimodal perception data. Based on the panoramic description information, a basic chassis control strategy matching the current driving scenario is determined; Based on the aforementioned basic chassis control strategy, the target chassis control parameters are determined; Based on the target chassis control parameters, the chassis actuators of the vehicle are controlled to perform corresponding chassis control operations.

2. The method according to claim 1, characterized in that, Based on the multimodal perception data, a panoramic description of the vehicle's current driving scenario is generated, including: The multimodal sensing data is input into the multimodal sensing model so that the multimodal sensing model outputs panoramic description information; The multimodal sensing model includes: The feature extraction and alignment unit is used to extract multimodal features from multimodal sensing data and align the multimodal features; The scene description generation unit is used to gradually generate a structured scene description of the current driving scene from the aligned multimodal features through the decoder; and to use the point cloud computing module to convert the collected point cloud data into road surface unevenness level. The multimodal fusion unit fuses the road surface unevenness level and the structured scene description to output panoramic description information.

3. The method according to claim 1, characterized in that, Based on the aforementioned basic chassis control strategy, the target chassis control parameters are determined, including: Obtain personalized user information corresponding to the current driving scenario of the vehicle; Based on the user's personalized information, the default chassis control parameters in the basic chassis control strategy are adjusted to generate target chassis control parameters.

4. The method according to claim 3, characterized in that, Obtain personalized user information corresponding to the current driving scenario of the vehicle, including: Obtain user commands issued by the user in the current driving scenario; Based on the user's personalized information, the default chassis control parameters of the basic chassis control strategy are adjusted to generate target chassis control parameters, including: Based on the panoramic description information at the time the user command is issued, a basic chassis control strategy matching the driving scenario at the time the user command is issued is determined; According to the user instructions, the default chassis control parameters in the matching basic chassis control strategy are adjusted to generate target chassis control parameters.

5. The method according to claim 3, characterized in that, Obtain personalized user information corresponding to the current driving scenario of the vehicle, including: Based on the panoramic description information, retrieve historical preference data that matches the current driving scenario from the user's historical preference database; Based on the user's personalized information, the default chassis control parameters of the basic chassis control strategy are adjusted to generate target chassis control parameters, including: Based on the historical preference data, the default chassis control parameters in the basic chassis control strategy are adjusted to generate the target chassis control parameters.

6. The method according to claim 5, characterized in that, The historical preference data was obtained in the following way: Obtain historical user commands and panoramic description information of the time when the historical user commands were issued; Based on the historical user instructions, the default chassis control parameters in the basic chassis control strategy at the time the historical user instructions were issued are adjusted to generate historical adjustment results. The panoramic description information at the time the historical user command was issued and the historical adjustment results are associated and stored in the historical preference database.

7. The method according to claim 3, characterized in that, Based on the user's personalized information, the default chassis control parameters of the basic chassis control strategy are adjusted to generate target chassis control parameters, including: The user's personalized information and the panoramic description information are input together into the adaptive parameter adjuster; The adaptive parameter adjuster adjusts the default chassis control parameters within the preset parameter range corresponding to the basic chassis control strategy to generate the target chassis control parameters. The preset parameter range represents the safety constraint boundary that cannot be modified by the user.

8. A vehicle control device, characterized in that, The device includes: The panoramic description information generation module is used to acquire the multimodal perception data of the vehicle during vehicle operation, and generate panoramic description information describing the current driving scenario of the vehicle based on the multimodal perception data. The control strategy determination module is used to determine a basic chassis control strategy that matches the current driving scenario based on the panoramic description information. The target chassis control parameter determination module is used to determine the target chassis control parameters based on the basic chassis control strategy. The chassis control execution module is used to control the chassis actuators of the vehicle to perform corresponding chassis control operations according to the target chassis control parameters.

9. An electronic device, characterized in that, include: A processor and a memory, the processor being configured to execute a vehicle control program stored in the memory to implement the vehicle control method according to any one of claims 1-7.

10. A storage medium, characterized in that, The storage medium stores one or more programs, which can be executed by one or more processors to implement the vehicle control method according to any one of claims 1-7.

Images