Vehicle control method, device and vehicle

Abstract

The embodiment of the application relates to the technical field of vehicles, and discloses a vehicle control method and device and a vehicle, the method comprising: when a subsidence working condition appears in front of a road surface, performing fusion processing on obtained cloud road surface subsidence data, self-vehicle sensing data and current vehicle state data to obtain a subsidence scene model corresponding to the subsidence working condition; determining suspension parameters of each wheel in each subsidence driving stage according to the subsidence scene model and the current vehicle state data; wherein the subsidence driving stage represents a driving stage when the vehicle passes through the subsidence working condition; and performing staged control on the suspension of the vehicle based on the suspension parameters of each wheel in each subsidence driving stage. The technical scheme of the application can solve the problem of poor driving comfort and poor driving stability of a vehicle when passing through a subsidence working condition in the prior art.

Description

Technical Field

[0001] This invention relates to the field of vehicle technology, specifically to a vehicle control method, device, and vehicle. Background Technology

[0002] With the acceleration of urbanization and the increase in the service life of roads, road surface subsidence (such as potholes, collapses, cracks, etc.) has become a typical complex working condition that affects the safety and comfort of vehicle driving.

[0003] Currently, the main approach relies on onboard sensors (such as accelerometers and height sensors) to provide real-time feedback on current road conditions, and passively responds to bumps by adjusting damping force or spring stiffness. However, this method has a delayed response to the impact of road subsidence, which can easily lead to severe vehicle vibration, wheel lift-off, and even damage to chassis components, resulting in poor ride comfort and driving stability when driving through subsidence conditions. Summary of the Invention

[0004] In view of the above problems, embodiments of the present invention provide a vehicle control method, device and vehicle to solve the problem of poor driving comfort and driving stability of vehicles in the prior art when passing through subsidence conditions.

[0005] According to one aspect of the present invention, a vehicle control method is provided, the method comprising:

[0006] When a road subsidence occurs ahead, the acquired cloud-based road subsidence data, vehicle perception data, and current vehicle status data are fused together to obtain a subsidence scenario model corresponding to the subsidence condition.

[0007] Based on the subsidence scenario model and the current vehicle status data, the suspension parameters of each wheel in each subsidence driving stage are determined; wherein, the subsidence driving stage represents the driving stage when the vehicle passes through the subsidence condition.

[0008] The vehicle's suspension is controlled in stages based on the suspension parameters of each wheel during each sinking driving phase.

[0009] According to another aspect of the present invention, a vehicle control device is provided, comprising:

[0010] The fusion module is used to fuse the acquired cloud-based road subsidence data, vehicle perception data, and current vehicle status data when a subsidence condition occurs on the road ahead, in order to obtain a subsidence scenario model corresponding to the subsidence condition.

[0011] The determination module is used to determine the suspension parameters of each wheel in each sinking driving stage based on the sinking scenario model and the current vehicle status data; wherein, the sinking driving stage represents the driving stage of the vehicle when passing through the sinking condition.

[0012] The control module is used to perform phased control of the vehicle's suspension based on the suspension parameters of each wheel during each sinking driving stage.

[0013] According to another aspect of the present invention, an electronic device is provided, including: a processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other through the communication bus;

[0014] The memory is used to store at least one executable instruction that causes the processor to perform the operation of the vehicle control method described above.

[0015] According to another aspect of the present invention, a vehicle is provided, comprising: a vehicle body and an electronic device, the electronic device being used to perform the operation of the vehicle control method described above.

[0016] According to another aspect of the present invention, a computer-readable storage medium is provided, the storage medium storing at least one executable instruction that causes an electronic device / vehicle control device to perform the following operations:

[0017] When a road subsidence occurs ahead, the acquired cloud-based road subsidence data, vehicle perception data, and current vehicle status data are fused together to obtain a subsidence scenario model corresponding to the subsidence condition.

[0018] Based on the subsidence scenario model and the current vehicle status data, the suspension parameters of each wheel are determined in each subsidence driving stage; wherein, the subsidence driving stage represents the driving stage of the vehicle when passing through the subsidence condition.

[0019] The vehicle's suspension is controlled in stages based on the suspension parameters of each wheel during each sinking driving phase.

[0020] When a road subsidence occurs ahead, this invention first constructs a high-precision subsidence scenario model by fusing prior road data from the cloud, real-time vehicle-mounted perception data, and vehicle status data. This accurately recreates the true attributes of the subsided road surface. Then, based on the vehicle's real-time status, it adaptively matches exclusive suspension control parameters for different driving stages and different wheels. Finally, relying on time-sequential, phased control logic, it completes refined dynamic control of the four-wheel suspension. This abandons the passive, lagging control mode of traditional suspensions that solely rely on vehicle-mounted sensors. It leverages the advantages of multi-source fusion perception from the vehicle and cloud to compensate for the low accuracy and weak anti-interference capabilities of single visual perception. It predicts subsidence conditions in advance and achieves differentiated, phased active control of the four wheels, effectively mitigating the impact disturbance when the vehicle travels on a subsided road surface, suppressing attitude fluctuations such as pitch and roll, and significantly improving the chassis control accuracy and robustness under severe subsidence conditions. While ensuring vehicle driving stability and chassis structural safety, it significantly optimizes vehicle ride comfort.

[0021] The above description is merely an overview of the technical solutions of the embodiments of the present invention. In order to better understand the technical means of the embodiments of the present invention and to implement them in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the embodiments of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0022] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0023] Figure 1 A system architecture diagram provided by the present invention is shown;

[0024] Figure 2 A flowchart of a first embodiment of the vehicle control method provided by the present invention is shown;

[0025] Figure 3 A flowchart of a second embodiment of the vehicle control method provided by the present invention is shown;

[0026] Figure 4 A schematic diagram of an application scenario provided by the present invention is shown;

[0027] Figure 5 A schematic diagram of an embodiment of the vehicle control device provided by the present invention is shown;

[0028] Figure 6 A schematic diagram of an embodiment of the electronic device provided by the present invention is shown. Detailed Implementation

[0029] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0030] With the rapid iteration of automotive electrification and intelligent technologies, users' comprehensive requirements for vehicle ride comfort, driving safety, and chassis stability continue to rise. The steerable suspension system, relying on the synergistic configuration of air springs and Continuous Damping Control (CDC) dampers, can flexibly and dynamically adjust suspension stiffness, damping, and vehicle height, achieving a dynamic balance between vehicle handling performance and ride comfort. It has now become a core configuration for improving the chassis and driving quality of mid-to-high-end intelligent electric vehicles.

[0031] Existing drive-by-wire suspension systems largely rely on real-time vehicle status signals collected by onboard sensors, providing only passive and immediate responses to current road surface stimuli. When facing sudden, large-deformation, and long-travel road conditions such as road subsidence, traditional passive response control methods exhibit significant predictive deficiencies and response lags. When a vehicle travels at medium to high speeds through a subsided area, the severe road impact significantly deteriorates ride comfort and can easily cause fatigue damage to chassis components. In extreme cases, it can also lead to momentary wheel lift-off and sudden changes in ground load, greatly increasing the risk of loss of vehicle control.

[0032] Among related technologies, intelligent predictive suspension solutions, represented by "magic carpet suspension," primarily rely on onboard cameras to perceive the undulations of the road surface ahead and make pre-adjustments to the suspension. However, this type of stand-alone visual perception solution is greatly affected by external factors such as lighting, weather, shooting distance, and road obstructions. For complex and harsh conditions such as deep potholes and large-area subsidence, it cannot accurately identify key features such as subsidence depth, geometric dimensions, and edge slope, resulting in insufficient perception accuracy and robustness in adapting to working conditions.

[0033] Therefore, existing suspension control technologies suffer from problems such as delayed passive control response and poor perception accuracy, resulting in poor ride comfort and driving stability when vehicles are traversing sinkholes. The inventors recognized that onboard cameras can only identify the road surface at close range ahead of the vehicle and lack the ability to anticipate the location and shape of sinkholes from a distance. Cloud-based crowdsourced road maps, however, aggregate a large amount of road surface sinkhole annotation information collected and continuously updated by vehicles over a long period. This allows for the output of complete prior geometric data on sinkholes even when the vehicle is far from the sinkhole area, thus compensating for the limited perception range of onboard cameras. Therefore, instead of using a single perception source for control, the inventors integrated cloud-based road surface sinkhole data, vehicle perception data, and vehicle status data to construct a precise sinkhole scenario model. Based on the temporal driving characteristics of traversing sinkholes, differentiated, staged suspension parameters are matched to each wheel. Through a staged and refined active suspension control method, the inventors fundamentally solve the problems of delayed control, limited perception dimensions, and poor adaptability of traditional suspension systems, effectively improving driving stability and ride comfort when vehicles are traversing sinkholes.

[0034] The execution subject of this invention can be a vehicle controller, such as an existing controller in the vehicle or a separately set controller or domain controller, such as a suspension controller or chassis domain controller, or an additional controller. The controller can be an electronic control unit (ECU), a microcontroller unit (MCU), etc., and this invention does not limit the scope of the invention.

[0035] Figure 1 A system architecture diagram provided by this invention is shown. Figure 1 As shown, the suspension control system includes a perception layer module, a decision layer module, and an execution layer module.

[0036] The perception layer module includes a vehicle-to-cloud map module, a camera perception module, and a vehicle status sensor module.

[0037] (1) Vehicle Cloud Map Module: Includes vehicle network T-BOX (a vehicle network terminal used to establish a data communication link between the vehicle and the cloud server to realize downlink of road condition data and uplink of vehicle status information), cloud road condition server (a dedicated server deployed in the cloud to store and push chassis-level road condition data such as road subsidence and smoothness to provide a long-distance forward road condition prediction data source for vehicles) and GPS / Beidou positioning module, with a data update frequency of 10Hz and a maximum prediction distance of 200m, used to synchronize historical road subsidence points and road surface smoothness data;

[0038] (2) Camera perception module: A binocular / tri-lens camera (installed above the windshield of the vehicle, facing the front of the vehicle) is required to detect a distance of 5-20m, with a subsidence recognition accuracy of ±2cm, and output the distance from the wheel to the subsidence and the relevant physical parameters of the road subsidence (including the length, width and actual depth of the subsidence).

[0039] (3) Vehicle status sensor module: includes vehicle height sensor, vehicle acceleration sensor or inertial measurement unit (IMU) sensor, sampling frequency 100Hz-500Hz, to collect real-time data such as vehicle vibration, shock absorber displacement, and vehicle attitude.

[0040] The decision-making module includes the suspension central controller, which is equipped with an automotive-grade high-performance processing chip with a main operating frequency of ≥1.2GHz and an internal control latency of ≤20ms. It has the ability to acquire multi-channel signals in parallel and can simultaneously receive cloud communication data, visual image parameters, and vehicle posture sensing signals. It undertakes all logical functions such as data reception and parsing, fusion calculation, working condition determination, suspension control decision-making, and adjustment command issuance.

[0041] The execution layer module includes a drive-by-wire suspension module, which includes air springs and continuous damping control (CDC) / magnetic ride control (MRC) dampers. Based on the instructions issued by the decision layer, the module adjusts the stiffness, height, and damping force to optimize the ride comfort and vehicle posture stability during vehicle operation.

[0042] Figure 2 A flowchart of a first embodiment of the vehicle control method provided by the present invention is shown. Figure 2 As shown, the method includes the following steps:

[0043] Step 110: When a subsidence occurs on the road ahead, the acquired cloud-based road subsidence data, vehicle perception data, and current vehicle status data are fused together to obtain a subsidence scenario model corresponding to the subsidence condition.

[0044] For example, subsidence conditions refer to road surface irregularities caused by localized subsidence or collapse of the road structure, manifested as geometric features such as depressions, potholes, or abrupt slope changes. Cloud-based road subsidence data refers to road subsidence information generated by a cloud server through fusion and iterative optimization of historical subsidence events reported by other vehicles. This information may include, for example, the GPS coordinates of the subsidence center, geometric dimensions, subsidence depth, confidence level, and last update time. Vehicle perception data reflects the vehicle's real-time perception of the current road environment and may specifically include at least one output from a forward-facing camera, surround-view camera, LiDAR, or millimeter-wave radar. This data may include key geometric parameters such as subsidence depth, subsidence edge slope, longitudinal length and lateral width of the subsidence, and the flatness of the subsidence bottom. Current vehicle status data characterizes the vehicle's current operating state and may include, for example, vehicle speed, sprung mass, steering wheel angle, accelerator pedal opening, brake pedal opening, vertical acceleration of the vehicle body at corresponding wheel positions, suspension travel at each wheel, and other data characterizing the vehicle's operating state.

[0045] Data fusion processing refers to the integrated processing of multi-source heterogeneous road surface data and vehicle status data through weight matching, feature fusion, and spatiotemporal alignment. The subsidence scenario model is the final output of the fusion processing. It is a structured data organization form that packages the fused and corrected subsidence geometric feature vectors (including subsidence depth, edge slope, longitudinal length, lateral width, bottom flatness, etc.) with the current vehicle status (vehicle speed, heading angle, positioning coordinates, etc.) into a unified pre-aiming scenario model. It also includes subsidence geometric parameters, positional relationships, confidence level timestamps, and the temporal relationship of the expected indentation of each wheel into the subsidence.

[0046] In one example, when a vehicle determines that there is road subsidence on the path ahead through the onboard perception system or cloud data while driving, the controller initiates a fusion processing procedure. Specifically, the controller first sends a request containing the vehicle's current position and driving direction to the cloud at preset intervals via the onboard communication device, and receives subsidence event information (cloud road subsidence data) from the cloud within the range of the path ahead of the vehicle, including the location coordinates, depth, level, and historical passage characteristics of each subsidence point; at the same time, the controller obtains the subsidence parameters identified in real time from the onboard camera (vehicle perception data), including maximum depth, edge slope, longitudinal length, lateral width, and bottom flatness; and reads state variables such as current vehicle speed, vertical acceleration of each wheel, and suspension travel (current vehicle state data).

[0047] The controller performs weighted calculations on cloud-based sinking data and vehicle-sensed data. The weighting coefficients are determined by the real-time confidence level and data age of the data source—the weight of cloud-based data decays exponentially over time, while the weight of visual data is dynamically adjusted according to the recognition confidence level (a fixed weighting ratio can also be used, such as 40% for historical cloud data and 60% for real-time visual data, to weaken the impact of lagging data and enhance the reference value of real-time road conditions). Through normalized weighted calculations, the optimal fused feature vector is output. Then, combined with the current vehicle status data, the estimated time for the vehicle to reach the sinking area and the contact sequence of each wheel are calculated. This feature vector, along with vehicle speed, heading angle, positioning coordinates, and other vehicle status and contact sequence data, is packaged into a unified sinking scene model.

[0048] This step incorporates prior information from the cloud, real-time perception results from the vehicle, and the vehicle's own operating status into the same scenario modeling process, enabling subsidence identification to not rely on a single information source. It also allows subsequent suspension control to be directly based on quantifiable subsidence depth, edge slope, area length, and wheel passage sequence, thus providing a unified input for forward-looking control under subsidence conditions.

[0049] Step 120: Based on the sinking scenario model and the current vehicle status data, determine the suspension parameters of each wheel in each sinking driving stage.

[0050] For example, the sinking driving stage characterizes the driving stage when a vehicle passes through a sinking condition. In other words, the sinking driving stage is used to divide the process of a vehicle passing through a sinking into different control segments according to the driving sequence. For example, when passing through a sinking, the vehicle may sequentially experience multiple consecutive stages such as before entering the sinking, during entering the sinking, before exiting the sinking, during exiting the sinking, and after exiting the sinking. It should be noted that the embodiments of the present invention do not limit the method and number of sinking driving stages. It is understood that the impact type, vehicle body response characteristics, and control objectives of each stage are significantly different.

[0051] Suspension parameters refer to the underlying control quantities used to control the adjustment actions of the suspension system during vehicle slumping. These parameters may include, for example, the target stiffness of the spring, the target damping force of the shock absorber, and the target vehicle height. The target stiffness refers to the equivalent elastic coefficient of the adjustable spring, which determines the suspension system's ability to resist vertical compression. It may include multiple stiffness modes; however, this embodiment of the invention does not limit the number of stiffness modes. The target damping force refers to the adjustable damping force value provided by the shock absorber, used to dissipate road impact energy and suppress vehicle vibration. The target vehicle height refers to the target value for adjusting the vertical position of the vehicle body relative to the axle or the ground, used to allow for suspension buffer travel.

[0052] In one example, after obtaining the sinking scenario model, the vehicle controller divides the passage process of each wheel into multiple sinking stages based on the temporal relationship of each wheel pressing into the sinking area in the sinking scenario model. These stages could be called the entry stage, the bottom of the sinking area, and the exit stage. When determining the suspension parameters, it is necessary to jointly analyze the sinking depth, length, lateral coverage area corresponding to the left and right wheels, entry slope, exit slope, and current vehicle state data in the sinking scenario model. Specifically, the vehicle controller can first establish a wheel-by-wheel vertical dynamics prediction model, using the change in tire contact height for each wheel as the external excitation input, and the vehicle body mass, unsprung mass, spring stiffness, damping coefficient, and actuator force as state variables to predict the wheel load changes and vehicle attitude change trends in subsequent control cycles. During the entry phase, the controller determines the adjustment amount of the corresponding suspension parameters based on the entry slope and the expected impact speed, so that the wheels have the vertical adaptability to match the terrain changes before entering the sinkhole. For the bottom of the sinkhole, the controller calculates the suspension parameters used to maintain the vehicle body posture and limit the suspension from bottoming out based on the sinkhole depth, length and current vehicle status data. During the exit phase, the controller determines the suspension parameters for the recovery phase based on the exit slope and the current vertical response of the vehicle, so that the rebound process of the vehicle body is coordinated with the tire re-climbing process.

[0053] Optionally, the switching conditions for the sinking driving stage can be determined by the expected arrival time and the real-time wheel position. When the actual vehicle speed changes and causes the expected time to deviate, the controller recalculates the boundaries of each stage to keep the parameter scheduling consistent with the actual passage process.

[0054] Understandably, this step is not simply about providing a uniform suspension setting for the whole vehicle, but rather about transforming the sinking scenario model into suspension parameters for each wheel and stage, so that different wheels can obtain corresponding control targets when they enter the sinking at different times, thereby ensuring that the timing of suspension adjustment is consistent with the sinking geometry and the real-time state of the vehicle.

[0055] Step 130: Control the vehicle's suspension based on the suspension parameters of each wheel during each sinking driving stage.

[0056] For example, a suspension is an elastic connection assembly in a vehicle chassis system that connects the frame (or monocoque body) to the axle (or wheels). The suspension specifically refers to an adjustable suspension system with active or semi-active adjustment capabilities, including actuators such as springs (e.g., air springs) with adjustable stiffness and height and CDC / MRC dampers (providing adjustable damping force).

[0057] In one example, after obtaining the suspension parameters of each wheel during each sinking driving stage, the controller generates a control trajectory corresponding to the time axis. The control trajectory includes at least the suspension parameters of each wheel during the entry, sinking bottom, and exit stages. Before the vehicle enters the sinking area, the controller can send adjustment commands to the suspension actuators of the relevant wheels one or more control cycles in advance, based on the current vehicle speed and the longitudinal distance in the sinking scenario model, so that the suspension parameters reach the target vicinity before the wheel contacts the sinking edge. During execution, each wheel actuator receives control commands via the vehicle bus. To avoid abrupt parameter changes during stage transitions, the controller performs smooth interpolation on the suspension parameters of adjacent stages. The interpolation method can be linear interpolation, spline interpolation, or restricted slope interpolation, allowing the suspension parameters to change continuously within a set transition time. The transition time can be adaptively determined based on the vehicle speed and the slope of the sinking edge. During control, the current vehicle state data is continuously fed back to the controller. The controller compares the real-time vertical acceleration of the vehicle body, suspension travel, and vehicle attitude with the predicted values. When the actual response deviates from the target range, the suspension parameters in subsequent control cycles are corrected online. For example, if the measured vertical acceleration of a wheel during the entry phase is higher than the predicted value, the damping adjustment of that wheel is increased; if the suspension travel is detected to be close to the mechanical limit in the bottom of the sinkhole, the support adjustment of the corresponding wheel is increased. If the vehicle accelerates, brakes, or turns during the passage through the sinkhole, the controller incorporates the new current vehicle state data into the current control loop, recalculates the suspension parameters and switching times for the remaining phases, and maintains the control strategy consistent with the actual operating conditions. After the vehicle leaves the sinkhole, the controller does not immediately restore the initial parameters, but gradually returns the suspension parameters of each wheel to the normal driving calibration values ​​based on the vertical vibration attenuation state after exiting the sinkhole, until the vehicle body attitude and suspension travel return to the normal range.

[0058] In this embodiment, when a road subsidence occurs ahead, a high-precision subsidence scenario model is first constructed by fusing prior road data from the cloud, real-time vehicle-mounted perception data, and vehicle status data. This model accurately recreates the true attributes of the subsided road surface. Then, based on the vehicle's real-time status, specific suspension control parameters are adaptively matched for different driving stages and different wheels. Finally, the four-wheel suspension is dynamically and precisely controlled using a time-sequential, phased control logic. This approach abandons the traditional passive, lagging control mode that relies solely on onboard sensors. By combining the advantages of multi-source fusion perception between the vehicle and the cloud, it overcomes the shortcomings of low accuracy and weak anti-interference capability of single visual perception. It can predict subsidence conditions in advance and achieve differentiated, phased active control of the four wheels, effectively reducing the impact disturbance when the vehicle travels on a subsided road surface, suppressing attitude fluctuations such as pitch and roll, and significantly improving the chassis control accuracy and robustness under severe subsidence conditions. While ensuring vehicle driving stability and chassis structural safety, it also significantly optimizes vehicle ride comfort.

[0059] Figure 3A flowchart of a second embodiment of the vehicle control method provided by the present invention is shown. Figure 3 As shown, the method includes the following steps:

[0060] Step 210: When a road subsidence occurs ahead, the road subsidence data from the cloud and the vehicle's perception data are fused to obtain subsidence feature information.

[0061] For example, the subsidence feature information is an intermediate output result after fusion processing, which may include fields such as the fused and corrected subsidence depth, subsidence edge slope, subsidence longitudinal length and transverse width, subsidence bottom flatness, and corresponding confidence level and timestamp.

[0062] In one example, when fusing the two types of data, the controller first performs spatial and temporal registration based on road coordinates, timestamps, and vehicle positioning information. Then, it performs feature stitching or weighted fusion on the depth, range, and edge morphology within the same subsidence area to form subsidence feature information. This fusion process can be implemented using feature-level fusion, confidence fusion, or multi-source consistency correction. Other fusion models can also be selected in practical applications, and this embodiment of the invention does not limit the choice.

[0063] Specifically, based on the data age of the cloud-based road subsidence data and the confidence level of the vehicle perception data, a first weighting coefficient for the cloud-based road subsidence data and a second weighting coefficient for the vehicle perception data are determined. Based on the first and second weighting coefficients, the cloud-based road subsidence data and the vehicle perception data are weighted and fused to obtain subsidence feature information.

[0064] Among them, data age refers to the time interval from the collection and update of road subsidence data in the cloud to the current moment. It is used to characterize the freshness and timeliness reliability of the prior data in the cloud. The longer the data age, the weaker the timeliness of the cloud data. Confidence score refers to the reliable probability score of the vehicle's perception device in identifying the subsidence features ahead. It is used to characterize the accuracy and reliability of the vehicle's real-time perception results. The higher the confidence score, the stronger the validity of the vehicle's perception data.

[0065] The first weighting coefficient is the fusion weight assigned to the cloud-based road subsidence data, used to quantify the contribution of cloud-based data in the feature fusion process. The second weighting coefficient is the fusion weight assigned to the vehicle perception data, used to quantify the contribution of onboard real-time perception data in the feature fusion process. Weighted fusion processing refers to an integrated calculation method that uses dual weighting coefficients to superimpose components, complement features, and correct errors in two heterogeneous road surface data streams. Subsidence feature information refers to the high-precision, high-real-time, and comprehensive road condition feature data covering subsidence depth, edge slope, and geometric dimensions output after fusing prior geometric information from the cloud and real-time road surface information from the vehicle.

[0066] For example, the controller reads the update timestamp of the cloud-based subsidence data to calculate the data age. Based on the preset age decay rule, it dynamically lowers the first weight coefficient of the old cloud data and increases the weight of the fresh cloud data. At the same time, it outputs the vehicle perception confidence based on the matching similarity of the vehicle image recognition. It increases the second weight coefficient for high-confidence perception data and decreases the second weight coefficient for low-confidence perception data. After normalizing the first and second weight coefficients, it performs weighted synthesis on the data with the same semantic dimension according to the weight coefficients. This complements the advantages of the two types of data and offsets the error defects of single data, ultimately generating accurate and reliable subsidence feature information.

[0067] By incorporating the time freshness of cloud data and the reliability of vehicle-mounted perception data into the same fusion framework, the contribution of the two types of data is dynamically allocated. This allows the subsidence feature information to retain prior information from the cloud while also reflecting the vehicle's current real-time observation results. The resulting subsidence feature information is closer to actual road conditions and can provide a more stable input basis for subsequent subsidence scenario models and suspension parameter calculations, thereby improving the consistency and accuracy of control decisions under complex subsidence conditions.

[0068] Step 220: Align the subsidence feature information with the current vehicle status data to obtain the subsidence scenario model corresponding to the subsidence condition.

[0069] In one example, after obtaining the sinking feature information, the controller further aligns it with the current vehicle state data. The current vehicle state data may include vehicle speed, sprung mass, vertical acceleration of each wheel, and suspension travel of each wheel. During alignment processing, the controller performs spatiotemporal alignment matching based on subsidence feature information and current vehicle state data: First, spatial alignment matching is performed. Based on the vehicle's real-time positioning coordinates, heading angle, and speed, the relative lateral deviation and longitudinal distance between the vehicle's current trajectory and the subsidence area ahead are calculated. The road surface geographic coordinates of the subsidence features are projected onto the vehicle's body coordinate system, completing the spatial binding between road surface features and vehicle position. Simultaneously, temporal alignment matching is performed. Based on the vehicle's real-time speed, the predicted arrival time of each wheel in the subsidence area is calculated. The road subsidence feature data and the frequently updated vehicle speed, sprung mass, and vehicle attitude state data are subjected to temporal interpolation and timestamp-based unified calibration to eliminate the sampling time difference between road surface perception data and vehicle state data. This ensures that the vehicle's driving state at each moment corresponds precisely one-to-one with the subsidence road surface features it is about to encounter. Finally, a subsidence scenario model that is temporally synchronized, spatially matched, and highly consistent with real driving conditions is constructed. This subsidence scenario model establishes a mapping relationship between subsidence feature information and current vehicle state data under the same control time, the same wheel dimension, and the same traffic stage.

[0070] By first fusing cloud-based road subsidence data with vehicle perception data, and then aligning the fusion result with current vehicle status data, road condition information and vehicle response information are processed in a layered manner, forming a subsidence scenario model with spatiotemporal consistency. This improves the completeness of subsidence area identification and ensures that subsequent suspension control parameters maintain a correspondence with current vehicle status data, making phased suspension control easier to implement.

[0071] Step 230: Determine the subsidence condition level based on the subsidence scenario model.

[0072] For example, the settlement condition level is a condition level divided according to the degree of danger and impact intensity of the settled road surface, which is used to differentiate the matching of suspension control parameters. For example, it can be divided into mild, moderate and severe settlement levels. It should be noted that the embodiments of the present invention do not limit the division level of the settlement condition level.

[0073] In one example, the controller can extract the subsidence depth, edge slope, and subsidence geometric parameters from the subsidence scenario model, and combine them with preset working condition classification thresholds for comprehensive evaluation to quantify the subsidence working condition level corresponding to the current road surface.

[0074] Specifically, the subsidence condition level is determined based on the subsidence depth, subsidence edge slope, and subsidence geometric parameters.

[0075] The subsidence depth, defined as the vertical height difference between the lowest point of the subsided road surface and the normal, smooth road surface, is a core indicator of road impact intensity; the greater the depth, the greater the vertical impact from the wheel. The subsidence edge slope, the inclination angle of the subsidence inlet and outlet slopes, characterizes the instantaneous impact intensity when a wheel enters or exits the subsidence; the greater the slope, the higher the instantaneous pitch and peak impact. Subsidence geometric parameters, including longitudinal length, transverse width, and profile morphology, are used to determine the coverage area of ​​the subsidence and the duration of wheel contact, distinguishing between localized point subsidence and large-area continuous subsidence.

[0076] For example, the controller can preset a mapping table of the value range of each dimension parameter under each working condition level, and then determine the subsidence working condition level based on the subsidence depth, subsidence edge slope and subsidence geometric parameters in the current subsidence scenario model, combined with the mapping table.

[0077] By combining subsidence depth, subsidence edge slope, and subsidence geometry parameters in the classification of working conditions, the judgment of the severity of subsidence is consistent with the actual road surface morphology, so that the subsidence working condition level can more accurately reflect the impact risk and difficulty of vehicles passing through the area.

[0078] Step 240: Determine the target stiffness and reference damping of each wheel in each stage of subsidence based on the subsidence condition level, vehicle speed and sprung mass.

[0079] For example, the target stiffness is the optimal stiffness parameter adapted to the current subsidence condition and used to adjust the elastic support capacity of the suspension, which can be adapted to various adjustable stiffness suspension structures. The reference damping is the preset value of basic damping obtained based on the matching of the subsidence condition and the overall vehicle state under each subsidence driving stage, serving as the baseline for subsequent damping corrections. Vehicle speed is the real-time vehicle speed used to determine impact intensity and adjust response speed. Sprung mass is the vehicle body load-bearing mass used to match the adaptability of the suspension support stiffness and the basic damping.

[0080] In one example, parameter matching is performed independently for the four wheels of the vehicle. Based on the determined subsidence condition level, real-time vehicle speed and vehicle sprung mass, the suspension target stiffness and reference damping baseline for each wheel at each stage of subsidence driving are calculated by referring to tables according to the working condition requirements of each subsidence driving stage.

[0081] Specifically, based on the subsidence condition level, vehicle speed, and sprung mass, the target stiffness of each wheel in each subsidence driving stage is determined; based on the subsidence condition level, vehicle speed, and target stiffness, the reference damping of each wheel in each subsidence driving stage is determined.

[0082] For example, the controller can be based on the mapping relationship between the subsidence condition level, current vehicle speed, sprung mass, and stiffness. This mapping relationship can be pre-calibrated as lookup table data, piecewise functions, or fitted functions. The controller retrieves the target stiffness of the corresponding wheel at each subsidence driving stage based on the input parameters, so that the front and rear wheels, and the left and right wheels obtain stiffness parameters that match the load and impact level when entering, crossing, and leaving the subsidence area.

[0083] After determining the target stiffness, the controller also pre-stores the correspondence between the sinking condition level, vehicle speed, and damping baseline. Furthermore, the controller can use this correspondence to determine the initial reference damping and, based on the current target stiffness, determine the final reference damping. It's understandable that stiffness and damping are not two independently adjustable variables; they have a strong coupling relationship in suspension dynamics. When switching to high stiffness mode, the damping baseline is simultaneously increased by one level to match the higher spring restoring force, preventing slower body oscillation decay due to increased stiffness and uncoordinated damping. When returning to low stiffness mode, the damping baseline is simultaneously reduced to prevent comfort degradation caused by a mismatch between damping force and low stiffness. During the transient process of stiffness switching, the damping force command undergoes smooth transition processing, with the transition time window determined by the actual vehicle step calibration, avoiding longitudinal step response of the body caused by a misalignment of stiffness-damping rhythm.

[0084] Alternatively, the controller may have a pre-stored correspondence between the settlement condition level, vehicle speed, stiffness, and damping baseline, so that after determining the target stiffness, the matching reference damping can be directly determined.

[0085] After receiving the subsidence condition level, vehicle speed, and sprung mass, the controller first generates target stiffness for each wheel at different stages. Then, it generates reference damping based on the target stiffness, so that the damping parameters are no longer set independently but are determined in conjunction with the stiffness parameters. This provides a unified set of basic parameters for subsequent damping correction and adjustment. As a result, each wheel can form a consistent initial suspension response reference in different subsidence driving stages and match the changes in the subsidence scenario.

[0086] By adopting this method, the initial setting of suspension parameters can simultaneously consider the sinking level, vehicle speed, and load changes, so that each wheel can obtain the corresponding stiffness and damping benchmarks at different sinking driving stages, thereby improving the consistency and adaptability of the staged control parameters and providing stable input for subsequent damping correction.

[0087] Step 250: For each wheel, based on the wheel's vertical acceleration, the wheel's suspension travel, and the PID control parameters corresponding to each sinking driving stage, the reference damping for each sinking driving stage is corrected to obtain the wheel's intermediate damping for each sinking driving stage.

[0088] For example, the vehicle body vertical acceleration is a real-time vertical vibration and impact parameter corresponding to the wheel position on the vehicle body, used to characterize the road impact intensity. The suspension dynamic travel is the real-time extension and contraction displacement of the suspension relative to the reference center position, reflecting the current degree of compression or extension of the suspension.

[0089] The proportional-integral-differential (PID) control parameters are independently configured proportional, integral, and derivative gain parameters for each sinking and driving stage, used to adapt to the dynamic adjustment requirements of different operating conditions. The intermediate damping is the reference damping after real-time deviation correction.

[0090] In one example, for each wheel, the comprehensive deviation of vehicle body vibration is calculated based on individually collected real-time vertical acceleration and suspension travel. The damping correction is calculated by combining the PID control parameters matched for the current driving stage. The correction is then superimposed on the reference damping for the corresponding stage to obtain the real-time updated intermediate damping for each wheel.

[0091] Specifically, for each sinking driving stage, the current deviation value is determined based on the vehicle's vertical acceleration and the wheel's suspension travel; the current damping correction increment of the wheel is determined based on the current deviation value and the PID control parameters corresponding to the sinking driving stage; the current damping correction of the wheel is obtained based on the damping correction amount at the previous moment and the current damping correction increment; and the intermediate damping of the wheel in the sinking driving stage is obtained based on the current damping correction of the wheel and the reference damping of the wheel in the sinking driving stage.

[0092] The current deviation value is the comprehensive operating condition deviation, which integrates vertical vibration deviation and suspension displacement deviation, and is used to characterize the degree of deviation between the current actual operating state of the suspension and the ideal stable state. The damping correction increment is the real-time damping fine-tuning change calculated by the PID algorithm within a single control cycle. The damping correction amount at the previous moment is the accumulated damping correction value obtained in the previous sampling cycle, used to achieve continuous iterative adjustment over time. The current damping correction amount is the real-time total correction amount after adding the current increment to the historical correction. The intermediate damping is the real-time damping parameter of a single wheel after adding the dynamic correction amount to the reference damping.

[0093] For example, for each wheel and each stage of the sinking process, the controller, in each control cycle (e.g., 5ms per control cycle), first collects the vehicle's vertical acceleration and suspension travel at the current moment, and then calculates the current deviation value through weighted synthesis. That is, the current deviation value. It can be represented as ;in, This indicates the vehicle's vertical acceleration, used to reflect ride comfort and impact intensity; Indicates the suspension travel; This represents the target value of the dynamic travel, taken as the midpoint of the total allowable travel of the suspension, to ensure equal upper and lower buffer spaces; , These are weighting coefficients. It should be noted that the weighting coefficients can be dynamically adjusted based on the subsidence condition level; for example, they can be increased for severe subsidence. Prioritize shock suppression, increase efforts for minor subsidence. Prioritize controlling travel margin.

[0094] The system retrieves the proportional, integral, and derivative PID parameters specific to the current sinking driving phase. It then performs proportional adjustment, integral cumulative correction, and derivative lead suppression calculations on the current deviation value and historical deviations, outputting the damping correction increment for the current control cycle. That is, the current damping correction increment. It can be represented as ;in, , Indicates historical deviation value; This represents the proportional gain, which is a discrete-domain parameter. The integral gain is a discrete-domain parameter that already includes the sampling period T_s. The differential gain is a discrete-domain parameter that already includes the sampling period T_s.

[0095] The damping correction value obtained from the previous iteration update The current damping correction increment is then time-series superimposed with the current damping correction increment to obtain the current damping correction amount. That is, the current damping correction amount. It can be represented as .

[0096] Finally, the current total damping correction amount is... The damping is superimposed on the preset wheel reference damping for this stage and the intermediate damping of the wheel in the current sinking driving stage is output in real time, so as to realize the continuous dynamic fine adjustment of the damping parameters as the vehicle body vibrates and the suspension travel fluctuates.

[0097] With this implementation method, the damping correction can be continuously updated according to the changes in the vehicle's vertical acceleration and suspension travel, and combined with the PID control parameters for different sinking driving stages, so that the reference damping can maintain a smooth transition when switching stages, thereby improving the consistency of suspension response and control accuracy under sinking road conditions.

[0098] Step 260: Based on the vehicle's sinking and passing scenario, the intermediate damping of each wheel in each sinking and passing stage is corrected to obtain the target damping of each wheel in each sinking and passing stage.

[0099] For example, the subsidence driving scenario refers to the operating condition of a vehicle traveling on a subsided road surface, which may include three types of scenarios: single-sided subsidence, double-sided subsidence, and continuous subsidence. The correction process is a vehicle-level optimization process based on the vehicle's attitude balance requirements, involving four-wheel coordinated correction, amplitude fine-tuning, and timing compensation of the single-wheel intermediate damping. The target damping is the precise damping control parameter ultimately used by the shock absorber after single-wheel dynamic correction and vehicle scenario correction. The suspension parameters for each subsidence driving stage include the target stiffness and target damping for each subsidence driving stage.

[0100] In one example, if a single-sided sinking scenario is identified, the intermediate damping on the impacted side remains unchanged (the intermediate damping is used as the target damping), while the reference damping on the non-impacted side is used (the reference damping is used as the target damping), and the damping difference between the two sides is constrained. If a double-sided sinking scenario is identified, the damping response of the front axle wheel is used as the reference, and the intermediate damping of the rear axle wheel is corrected by combining the wheelbase timing offset and amplitude fine-tuning. If a continuous sinking scenario is identified, the damping parameters of the previous stage are inherited to complete the transition correction, and finally the target damping of each wheel at each stage is obtained.

[0101] Specifically, when the vehicle sinking scenario is unilateral sinking, the intermediate damping of the impacted wheel in each sinking driving stage is used as the target damping of the impacted wheel in each sinking driving stage; and the reference damping of the non-impacted wheel in each sinking driving stage is used as the target damping of the non-impacted wheel in each sinking driving stage. When the vehicle sinking scenario is bilateral sinking, the intermediate damping of the front axle wheel in each sinking driving stage is used as the target damping of the front axle wheel in each sinking driving stage; and based on the actual damping of the front axle wheel and the timing offset corresponding to the vehicle wheelbase, the intermediate damping of the rear axle wheel in each sinking driving stage is corrected to obtain the target damping of the rear axle wheel in each sinking driving stage.

[0102] Unilateral subsidence refers to an asymmetrical subsidence condition where only one side of the vehicle's path is affected by road subsidence, while the other side remains on a smooth surface. This condition is highly prone to causing vehicle tilting and lateral imbalance. The wheel on the impact side is the one that enters the subsidence area and experiences the vertical impact from the road surface; the wheel on the non-impact side is the one that remains on a smooth surface and is not subject to subsidence impact.

[0103] Bilateral sinking refers to a symmetrical sinking condition where the travel paths of both wheels on the left and right sides of the vehicle cover the sinking area. The entire vehicle continuously experiences vertical impacts from the road surface, resulting in a timing misalignment issue between the front and rear axles. The actual damping of the front axle wheels is the target damping of the front axle after vehicle-wide scenario correction and sent to the shock absorbers for execution. This value serves as a feedforward signal for the actual front axle response in the PID control loop of the rear axle wheels, fully reflecting the true damping output effect of the front wheels in response to sinking impacts. The timing offset is the time difference between the front and rear axle wheels reaching the same sinking position, calculated based on the vehicle wheelbase and real-time vehicle speed, used to characterize the impact lag characteristics of the front and rear wheels. The rear axle correction processing is a collaborative control method that uses the actual damping state of the front axle and the timing lag characteristics to proactively correct the rear axle damping, achieving front-to-rear impact matching and attitude synchronization.

[0104] In one example, the controller identifies the current sinking traffic scenario through a sinking scenario model. If it is determined to be a single-sided sinking scenario, for the wheel on the impact side, the intermediate damping obtained by real-time PID dynamic correction is directly used as the final target damping to ensure that the impact side can offset the sudden vibration and displacement deviation of the road surface in real time and efficiently dissipate the impact energy. For the wheel on the non-impact side, the dynamic correction damping is abandoned, and the preset benchmark damping of each stage is returned to avoid excessive adjustment of the non-impact side and the introduction of unnecessary vehicle body disturbance. At the same time, the vehicle body roll trend is suppressed by the differential matching of left and right damping. If the scenario is determined to be a double-sided sinking, the front axle wheels are used as the real-time sensing reference. The front axle intermediate damping is directly locked as the final target damping of the front axle, accurately adapting to the real-time sinking impact state of the front wheels. At the same time, the timing offset between the front and rear axles is calculated in real time based on the vehicle's fixed wheelbase and current speed. The timing and intensity of the road impact that the rear wheels will be about to experience are predicted. Based on the actual damping of the front axle wheels, the timing compensation and amplitude fine-tuning of the current intermediate damping of the rear axle are performed to generate the target damping of the rear axle wheels. This offsets the lag impact error of the rear wheels sinking, ensuring that the rear axle damping state accurately matches the front wheel damping rhythm and the overall vehicle attitude change.

[0105] Through the above correction method, the damping distribution difference between the impacted side and the non-impacted side can be maintained when one side sinks, and the rear axle control and the front axle control can be kept consistent in time when both sides sink. This allows each wheel to obtain the target damping that matches its impact relationship in different sinking driving stages, thereby improving the continuity of suspension response and attitude stability during passage.

[0106] Specifically, when the vehicle sinking and passing scenario is continuous sinking, the PID integral term is cleared to zero and reinitialized after the previous sinking exit stage ends, and the target damping of the previous sinking exit stage is used as the reference damping for the next sinking entry stage.

[0107] Continuous subsidence refers to a driving path with multiple closely spaced subsidence areas, where the vehicle repeatedly enters, passes through, and exits these subsidence areas. Continuous impacts can easily cause the PID integral term to accumulate and saturate, leading to damping overshoot and vehicle body posture oscillations. The PID integral term is the integral component of the PID control parameters, used to continuously accumulate vehicle body vibration deviations and eliminate steady-state vibration errors. Under long-term continuous excitation, the integral value will continuously increase, resulting in integral saturation. Clearing and reinitializing the integral term means that after completely exiting a subsidence, the current accumulated integral value is cleared, the initial integral reference value is restored, and the interference of integral saturation on subsequent adjustments is eliminated. The target damping of the previous subsidence exit stage is the damping parameter finally executed by the shock absorber when the vehicle leaves the previous subsidence, reflecting the optimal damping matching state at the end of the previous subsidence. The reference damping of the next subsidence entry stage is the basic damping baseline upon which the PID correction calculation is based when the vehicle enters the next subsidence, serving as the initial reference for a new round of dynamic damping fine-tuning.

[0108] In one example, the controller uses a sinkhole scenario model to identify multiple sinkholes arranged in succession ahead. When it detects that the vehicle has completed all the time-sequence driving and exiting the previous sinkhole, it immediately clears all the integral stored values ​​of the PID control loops for each wheel corresponding to the driving stage, restores the initial value of the integral calculation to the preset zero reference to complete the initialization, and avoids the problem of integral accumulation saturation caused by the continuous impact of the previous sinkhole. At the same time, it retrieves the target damping value effective for each wheel in the exiting stage of the previous sinkhole and directly assigns it as the reference damping for the entering stage of the next sinkhole. This allows the damping baseline of the next round of sinkhole adjustment to fit the optimal vibration reduction range matched by the impact of the previous road surface, reducing the sudden change in vehicle body impact caused by the large adjustment of damping in the early stage of the new round of sinkhole.

[0109] In this way, the control connection between successive subsidence areas is directly carried over from the final damping value of the previous stage to the next stage. At the same time, the residual influence of the integral term of the previous stage on subsequent calculations is eliminated, so that the damping adjustment of the subsequent subsidence entry stage can be restarted with a stable reference value. This ensures that the damping switching between multiple consecutive subsidence areas maintains continuity and consistency.

[0110] Optionally, when the subsidence condition level and / or the current vehicle status data meet the preset conditions, the target vehicle height of each wheel in each subsidence driving stage is determined according to the subsidence depth in the subsidence scenario model.

[0111] The preset conditions include any of the following: the subsidence condition level is higher than the preset level; the static compression of the suspension is greater than the preset static compression threshold; the predicted dynamic compression of the suspension travel is greater than the preset dynamic compression threshold; the current ground clearance is greater than the safety limit threshold; and the vehicle speed is within the preset range.

[0112] Among them, the preset conditions are the thresholds for triggering the active vehicle height adjustment function; meeting any one of them will activate the vehicle height adaptation logic. The preset level is a pre-calibrated critical value for grading sinking conditions, used to distinguish between mild sinking and moderate to severe sinking that requires lifting the vehicle body. The static compression of the suspension is the compression displacement of the suspension relative to the center position when the vehicle is stationary and under load, reflecting the amount of basic sinking caused by the sprung mass; the preset static compression threshold is the safe upper limit of the static compression of the suspension, exceeding this value indicates that the basic ground clearance of the vehicle body is too low.

[0113] The predicted suspension dynamic travel compression is the maximum compression displacement of the suspension when the wheel passes through the pothole, determined by combining the sinkhole depth and vehicle speed predictions. The preset dynamic compression threshold is the safe critical value of the suspension buffer travel; exceeding this value poses a risk of suspension bottoming out and chassis impact. The current ground clearance is the vertical distance between the lowest chassis position and the road surface. The safety limit threshold is the minimum safe ground clearance to prevent the chassis from scraping against the road surface indentation. The preset range is a reasonable speed range for adapting to vehicle height adjustment; height adjustment is turned off when the vehicle speed is too high or too low to avoid control malfunction. The sinkhole scenario model fully stores the sinkhole depth parameters representing the current pothole road surface. The greater the sinkhole depth, the higher the vehicle height needs to be raised. The target vehicle height is the target suspension lift height obtained by matching the sinkhole depth at each sinkhole driving stage; for example, the chassis ground clearance can be adaptively adjusted by inflating and deflating the air springs.

[0114] For example, the controller reads the sinking condition level, static compression of the suspension, predicted suspension travel, chassis ground clearance, and real-time vehicle speed in real time and compares them with each preset threshold. If any indicator reaches the trigger standard, the sinking depth value in the sinking scenario model is retrieved. According to the pre-calibrated sinking depth-vehicle height correspondence, the target vehicle height corresponding to each wheel is independently matched for each driving stage: the entry stage, the bottom of the sinking stage, the early stage of climbing, and the middle stage of climbing. The greater the sinking depth, the higher the target vehicle height is set, so as to reserve sufficient suspension buffer travel and avoid the risk of chassis collision.

[0115] The target vehicle height can also be calculated based on the subsidence depth. For example, the target vehicle height. It can be represented as ,in, Indicates the depth of subsidence. This indicates the maximum allowable adjustment range in a single operation. This value is constrained by the air spring travel range and inflation rate, and can be calibrated according to the physical characteristics of the vehicle's suspension system. This represents the calibration ratio coefficient, with a value less than 1, ensuring that the height adjustment is always less than the subsidence depth, avoiding excessive lifting that could shift the vehicle's center of gravity too high and affect lateral stability. Optionally, the value of this calibration ratio coefficient is related to the subsidence condition level; for moderate subsidence conditions... Take the smaller value, the adjustment amount is conservative, and the main goal is to make a small allowance; for severe subsidence conditions Take a larger value, moderately increase the adjustment amount, and maximize the available compression stroke of the suspension, but always keep it within the acceptable range. As a hard upper limit, the embodiments of the present invention do not limit the specific value range.

[0116] In this embodiment, the target vehicle height is triggered and determined by both the sinking depth and the current vehicle status data. This allows height control to match the severity of the sinking scenario and the vehicle's real-time attitude. Based on this, the controller outputs corresponding height setpoints to each wheel in stages, ensuring that the suspension maintains a continuous height compensation relationship when entering, passing through, and exiting the sinkhole, thereby improving vehicle attitude stability and chassis passability. Because the generation of the target vehicle height is subject to multiple constraints, the system can only perform height adjustments when there is sufficient adjustment space and when adjustment is truly necessary, thus improving the targeting and consistency of height control.

[0117] Step 270: Control the vehicle's suspension based on the suspension parameters of each wheel during each sinking driving stage.

[0118] It should be noted that this step is similar to step 130 mentioned above, and will not be repeated here.

[0119] Optionally, the vehicle's suspension can be controlled in stages based on the target stiffness, target damping, and target vehicle height of each wheel during each sinking driving phase.

[0120] In this context, phased control refers to the controller sending control signals to the corresponding actuators according to the specific sinking driving stage of each wheel of the vehicle and the corresponding suspension parameters for that stage, so that the suspension system executes differentiated adjustment strategies at each stage.

[0121] For example, as mentioned above, the controller can pre-calculate the suspension parameters such as the target stiffness, target damping, and target vehicle height of each wheel in each sinking driving stage. Then, when the wheel enters the corresponding sinking driving stage, the suspension system can be controlled according to the suspension parameters of that stage.

[0122] In this embodiment, firstly, accurate subsidence characteristic information is obtained by fusing cloud-based prior subsidence data with vehicle-mounted real-time perception data. Then, a subsidence scenario model is constructed by combining real-time vehicle status data to achieve spatiotemporal alignment. The subsidence condition level is quantitatively determined based on the scenario model. Target stiffness and reference damping are matched for each wheel at each subsidence driving stage by combining vehicle speed and sprung mass. Intermediate damping is obtained by dynamically correcting single-wheel damping through real-time vehicle vibration parameters and segmented PID control parameters. Finally, the final suspension parameters are obtained by coordinating the damping of all four wheels of the vehicle in different subsidence driving scenarios. The suspension is then precisely controlled in stages according to the subsidence driving sequence. This approach overcomes the shortcomings of single-sensor methods, such as low accuracy, poor anti-interference, and lack of predictive ability, by integrating multi-source data from the vehicle and cloud. Relying on multi-level control logic, including graded working condition judgment, segmented PID dynamic correction, and four-wheel scenario collaborative correction, it achieves refined management and control throughout the entire process, from road scene modeling, working condition classification, parameter adaptive matching to phased active control. This completely solves the problems of traditional suspension passive response lag, single control parameters, and inability to adapt to complex impact conditions of sinking. It effectively suppresses pitch and roll attitude fluctuations during vehicle sinking, reduces road impact damage to the chassis, and significantly improves vehicle driving stability, chassis safety, and ride comfort under harsh road conditions.

[0123] Figure 4 A schematic diagram of an application scenario provided by the present invention is shown, such as... Figure 4 As shown, the vehicle control method provided in this embodiment of the invention may include the following steps.

[0124] 1. Cloud-based long-distance perception: While driving, other vehicles can use their onboard suspension height sensors, acceleration sensors, and onboard cameras to collect real-time data on road bumps, undulations, and subsidence. This data includes geometric parameters such as subsidence length, width, and depth, and marks road subsidence events. The onboard computing platform anonymizes the collected raw perception data and extracts subsidence features. It then uploads road subsidence data, including GPS coordinates of the subsidence center, estimated geometric dimensions of the subsidence, average subsidence depth (or subsidence condition level), perception confidence, and the last update time, to a cloud server. The cloud server aggregates and integrates the road data collected by crowdsourced vehicles and uses AI models to perform temporal memory and iterative optimization. It continuously generates and updates a high-precision road subsidence layer covering the entire road surface, providing high-precision data support for vehicles to subsequently obtain cloud-based road subsidence prior data, predict subsidence conditions, and actively adjust the suspension.

[0125] The vehicle periodically (e.g., 1Hz) requests subsidence information about the path ahead (within 200 meters) from the cloud via the T-BOX. Based on the vehicle's real-time positioning and driving direction, the cloud sends information such as the location, severity, and historical passage characteristics of subsidence points within several kilometers ahead to the vehicle, forming an "electronic pre-aiming path." The onboard controller obtains basic road condition information such as the location of road subsidence points, historical subsidence depth, confidence level, and road slope within a range of 50-200 meters ahead in advance, completing an initial road condition prediction and allowing sufficient response time for suspension adjustments.

[0126] After receiving the sinking data from the cloud, the vehicle controller combines its own positioning and current speed to calculate the estimated distance and time to reach the first sinking area. If the time is greater than the suspension system's maximum pre-adjustment response time (e.g., 1 second), pre-adjustment preparation is initiated; at the same time, the timestamp of the sinking event is verified. If it exceeds the validity period (e.g., 72 hours) or the confidence level is lower than the threshold, it is downgraded to a real-time correction mode that relies solely on vision.

[0127] 2. Accurate close-range identification by camera: The vehicle-mounted camera monitors the road surface information in real time. When a road subsidence event enters the camera's sensing range, it sends the following key sensing parameters back to the controller: maximum subsidence depth, subsidence edge slope, longitudinal length and lateral width of the subsidence, and flatness of the bottom of the subsidence.

[0128] (1) Maximum subsidence depth: Through three-dimensional reconstruction and scale alignment, the vertical drop of the lowest point of the subsidence area relative to the normal road surface is accurately calculated, which serves as the core basis for assessing the severity of subsidence and the suspension travel adjustment.

[0129] (2) Slope of the sinking edge: Measure the slope of the road surface at the sinking edge to reflect the steepness of the sinking edge. It is used to judge the intensity of the impact when the wheel enters the sinking and the suspension damping pre-adjustment strategy.

[0130] (3) Longitudinal length and lateral width of sinkhole: The geometric dimensions of the sinkhole area are measured along the vehicle's driving direction and perpendicular to the driving direction, respectively. The longitudinal length determines the duration of the wheel passing through the sinkhole, and the lateral width is used to determine whether the sinkhole affects one or both wheels, thereby guiding the formulation of the four-wheel independent control strategy.

[0131] (4) Sinking bottom flatness: assess the degree of undulation inside the sinking bottom surface, to predict the intensity of continuous bumps when the wheel is driving on the sinking bottom, and to provide a reference for real-time suspension damping adjustment.

[0132] The above parameters are transmitted back to the multi-source data fusion module in real time in the form of structured data, and are spatiotemporally aligned and fused with the subsidence data in the cloud to jointly construct a complete subsidence scenario prediction model.

[0133] 3. Multi-source perception data fusion: Through multi-source data preprocessing and spatiotemporal alignment mechanism, the cloud-based data, the vehicle perception data obtained from vehicle vision recognition, and the vehicle status data such as vehicle speed, steering wheel angle, accelerator / brake pedal signals, vehicle acceleration, suspension current travel, and sprung mass output by the vehicle CAN bus are uniformly calibrated to the same timestamp and vehicle coordinate system. Relying on the timestamp synchronization mechanism of the global high-precision clock, the problems of low frequency and high latency of vehicle cloud map, low frequency and low latency of vehicle camera, and inconsistent sampling rate and time sequence between high frequency real-time chassis sensors are solved.

[0134] Specifically, hardware-level timestamps are added to various data acquisition terminals and uniformly transmitted to the fusion center. The fusion center configures a circular buffer to store the complete multi-source data streams of the recent period. At the same time, the Extended Kalman Filter (EKF) framework is used to realize asynchronous dynamic state estimation of multi-source data. Driven by IMU high-frequency data, the system state is continuously updated with time prediction. When cloud data or visual perception data arrives, the measurement correction update is completed by matching the historical state in the buffer according to the corresponding timestamp, and the corrected state is forward propagated to the current moment, effectively compensating for the phase error caused by data transmission and processing delays. Based on this, an adaptive weighted fusion algorithm is used to accurately fuse subsidence geometric features such as subsidence depth and width. Using high-precision real-time parameters from visual recognition as a benchmark, the fusion accuracy is constrained by cloud-based geometric parameters obtained from the subsidence influence radius and historical vehicle average suspension response data. The weight coefficients are dynamically determined by the data age and real-time confidence level of the data source. The weight of cloud data decreases exponentially with the data age, while the weight of visual perception data is adaptively and dynamically adjusted with the recognition confidence level. Alternatively, a fixed ratio of 40% for historical cloud data and 60% for real-time visual data can be used to weaken the interference of lagging data and enhance the reference value of real-time road conditions. Finally, the optimal subsidence fusion feature vector is output through normalized weighted calculation. This algorithm can stably output high-precision and high-reliability road surface prediction information even under conditions of single sensor failure or high data latency, providing accurate data support for subsequent subsidence scenario modeling, condition level determination, and refined active control of the suspension.

[0135] 4. Road Subsidence Pre-aiming Scene Construction: The fused subsidence feature vector and the vehicle's current state (vehicle speed, heading angle, positioning coordinates) are packaged into a unified pre-aiming scene model, which includes subsidence geometric parameters, positional relationships, confidence level and timestamp, and serves as input for the subsequent fusion decision stage.

[0136] By integrating multi-source data fusion and road subsidence prediction scenario construction, the system combines long-distance pre-perception from vehicle-cloud maps with close-range accurate identification from vehicle cameras to achieve full-element perception and unified scenario model construction of subsidence scenarios, providing complete prediction information for subsequent fusion decision-making.

[0137] 5. Integrated Decision and Parameter Calculation Stage: This stage uses the completed road subsidence prediction scenario model as the core input, combines the vehicle's real-time driving conditions and chassis status parameters, completes the intelligent judgment of the working condition level and the autonomous matching of the control strategy, and accurately calculates the three core adjustment parameters of suspension equivalent stiffness, adjustable damping and vehicle target height based on the hierarchical algorithm system, outputs independent control commands for four wheels, and completes the pre-control logic decision.

[0138] (1) Classification of road subsidence conditions: Based on the subsidence depth, edge slope, and geometric dimensions after fusion correction in the subsidence scenario model, a multi-dimensional condition evaluation system is established. The fuzzy hierarchical analysis method is used to classify the subsidence traffic conditions and define three levels of control standards:

[0139] Mild subsidence condition: Subsidence depth ≤30mm, gentle edge slope, only produces minor road surface undulation impact, with minimal impact on vehicle body posture;

[0140] Moderate subsidence condition: Subsidence depth 30-60mm, with obvious inclination at the edges, vehicle passage is prone to vertical vibration and slight attitude shift;

[0141] Severe subsidence conditions: Subsidence depth > 60mm, steep edges, large bottom undulations, high-speed traffic is prone to wheel impact, side tilting and ground lifting risks.

[0142] (2) Intelligent calculation of core suspension parameters:

[0143] Target stiffness: The stiffness of the multi-chamber air spring is simplified into two levels of adjustment: high stiffness and low stiffness, in order to reduce the complexity of the actuator response and improve the efficiency of mass production calibration. The timing of the stiffness command switching is determined by back-calculating the air spring response delay margin from the "front wheel sinking moment" output by the pre-aiming timing calculation module, ensuring that the stiffness preset is completed before the wheel contacts the sinking edge.

[0144] Specifically, the controller can determine the target stiffness based on the correlation between the sinking condition level, vehicle speed, sprung mass, and stiffness. For example, mild sinking condition with vehicle speed below the threshold → low stiffness; moderate and above sinking condition → high stiffness; mild sinking condition with vehicle speed above the threshold → high stiffness; sprung mass exceeding the full load threshold → high stiffness. The low stiffness mode prioritizes smooth vibration filtering, maintaining the suspension frequency within the comfort range; the high stiffness mode prioritizes vertical support, increasing the suspension's resistance to impact energy and preventing wheel bottoming and excessive vehicle pitch. During the transient process of stiffness switching, the damping force command undergoes smooth transition processing, with the transition time window determined by the actual vehicle step calibration to avoid longitudinal step response of the vehicle body caused by stiffness-damping rhythm misalignment. The stiffness classification command and damping baseline command are uniformly scheduled in the controller, forming a coordinated matching relationship of "stiffness and flexibility."

[0145] b. Target Damping: The damping force calculation adopts a composite control architecture of "feedforward baseline setting, feedback deviation adjustment, and segmented parameter switching". The controller has a preset reference damping MAP table. First, the reference damping MAP table determines the reference damping force based on the sinking condition level and vehicle speed. Then, the incremental PID outputs the correction increment based on the real-time status of the vehicle body. The two are superimposed to form the intermediate damping. The three gain parameters of the PID are switched segmentally according to the sinking driving stage (driving in / bottoming out) to adapt to the different control objectives of each stage.

[0146] 1) Feedforward Reference Damping Force: The feedforward component handles the main adjustment, aiming to minimize the adjustment burden on the PID feedback loop and reduce the risk of overshoot. The reference damping MAP table uses the indentation condition level (mild / moderate / severe) as rows and real-time vehicle speed as columns. It is a two-dimensional lookup table generated through actual vehicle calibration, outputting the reference damping force. The calibration principles are as follows: Mild condition → low damping baseline, prioritizing vibration filtering; Moderate condition → medium damping baseline, balancing support and buffering; Severe condition → high damping baseline, maximizing impact energy dissipation. Vehicle speed correction rule: Under the same condition level, for each speed increase, the reference damping force is appropriately increased according to the calibration slope. The increase range and upper limit are determined by subjective evaluation calibration on the actual vehicle. Optionally, after determining the target stiffness, the damping baseline is switched to a reference damping value that matches the target stiffness. That is, when switching to the high stiffness mode, the damping force feedforward baseline is simultaneously increased by one level to match the higher spring restoring force, so as to avoid the body oscillation attenuation slowing down due to increased stiffness and failure of damping to follow. When returning to the low stiffness mode, the damping force feedforward baseline is simultaneously retracted to prevent the deterioration of comfort caused by the mismatch between damping force and low stiffness.

[0147] 2) Incremental PID Feedback Correction: The feedback section, based on the feedforward reference damping, fine-tunes the damping force according to the real-time vehicle condition to compensate for uncertainties such as load differences, suspension aging, and random road disturbances that the feedforward map cannot cover. The feedback correction amount is constrained by upper and lower limits. The upper limit prevents oversaturation of the damping force, which could lead to deterioration of tire contact with the ground, while the lower limit ensures that the minimum damping force meets the basic vibration reduction requirements. In each control cycle (e.g., 5ms per control cycle), the controller first collects the current vehicle vertical acceleration a_z(k) and suspension travel S_sus(k), and then calculates the current comprehensive deviation e(k) through weighted synthesis. Based on this deviation and historical deviations e(k-1) and e(k-2), the damping force correction increment for the current cycle is calculated. u(k). This is accumulated and added to the absolute PID output u(k-1) of the previous cycle to obtain u(k), which is then superimposed with the feedforward reference damping force to form the intermediate damping force. At the end of the cycle, e(k-1) is assigned to e(k-2), e(k) is assigned to e(k-1), and u(k) is assigned to u(k-1), completing the rolling update of the historical state and preparing for the next cycle.

[0148] 3) Integral Saturation Resistance and Output Limiting: The absolute value of u(k) is constrained by upper and lower limits—upper limit ≤ maximum adjustable damping force of the damper; lower limit ≥ minimum damping force of the damper, ensuring that the foundation's damping capacity is not lost. Simultaneously, the integral increment is increased when any of the following conditions are triggered. Forced zeroing of I—u(k) reaches output limit; switching occurs during sinking driving phase (driving in → bottom section, bottom section → driving out); moment of stiffness mode (high / low stiffness) switching. Anti-saturation logic ensures that the damping force command will not jump due to the accumulation of historical integrals when switching operating conditions.

[0149] 4) Segmented parameter scheduling strategy: The control objectives of damping force differ in different subsidence driving stages such as the entry stage, the subsidence bottom stage, and the exit stage. Based on the different stages of the subsidence passage process, the PID control parameters are scheduled online. , , This achieves differentiated control response characteristics at each stage. The strategy for scheduling PID control parameters can be represented as shown in Table 1.

[0150] Table 1 Scheduling strategies for PID control parameters

[0151]

[0152] It should be noted that Table 1 is an example of a scheduling strategy. The specific values ​​of the control parameters can be determined based on calibration experiments, and this embodiment of the invention does not impose any limitations on them. During phase switching, , , Instead of a step transition, a linear transition occurs along a preset ramp within a transition time. The typical transition time is 50~100ms, determined by actual vehicle calibration, to avoid sudden changes in the damping force command caused by abrupt changes in gain.

[0153] Specifically, before entering the subsidence stage, the reference damping is determined based on the reference damping MAP table, the subsidence condition level, and the vehicle speed.

[0154] Furthermore, the post-sinking stage can be divided into the entry stage and the sinking bottom stage. In the entry stage, the controller determines the damping correction for each wheel based on the vehicle's vertical acceleration, the wheel's suspension travel, and the first PID control parameters. It then determines the intermediate damping for each wheel in the entry stage based on the damping correction and the baseline damping of the post-sinking stage. Similarly, in the sinking bottom stage, the controller determines the damping correction for each wheel based on the vehicle's vertical acceleration, the wheel's suspension travel, and the second PID control parameters. Finally, it determines the intermediate damping for each wheel in the sinking bottom stage based on the damping correction and the baseline damping of the post-sinking stage.

[0155] The first and second PID control parameters mentioned above can be determined based on the scheduling strategy in Table 1, and the first and second PID control parameters are different. By further dividing the stage after entering the sinkhole into the entry stage and the sinkhole bottom stage, and using different PID control parameters to form damping correction amounts, the controller can implement segmented corrections for the impact differences between the sinkhole entry process and the sinkhole bottom process, so that each wheel obtains intermediate damping that matches the current road surface condition.

[0156] Furthermore, during the exit from the subsidence stage, the climbing phase can be divided into the early and middle stages based on the distance between the wheels and the crest of the slope, the amplitude of the vehicle's vertical acceleration, and the deviation of the suspension travel. In the early stage of the climb, the damping correction for each wheel is determined based on the vehicle's vertical acceleration, the suspension travel, and the third PID control parameter. The intermediate damping for each wheel in the early stage is then determined based on the damping correction and the baseline damping in the exit from the subsidence stage. Similarly, in the middle stage of the climb, the damping correction for each wheel is determined based on the vehicle's vertical acceleration, the suspension travel, and the fourth PID control parameter. The intermediate damping for each wheel in the middle stage is then determined based on the damping correction and the baseline damping in the exit from the subsidence stage.

[0157] In practical implementation, the distance between the wheels and the crest of the hill can be determined by the vehicle's perception data and positioning information. The vehicle's vertical acceleration and suspension travel are collected in real time by onboard sensors. After inputting the above data into the segmented judgment logic, the controller first determines whether it has entered the early or middle stage of the climb based on the distance threshold and the acceleration amplitude and travel deviation. Then, it calls the corresponding PID parameters to calculate the damping correction. The third and fourth PID control parameters mentioned above can be determined based on the scheduling strategy in Table 1. The third and fourth PID control parameters are different. By subdividing the exit from the depression stage into the early and middle stages of the climb and performing damping corrections separately, the controller can output matching intermediate damping according to the different force states when the vehicle leaves the depression area, so that the suspension can maintain continuous adjustment during the transition to the crest of the hill. This allows the damping response of the front and rear wheels or the left and right wheels to be adapted to the wheel climbing process, and improves the vehicle's attitude stability and control consistency when exiting the depression.

[0158] Finally, after exiting the sinking phase, in the early stage of attitude recovery, the PID control parameters are switched to residual vibration attenuation parameters: reduce the proportional gain, turn off the integral term, and increase the derivative gain to dominate residual vibration attenuation. The damping force smoothly transitions from the feedforward baseline to the low damping region, and the body pitch amplitude is reduced in conjunction with the air spring rebound rate.

[0159] 5) Based on the hierarchical collaborative control strategy, damping force commands are independently distributed to the four wheels: When the vehicle sinks in a single-sided sinking scenario, the intermediate damping (feedforward + PID feedback) of the impacted wheel in each sinking driving stage is used as the target damping of the impacted wheel in each sinking driving stage; and the reference damping of the non-impacted wheel in each sinking driving stage is used as the target damping of the non-impacted wheel in each sinking driving stage. When the vehicle sinks in a double-sided sinking scenario, the intermediate damping of the front axle wheel in each sinking driving stage is used as the target damping of the front axle wheel in each sinking driving stage; and based on the actual damping of the front axle wheel and the timing offset corresponding to the vehicle wheelbase, the intermediate damping of the rear axle wheel in each sinking driving stage is corrected to obtain the target damping of the rear axle wheel in each sinking driving stage.

[0160] When the vehicle sinking and passing scenario is continuous sinking, the PID integral term is cleared to zero and reinitialized after the previous sinking exit stage ends, and the target damping of the previous sinking exit stage is used as the reference damping for the next sinking entry stage.

[0161] c. Target vehicle height: Vehicle height adjustment prioritizes "smooth posture and ensures ground clearance as a safety net." Unlike fully active suspension, which can actively lift over obstacles, the height adjustment capability of the air spring + CDC / MRC semi-active suspension configuration is physically constrained by the inflation and deflation rate and travel range of the air springs. Therefore, height adjustment is not the primary means of dealing with sinking impacts, but rather a supplementary adjustment to reserve suspension buffer travel. It only performs a limited pre-lift when multiple safety conditions are met, and then smoothly returns to its original position after driving out of the sinking area.

[0162] Height adjustment can only be triggered if all five of the following conditions are met simultaneously. If any condition is not met, the current height will be locked, and subsidence control will be achieved solely through stiffness grading and damping segment adjustment: Subsidence level is moderate or higher; changes in sprung mass cause the suspension's static compression to exceed the limit; predicted dynamic travel compression may exceed 80% of the available travel; current ground clearance is higher than the minimum safe limit; vehicle speed is within the adjustable range. When any of the above conditions are met, the controller can calculate the target vehicle height based on the subsidence depth in the subsidence scenario model.

[0163] d. Multi-parameter collaborative output and arbitration mechanism: The calculation results of the three main parameters—target stiffness, target damping, and target vehicle height—are uniformly arbitrated within the controller. The arbitration rules include: smooth transition of damping force command execution during stiffness switching; moderate pre-increase of damping force during height adjustment to compensate for attitude changes; when ground clearance limit protection is triggered, stiffness is locked to high stiffness mode, height command is locked to the current value, and damping force executes marginal protection mode.

[0164] After arbitration, four-wheel independent control command messages are generated, and fault redundancy checks are performed simultaneously. The validity of the sensing data source, ECU operating status, and suspension actuator feedback signals are verified in real time. If data anomalies or communication interruptions occur, the system immediately switches to default safety control parameters to ensure basic vehicle stability. The final verified target stiffness, target damping, and target vehicle height control commands for the four-wheel independent suspension are cached and sorted according to the vehicle's driving sequence. The timing of command issuance is precisely matched based on the real-time distance between the vehicle and the sinking area, ensuring that all pre-adjustment actions are completed before the wheels reach the impact point of the sinking, thus completing the closed-loop logic of the fusion decision-making process.

[0165] 6. Wire-controlled suspension pre-adjustment execution phase: After arbitration, the control command is sent to the suspension controller according to the sinking driving phase, which drives the air spring solenoid valve group to complete the stiffness and vehicle height adjustment, drives the CDC / MRC shock absorber drive module to output continuously adjustable damping, and the air compressor and air tank cooperate to complete the air spring inflation and deflation height adjustment; at the same time, the fault detection and degradation module continuously collects the feedback signals of each actuator, and activates the graded degradation strategy when communication or execution abnormalities occur to ensure driving stability.

[0166] (1) Before entering the sinkhole - feedforward pre-adjustment of the pre-aiming drive:

[0167] During the driving phase when the vehicle approaches the sinkhole area beyond visual range, the suspension system generates suspension pre-control commands based on the sinkhole feature information obtained by fusing the vehicle cloud map and the vehicle vision and sends them to the actuators to achieve feedforward active pre-adjustment before passing through the sinkhole.

[0168] The system combines the sinking condition level with the real-time vehicle speed to switch the air springs to the target stiffness mode that matches the condition in advance. The air spring inflation and deflation are completed before the wheels reach the edge of the sinking, allowing the suspension stiffness to enter a steady-state preset state in advance. At the same time, the system controls the CDC / MRC shock absorbers to complete the preset reference damping baseline for each sinking driving stage. Relying on the millisecond-level fast response characteristics of the shock absorbers, the damping baseline is locked instantly before the vehicle enters the sinking, providing a stable reference for subsequent feedback dynamic correction.

[0169] In terms of differentiated control under different operating conditions, if the scenario is identified as a single-sided sinking and passing scenario, the full stiffness and damping pre-adjustment logic is only executed on the wheel on the impacted side, while the wheel on the non-impacted side maintains the basic suspension state, effectively avoiding the imbalance of the vehicle body's tilt posture; if the scenario is identified as a double-sided sinking and passing scenario, the pre-adjustment operation is performed on all four wheels simultaneously, focusing on enhancing the vertical vibration reduction performance of the whole vehicle and adapting to symmetrical road impact conditions in advance.

[0170] (2) After entering the sinkhole - feedforward-feedback composite regulation:

[0171] When the wheel contacts the edge of the sinkhole and enters the sinkhole area, the suspension control mode switches from pure feedforward pre-adjustment to feedforward-feedback composite adjustment mode, achieving dynamic adaptive control based on the actual impact excitation of the road surface and the real-time vehicle body status.

[0172] During the sinking and entering phase, the sudden change in road surface at the sinking edge will cause the vehicle's vertical acceleration and suspension travel to deviate rapidly from the steady-state target value. At this time, the incremental PID controller uses the first PID control parameter with the weighted comprehensive deviation of the vehicle's vertical acceleration and suspension travel as the input quantity to calculate the damping correction increment in real time. This increment is then superimposed on the preset reference damping to obtain the intermediate damping, which quickly adjusts the shock absorber damping to the high damping range, maximizing the dissipation of road impact energy and effectively suppressing the vehicle's front-end dive and impact peak.

[0173] For differentiated control under different operating conditions, in the case of single-sided sinking, the PID feedback gain of the wheel on the impacted side is adjusted to the optimal high gain range (that is, the intermediate damping after feedback adjustment is used as the target damping), while the wheel on the non-impacted side maintains the preset benchmark damping. At the same time, the damping difference of the four wheels is constrained by roll suppression to avoid vehicle attitude deviation. In the case of double-sided sinking, the front and rear axle wheels independently execute PID feedback adjustment, and the rear axle control loop uses the actual response signal of the front axle as the feedforward reference. Combined with the timing offset and amplitude fine-tuning strategy corresponding to the vehicle wheelbase, the front and rear axle suspension control is coordinated and matched.

[0174] Upon reaching the bottom of the road, the vertical vibration of the vehicle body tends to stabilize. The system automatically switches the PID steady-state control parameters (second PID control parameters), reducing the proportional gain, introducing a small integral term to eliminate steady-state error, and reducing the differential gain to prevent the suspension from over-adjusting to minor random undulations in the road surface. Simultaneously, the suspension dynamic travel margin protection logic is activated, monitoring the remaining dynamic travel margin in real time. When the dynamic travel margin falls below a preset safety threshold, damping force is actively added to reduce the risk of the suspension bottoming out and impacting the limit block. Once the dynamic travel margin returns to a safe range, the damping addition decreases linearly and smoothly exits. During this stage, the feedforward preset reference damping is the primary control basis, and the PID feedback adjustment is only used to compensate for uncertain disturbances such as vehicle load fluctuations and random road surface disturbances, ensuring the smooth operation of the suspension.

[0175] (3) Before driving out of the sinkhole - anticipation and pre-adjustment of uphill impact:

[0176] The logic at this stage is similar to that before the vehicle enters the sinkhole. As the wheels are about to exit the sinkhole, the system uses an onboard camera to identify the slope of the incline at the exit point in real time. Combined with real-time vehicle speed and sprung mass, it predicts the upcoming uphill impact. The suspension system switches from a feedforward-feedback hybrid adjustment mode to an uphill pre-adjustment mode, preparing for vehicle posture recovery in advance. The system adaptively fine-tunes the target stiffness of the air springs based on the slope and vehicle speed threshold. Under gentle slopes and low-speed driving conditions, it maintains the current stiffness; under steep slopes and high-speed driving conditions, it maintains a high-stiffness mode, providing sufficient vertical support to the vehicle body and suppressing body pitch. The stiffness adjustment command is executed before the wheels contact the exit slope, ensuring the air springs enter a steady-state operating state. Simultaneously, the CDC / MRC shock absorbers pre-set the damping baseline for uphill conditions, locking the damping baseline the instant the wheels contact the slope surface, leveraging the shock absorbers' rapid response characteristics.

[0177] The differential control logic for different operating conditions is consistent with that for the entry phase. In the single-sided sinking and exit scenario, only the impacted side performs pre-adjustment, while the non-impacted side maintains its basic state to suppress vehicle roll. In the double-sided sinking and exit scenario, all four wheels are pre-adjusted simultaneously, with a focus on optimizing the vehicle's pitch suppression capability.

[0178] (4) Driving out of the sinkhole - uphill impact response and composite adjustment:

[0179] Once the wheels contact the surface of the slope and enter the climbing driving state, the suspension system continuously adopts a feedforward-feedback composite adjustment mode to dynamically respond to the impact of the uphill road surface and the changes in the vehicle's posture.

[0180] In the early stages of climbing, the instantaneous impact from exiting the slope causes significant deviations in the vehicle's vertical acceleration and suspension travel. The incremental PID controller uses the third PID control parameter, still taking the weighted combined deviation of the vehicle's vertical acceleration and suspension travel as input, to calculate the damping correction increment in real time and add it to the reference damping. This quickly increases the damper damping, efficiently dissipates the uphill impact energy, and suppresses the vehicle's upward speed and impact peak.

[0181] In the case of single-sided sinking and exiting the vehicle, the PID feedback gain on the impacted side is maximized, while the non-impacted side maintains the reference damping, and the roll is avoided by the damping difference constraint. In the case of double-sided sinking and exiting the vehicle, the front and rear axles are independently controlled by PID, and the actual response of the front axle is used as the feedforward reference for the rear axle. The timing offset and amplitude fine-tuning are matched to achieve coordinated control.

[0182] After entering the middle of the climb, the vehicle body posture and vertical vibration gradually stabilize. The system switches back to PID steady-state parameters (fourth PID control parameter), reduces the proportional gain, introduces a small integral term to eliminate steady-state error, and weakens the derivative gain to avoid over-adjustment of small road surface undulations. Simultaneously, the suspension dynamic travel margin protection logic continues to run, dynamically supplementing or discontinuing damping compensation according to the dynamic travel margin. In this stage, the feedforward reference damping is still the main factor, supplemented by PID feedback correction, to accurately compensate for various random operating condition disturbances.

[0183] (5) After driving out of the sinkhole - attitude recovery and state reset:

[0184] After the wheels are completely detached from the sunken slope, the system initiates the process of restoring the vehicle's posture and resetting the suspension parameters, taking into account both posture stability and driving comfort.

[0185] In the initial stage of attitude recovery, the system switches to dedicated parameters for PID residual vibration attenuation. By reducing the proportional gain, disabling the integral term, and increasing the derivative gain, it rapidly attenuates residual vehicle vibration, smoothly controlling the damper damping to return from the working baseline to the low-damping range. Combined with precise control of the air spring rebound rate, this quickly smooths out the residual pitch amplitude of the vehicle. If the system identifies a continuous sinking driving scenario, it will forcibly reset the PID integral terms of each wheel and reinitialize the parameters. At the same time, it will assign the target damping of the previous sinking exit stage as the reference damping for the next sinking entry stage, achieving a smooth transition of suspension parameters under multiple continuous sinking conditions and avoiding abrupt adjustments.

[0186] Once the vehicle's posture is fully stable, the global state reset logic is executed: the vehicle height is delayed for a preset time and then gradually returns to the normal driving height at a smooth rate; the air spring stiffness gradually switches to a normal low stiffness mode based on real-time vehicle speed and road conditions, and the shock absorber reference damping is simultaneously and smoothly adjusted back. Throughout the reset process, the PID controller maintains low-gain standby operation, correcting residual vehicle vibration and minor posture deviations in real time. If a new road disturbance is detected during the reset process, the reset process is immediately terminated, and the system re-enters the feedforward-feedback composite adjustment mode.

[0187] During the entire process of the vehicle passing through the sinking condition, the controller continuously monitors the working status and feedback signals of each suspension actuator in real time. When a fault such as excessive execution deviation or communication abnormality occurs, a graded degradation control strategy is activated: when a single wheel execution is abnormal, the damping baseline of the other wheels is adjusted to compensate for the overall vehicle attitude deviation; when multiple wheels execute abnormally or the system fails, the system actively degrades to the passive basic damping control mode to ensure the stability and driving safety of the vehicle throughout the entire process.

[0188] 7. Closed-loop feedback reset stage

[0189] After all the wheels of the vehicle have completely left the sunken road surface, the vehicle sensors continuously collect vehicle vibration, attitude and suspension operation feedback signals. The main control unit executes smooth reset control logic, gradually restoring the target stiffness, target damping and target vehicle height of the suspension to the reference parameters of the normal comfort mode, eliminating secondary body bumps caused by instantaneous changes in parameters, and completing the closed-loop control of the suspension throughout the single sinking condition.

[0190] Based on multi-source perception fusion information, combined with real-time vehicle speed, wheel vertical acceleration, and suspension travel variation characteristics, the system accurately determines the vehicle's sinking and departure status. When the suspension travel of all four wheels returns to the normal steady-state fluctuation range and remains there for a preset duration (0.5s), the system confirms that the vehicle has completely left the sinking area. Based on the residual vehicle pitch angle, roll angle, and vertical velocity after departure, the system plans a smooth suspension parameter reset trajectory, gradually adjusting the damping parameters, stiffness parameters, and vehicle travel parameters of each wheel suspension back to the default comfort setting range. The overall reset process can be configured for a duration of 2s-5s, and through gradual parameter adjustment, it completely avoids residual vehicle sway and secondary attitude disturbances caused by sudden changes in suspension parameters.

[0191] The vehicle anonymizes and transmits the visually recognized subsidence parameters, actual vehicle posture, and precise positioning information obtained during the passage through the subsidence area to the cloud server. The cloud server integrates crowdsourced data from multiple vehicles to iteratively update the subsidence depth, perception confidence, and passage adaptation reference information in the overall road surface subsidence layer. This update does not rely on single-vehicle data; instead, it forms a data closed loop through iterative optimization of massive crowdsourced data, continuously improving the pre-perception accuracy of the cloud-based road surface prior information. This provides high-precision prior data support for subsequent subsidence prediction and active suspension adjustment for passing vehicles.

[0192] Once the suspension parameters and vehicle body posture have fully reset and stabilized, the suspension system switches back to the conventional real-time damping control mode and continuously monitors the forward subsidence condition warning information sent from the cloud. It can cyclically trigger the subsidence condition graded adjustment process to achieve continuous adaptive closed-loop control of road subsidence conditions.

[0193] In summary, this invention utilizes a dual-sensing technology that integrates long-distance perception via vehicle-to-the-cloud map and close-range precise identification via onboard cameras. It achieves long-distance road condition prediction using the vehicle-to-the-cloud map, precise correction of near-distance subsidence parameters using high-definition cameras, and integrates road condition, vehicle posture, and driving status data through a multi-source data fusion algorithm to build a multi-source perception fusion architecture. Simultaneously, it designs an integrated end-to-end control logic encompassing prediction, real-time correction, fusion calculation, precise control, and closed-loop reset, dynamically matching suspension damping, height, and stiffness adjustment parameters to pre-adjust the suspension before the vehicle reaches the subsidence area. After the vehicle leaves, the suspension parameters are smoothly reset based on vehicle feedback, ultimately achieving stable vehicle posture control under different levels of subsidence conditions. This solves the technical challenges of existing suspensions' inability to accurately identify road subsidence and their lagging control. This invention not only accurately predicts road subsidence conditions and pre-adjusts the suspension in advance, optimizing driving comfort and vehicle posture stability, but also possesses sensing redundancy and dynamic updates of vehicle-to-the-cloud data, meeting automotive-grade safety standards and forming a core differentiating advantage from existing "magic carpet" functions. The specific advantages are as follows:

[0194] (1) High perception accuracy and strong robustness: The dual closed-loop collaborative mechanism of "prior information + real-time perception" is adopted. The vehicle cloud map provides the front-end road condition prediction data, and the vehicle camera corrects the size and depth of the sinkhole in real time, making up for the blind spot of single perception, reducing the false detection and missed detection rate in adverse scenarios such as backlight, night, and water accumulation, and breaking through the limitation of insufficient robustness of single-machine perception of the existing "magic carpet" function.

[0195] (2) Possesses self-evolution capability: Through the vehicle-cloud closed-loop iterative update mechanism, the real-time update of road subsidence map information is realized, forming a virtuous cycle of "perception-control-feedback-update", providing more accurate reference for subsequent vehicles and improving the collaborative performance of the entire vehicle network system. This is a core advantage that the existing "magic carpet" function (single-machine control) does not possess.

[0196] (3) Strong compatibility and better cost adaptability: No need to install expensive sensors such as LiDAR. It can achieve a control effect that is better than the existing "magic carpet" function by simply cooperating with the vehicle cloud map and the existing vehicle camera. It is compatible with mainstream electromagnetic suspension and air suspension. No need to make major changes to the original vehicle chassis architecture, reducing the overall vehicle hardware cost and making it easier to achieve mass production and installation on vehicles. It solves the problems of high hardware cost and low penetration rate of the existing "magic carpet" function.

[0197] (4) High system safety: The design of multi-level perception redundancy and control degradation strategy meets the automotive-grade functional safety requirements and avoids control failure due to the failure of a single perception component; at the same time, it has an actuator protection mechanism to extend the service life of the suspension system and make up for the shortcomings of the existing "magic carpet" function lacking a complete redundancy mechanism.

[0198] Figure 5 A schematic diagram of an embodiment of the vehicle control device provided by the present invention is shown. Figure 5 As shown, the vehicle control device 300 includes: a fusion module 310, a determination module 320, and a control module 330.

[0199] The fusion module 310 is used to fuse the acquired cloud-based road subsidence data, vehicle perception data and current vehicle status data when a subsidence condition occurs on the road ahead, so as to obtain a subsidence scene model corresponding to the subsidence condition.

[0200] The determination module 320 is used to determine the suspension parameters of each wheel in each sinking driving stage based on the sinking scenario model and the current vehicle status data; wherein, the sinking driving stage represents the driving stage when the vehicle passes through the sinking working condition.

[0201] The control module 330 is used to perform phased control of the vehicle's suspension based on the suspension parameters of each wheel during each sinking driving stage.

[0202] In one alternative approach, the current vehicle state data includes vehicle speed, sprung mass, vertical acceleration of each wheel, and suspension travel of each wheel; the determination module 320 is used for:

[0203] Based on the subsidence scenario model, determine the subsidence condition level;

[0204] Based on the subsidence condition level, vehicle speed, and sprung mass, determine the target stiffness and reference damping of each wheel in each subsidence driving stage.

[0205] For each wheel, the reference damping of each sinking driving stage is corrected based on the wheel's vertical acceleration, the wheel's suspension travel, and the PID control parameters corresponding to each sinking driving stage, so as to obtain the intermediate damping of the wheel in each sinking driving stage.

[0206] Based on the vehicle's sinking and driving scenario, the intermediate damping of each wheel is corrected in each sinking and driving stage to obtain the target damping of each wheel in each sinking and driving stage; wherein, the suspension parameters in each sinking and driving stage include the target stiffness and the target damping in each sinking and driving stage.

[0207] In one alternative approach, module 320 is defined for:

[0208] The target stiffness of each wheel in each stage of settlement driving is determined based on the settlement condition level, vehicle speed and sprung mass.

[0209] Based on the subsidence condition level, vehicle speed, and target stiffness, determine the reference damping of each wheel in each subsidence driving stage.

[0210] In one alternative approach, the subsidence scenario model includes subsidence depth, subsidence edge slope, and subsidence geometric parameters; the determination module 320 is used for:

[0211] The settlement condition level is determined based on the settlement depth, settlement edge slope, and settlement geometric parameters.

[0212] In one alternative approach, module 320 is defined for:

[0213] When the vehicle sinking and passing scenario is a single-sided sinking, the intermediate damping of the wheel on the impacted side in each sinking driving stage is taken as the target damping of the wheel on the impacted side in each sinking driving stage; and the reference damping of the wheel on the non-impacted side in each sinking driving stage is taken as the target damping of the wheel on the non-impacted side in each sinking driving stage.

[0214] When the vehicle sinks and passes through a sinking scenario with both sides sinking, the intermediate damping of the front axle wheels in each sinking driving stage is taken as the target damping of the front axle wheels in each sinking driving stage; and based on the actual damping of the front axle wheels and the time offset corresponding to the vehicle wheelbase, the intermediate damping of the rear axle wheels in each sinking driving stage is corrected to obtain the target damping of the rear axle wheels in each sinking driving stage.

[0215] In one alternative approach, the sinking driving phase is the post-sinking driving phase; the determining module 320 is used for:

[0216] In the entry section after entering the sinkhole, the damping correction amount of each wheel in the entry section after entering the sinkhole is determined based on the vertical acceleration of the vehicle body, the suspension dynamic travel of the wheel, and the first PID control parameters; and the intermediate damping of each wheel in the entry section after entering the sinkhole is determined based on the damping correction amount in the entry section after entering the sinkhole and the reference damping in the stage after entering the sinkhole.

[0217] After entering the bottom of the sinkhole, the damping correction amount of each wheel is determined based on the vertical acceleration of the vehicle body, the suspension travel of the wheel, and the second PID control parameters. Then, based on the damping correction amount and the reference damping of the bottom of the sinkhole, the intermediate damping of each wheel is determined.

[0218] In one alternative approach, the sinking driving phase is the exit from the sinking phase; the determining module 320 is used for:

[0219] Based on the distance between the wheels and the crest of the slope in the sinking condition, the vertical acceleration amplitude of the vehicle body, and the deviation of the suspension travel, the climbing phase and the climbing phase are divided.

[0220] In the early stage of climbing out of the sinkhole, the damping correction of each wheel in the early stage of climbing is determined based on the vertical acceleration of the vehicle body, the suspension travel of the wheel, and the third PID control parameters; and the intermediate damping of each wheel in the early stage of climbing is determined based on the damping correction in the early stage of climbing and the reference damping in the stage of exiting the sinkhole.

[0221] During the mid-climb phase after exiting the sinkhole, the damping correction amount for each wheel during the mid-climb phase is determined based on the vehicle's vertical acceleration, the wheel's suspension travel, and the fourth PID control parameters. Then, based on the damping correction amount during the mid-climb phase and the reference damping during the exit from the sinkhole phase, the intermediate damping for each wheel during the mid-climb phase is determined.

[0222] In one alternative approach, module 320 is defined for:

[0223] For each stage of sinking and driving, the current deviation value is determined based on the vertical acceleration of the vehicle body and the suspension travel of the wheel;

[0224] Based on the current deviation value and the PID control parameters corresponding to this sinking driving stage, determine the current damping correction increment of the wheel;

[0225] The current damping correction of the wheel is obtained based on the damping correction amount at the previous moment and the current damping correction increment.

[0226] The intermediate damping of the wheel during this sinking driving phase is obtained based on the wheel's current damping correction and the wheel's reference damping during this sinking driving phase.

[0227] In one alternative approach, module 320 is defined for:

[0228] When the vehicle sinking and passing scenario is continuous sinking, the PID integral term is cleared to zero and reinitialized after the previous sinking exit stage ends, and the target damping of the previous sinking exit stage is used as the reference damping for the next sinking entry stage.

[0229] In one alternative approach, the suspension parameters also include the target vehicle height; the determining module 320 is used for:

[0230] When the subsidence condition level and / or the current vehicle status data meet the preset conditions, the target vehicle height of each wheel in each subsidence driving stage is determined according to the subsidence depth in the subsidence scenario model.

[0231] The preset conditions include any of the following:

[0232] The subsidence condition level is higher than the preset level;

[0233] The static compression of the suspension exceeds the preset static compression threshold.

[0234] The predicted suspension dynamic travel compression is greater than the preset dynamic compression threshold.

[0235] The current ground clearance is greater than the safety limit threshold;

[0236] The vehicle speed is within the preset range;

[0237] The vehicle's suspension is controlled based on the suspension parameters of each wheel during each stage of sinking and driving, including:

[0238] The vehicle's suspension is controlled in stages based on the target stiffness, target damping, and target vehicle height of each wheel during each sinking and driving phase.

[0239] In one alternative embodiment, the fusion module 310 is used for:

[0240] By fusing cloud-based road subsidence data and vehicle-sensing data, subsidence characteristic information can be obtained.

[0241] By aligning the subsidence feature information with the current vehicle status data, a subsidence scenario model corresponding to the subsidence condition is obtained.

[0242] In one alternative embodiment, the fusion module 310 is used for:

[0243] Based on the data age of the cloud-based road subsidence data and the confidence level of the vehicle perception data, the first weighting coefficient of the cloud-based road subsidence data and the second weighting coefficient of the vehicle perception data are determined.

[0244] Based on the first and second weighting coefficients, the cloud-based road subsidence data and the vehicle perception data are weighted and fused to obtain subsidence feature information.

[0245] As can be seen from the above, the vehicle control device provided in this embodiment of the invention can combine the advantages of vehicle-cloud multi-source fusion perception to make up for the defects of low accuracy and weak anti-interference ability of single visual perception, predict the sinking conditions in advance and realize differentiated and phased active control of the four wheels, effectively weaken the impact disturbance when the vehicle passes through the sinking road surface, suppress the attitude fluctuations such as body pitch and roll, and greatly improve the chassis control accuracy and control robustness under the harsh sinking conditions. While ensuring the vehicle driving stability and chassis structural safety, it significantly optimizes the vehicle driving comfort.

[0246] Figure 6 The diagram shows a structural schematic of an embodiment of the electronic device provided by the present invention. The specific embodiments of the present invention do not limit the specific implementation of the electronic device.

[0247] like Figure 6 As shown, the electronic device 400 may include: a processor 402, a communications interface 404, a memory 406, and a communications bus 408.

[0248] The processor 402, communication interface 404, and memory 406 communicate with each other via communication bus 408. Communication interface 404 is used to communicate with other network elements such as clients or other servers. The processor 402 executes program 410, specifically performing the relevant steps described above in the vehicle control method embodiment.

[0249] Specifically, program 410 may include program code, which includes computer-executable instructions.

[0250] Processor 402 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention. The electronic device includes one or more processors, which may be processors of the same type, such as one or more CPUs; or processors of different types, such as one or more CPUs and one or more ASICs.

[0251] Memory 406 is used to store program 410. Memory 406 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0252] Specifically, program 410 can be called by processor 402 to cause the electronic device to perform the following operations:

[0253] When a road subsidence occurs ahead, the acquired cloud-based road subsidence data, vehicle perception data, and current vehicle status data are fused together to obtain a subsidence scenario model corresponding to the subsidence condition.

[0254] Based on the subsidence scenario model and current vehicle status data, the suspension parameters of each wheel are determined in each subsidence driving stage; where the subsidence driving stage represents the driving stage when the vehicle passes through the subsidence working condition.

[0255] The vehicle's suspension is controlled in stages based on the suspension parameters of each wheel during each sinking driving phase.

[0256] In one alternative approach, the current vehicle state data includes vehicle speed, sprung mass, vertical acceleration of each wheel, and suspension travel of each wheel; based on the subsidence scenario model and the current vehicle state data, the suspension parameters for each wheel during each subsidence driving stage are determined, including:

[0257] Based on the subsidence scenario model, determine the subsidence condition level;

[0258] Based on the subsidence condition level, vehicle speed, and sprung mass, determine the target stiffness and reference damping of each wheel in each subsidence driving stage.

[0259] For each wheel, the reference damping of each sinking driving stage is corrected based on the wheel's vertical acceleration, the wheel's suspension travel, and the PID control parameters corresponding to each sinking driving stage, so as to obtain the intermediate damping of the wheel in each sinking driving stage.

[0260] Based on the vehicle's sinking and driving scenario, the intermediate damping of each wheel is corrected in each sinking and driving stage to obtain the target damping of each wheel in each sinking and driving stage; wherein, the suspension parameters in each sinking and driving stage include the target stiffness and the target damping in each sinking and driving stage.

[0261] In one alternative approach, the target stiffness and reference damping of each wheel during each stage of settlement travel are determined based on the settlement condition level, vehicle speed, and sprung mass, including:

[0262] The target stiffness of each wheel in each stage of settlement driving is determined based on the settlement condition level, vehicle speed and sprung mass.

[0263] Based on the subsidence condition level, vehicle speed, and target stiffness, determine the reference damping of each wheel in each subsidence driving stage.

[0264] In one alternative approach, the subsidence scenario model includes subsidence depth, subsidence edge slope, and subsidence geometric parameters; based on the subsidence scenario model, the subsidence condition level is determined, including:

[0265] The settlement condition level is determined based on the settlement depth, settlement edge slope, and settlement geometric parameters.

[0266] In one alternative approach, based on the vehicle's sinking and traffic scenario, the intermediate damping of each wheel is corrected for each sinking and traffic stage to obtain the target damping of each wheel for each sinking and traffic stage, including:

[0267] When the vehicle sinking and passing scenario is a single-sided sinking, the intermediate damping of the wheel on the impacted side in each sinking driving stage is taken as the target damping of the wheel on the impacted side in each sinking driving stage; and the reference damping of the wheel on the non-impacted side in each sinking driving stage is taken as the target damping of the wheel on the non-impacted side in each sinking driving stage.

[0268] When the vehicle sinks and passes through a sinking scenario with both sides sinking, the intermediate damping of the front axle wheels in each sinking driving stage is taken as the target damping of the front axle wheels in each sinking driving stage; and based on the actual damping of the front axle wheels and the time offset corresponding to the vehicle wheelbase, the intermediate damping of the rear axle wheels in each sinking driving stage is corrected to obtain the target damping of the rear axle wheels in each sinking driving stage.

[0269] In one alternative approach, the sinking driving stage is the stage after entering the sinking area; based on the vehicle's vertical acceleration, the wheel's suspension travel, and the PID control parameters corresponding to each sinking driving stage, the reference damping for each sinking driving stage is corrected to obtain the intermediate damping of the wheel in each sinking driving stage, including:

[0270] In the entry section after entering the sinkhole, the damping correction amount of each wheel in the entry section after entering the sinkhole is determined based on the vertical acceleration of the vehicle body, the suspension dynamic travel of the wheel, and the first PID control parameters; and the intermediate damping of each wheel in the entry section after entering the sinkhole is determined based on the damping correction amount in the entry section after entering the sinkhole and the reference damping in the stage after entering the sinkhole.

[0271] After entering the bottom of the sinkhole, the damping correction amount of each wheel is determined based on the vertical acceleration of the vehicle body, the suspension travel of the wheel, and the second PID control parameters. Then, based on the damping correction amount and the reference damping of the bottom of the sinkhole, the intermediate damping of each wheel is determined.

[0272] In one alternative approach, the sinking driving phase is the exiting sinking phase; based on the vehicle's vertical acceleration, the wheel's suspension travel, and the PID control parameters corresponding to each sinking driving phase, the reference damping for each sinking driving phase is corrected to obtain the intermediate damping of the wheel in each sinking driving phase, including:

[0273] Based on the distance between the wheels and the crest of the slope in the sinking condition, the vertical acceleration amplitude of the vehicle body, and the deviation of the suspension travel, the climbing phase and the climbing phase are divided.

[0274] In the early stage of climbing out of the sinkhole, the damping correction of each wheel in the early stage of climbing is determined based on the vertical acceleration of the vehicle body, the suspension travel of the wheel, and the third PID control parameters; and the intermediate damping of each wheel in the early stage of climbing is determined based on the damping correction in the early stage of climbing and the reference damping in the stage of exiting the sinkhole.

[0275] During the mid-climb phase after exiting the sinkhole, the damping correction amount for each wheel during the mid-climb phase is determined based on the vehicle's vertical acceleration, the wheel's suspension travel, and the fourth PID control parameters. Then, based on the damping correction amount during the mid-climb phase and the reference damping during the exit from the sinkhole phase, the intermediate damping for each wheel during the mid-climb phase is determined.

[0276] In one alternative approach, the reference damping for each stage of slumping is corrected based on the vehicle's vertical acceleration, the wheel's suspension travel, and the PID control parameters corresponding to each slumping stage, to obtain the intermediate damping of the wheel in each slumping stage, including:

[0277] For each stage of sinking and driving, the current deviation value is determined based on the vertical acceleration of the vehicle body and the suspension travel of the wheel;

[0278] Based on the current deviation value and the PID control parameters corresponding to this sinking driving stage, determine the current damping correction increment of the wheel;

[0279] The current damping correction of the wheel is obtained based on the damping correction amount at the previous moment and the current damping correction increment.

[0280] The intermediate damping of the wheel during this sinking driving phase is obtained based on the wheel's current damping correction and the wheel's reference damping during this sinking driving phase.

[0281] In one alternative approach, the method further includes:

[0282] When the vehicle sinking and passing scenario is continuous sinking, the PID integral term is cleared to zero and reinitialized after the previous sinking exit stage ends, and the target damping of the previous sinking exit stage is used as the reference damping for the next sinking entry stage.

[0283] In one alternative approach, the suspension parameters also include the target vehicle height; the method further includes:

[0284] When the subsidence condition level and / or the current vehicle status data meet the preset conditions, the target vehicle height of each wheel in each subsidence driving stage is determined according to the subsidence depth in the subsidence scenario model.

[0285] The preset conditions include any of the following:

[0286] The subsidence condition level is higher than the preset level;

[0287] The static compression of the suspension exceeds the preset static compression threshold.

[0288] The predicted suspension dynamic travel compression is greater than the preset dynamic compression threshold.

[0289] The current ground clearance is greater than the safety limit threshold;

[0290] The vehicle speed is within the preset range;

[0291] The vehicle's suspension is controlled based on the suspension parameters of each wheel during each stage of sinking and driving, including:

[0292] The vehicle's suspension is controlled in stages based on the target stiffness, target damping, and target vehicle height of each wheel during each sinking and driving phase.

[0293] In one alternative approach, the acquired cloud-based road subsidence data, vehicle perception data, and current vehicle status data are fused to obtain a subsidence scenario model corresponding to the subsidence condition, including:

[0294] By fusing cloud-based road subsidence data and vehicle-sensing data, subsidence characteristic information can be obtained.

[0295] By aligning the subsidence feature information with the current vehicle status data, a subsidence scenario model corresponding to the subsidence condition is obtained.

[0296] In one alternative approach, cloud-based road subsidence data and vehicle-sensed data are fused to obtain subsidence feature information, including:

[0297] Based on the data age of the cloud-based road subsidence data and the confidence level of the vehicle perception data, the first weighting coefficient of the cloud-based road subsidence data and the second weighting coefficient of the vehicle perception data are determined.

[0298] Based on the first and second weighting coefficients, the cloud-based road subsidence data and the vehicle perception data are weighted and fused to obtain subsidence feature information.

[0299] As can be seen from the above, the electronic device provided in the embodiments of the present invention can combine the advantages of vehicle-cloud multi-source fusion perception to make up for the defects of low accuracy and weak anti-interference ability of single visual perception, predict the sinking conditions in advance and realize the four-wheel differentiated and phased active control, effectively weaken the impact disturbance when the vehicle passes through the sinking road surface, suppress the attitude fluctuations such as body pitch and roll, and greatly improve the chassis control accuracy and control robustness under the harsh sinking conditions. While ensuring the vehicle driving stability and chassis structure safety, it significantly optimizes the vehicle driving comfort.

[0300] This invention provides a vehicle, including a vehicle body and an electronic device, the electronic device being used to perform the following operations:

[0301] When a road subsidence occurs ahead, the acquired cloud-based road subsidence data, vehicle perception data, and current vehicle status data are fused together to obtain a subsidence scenario model corresponding to the subsidence condition.

[0302] Based on the subsidence scenario model and current vehicle status data, the suspension parameters of each wheel are determined in each subsidence driving stage; where the subsidence driving stage represents the driving stage when the vehicle passes through the subsidence working condition.

[0303] The vehicle's suspension is controlled in stages based on the suspension parameters of each wheel during each sinking driving phase.

[0304] In one alternative approach, the current vehicle state data includes vehicle speed, sprung mass, vertical acceleration of each wheel, and suspension travel of each wheel; based on the subsidence scenario model and the current vehicle state data, the suspension parameters for each wheel during each subsidence driving stage are determined, including:

[0305] Based on the subsidence scenario model, determine the subsidence condition level;

[0306] Based on the subsidence condition level, vehicle speed, and sprung mass, determine the target stiffness and reference damping of each wheel in each subsidence driving stage.

[0307] For each wheel, the reference damping of each sinking driving stage is corrected based on the wheel's vertical acceleration, the wheel's suspension travel, and the PID control parameters corresponding to each sinking driving stage, so as to obtain the intermediate damping of the wheel in each sinking driving stage.

[0308] Based on the vehicle's sinking and driving scenario, the intermediate damping of each wheel is corrected in each sinking and driving stage to obtain the target damping of each wheel in each sinking and driving stage; wherein, the suspension parameters in each sinking and driving stage include the target stiffness and the target damping in each sinking and driving stage.

[0309] In one alternative approach, the target stiffness and reference damping of each wheel during each stage of settlement travel are determined based on the settlement condition level, vehicle speed, and sprung mass, including:

[0310] The target stiffness of each wheel in each stage of settlement driving is determined based on the settlement condition level, vehicle speed and sprung mass.

[0311] Based on the subsidence condition level, vehicle speed, and target stiffness, determine the reference damping of each wheel in each subsidence driving stage.

[0312] In one alternative approach, the subsidence scenario model includes subsidence depth, subsidence edge slope, and subsidence geometric parameters; based on the subsidence scenario model, the subsidence condition level is determined, including:

[0313] The settlement condition level is determined based on the settlement depth, settlement edge slope, and settlement geometric parameters.

[0314] In one alternative approach, based on the vehicle's sinking and traffic scenario, the intermediate damping of each wheel is corrected for each sinking and traffic stage to obtain the target damping of each wheel for each sinking and traffic stage, including:

[0315] When the vehicle sinking and passing scenario is a single-sided sinking, the intermediate damping of the wheel on the impacted side in each sinking driving stage is taken as the target damping of the wheel on the impacted side in each sinking driving stage; and the reference damping of the wheel on the non-impacted side in each sinking driving stage is taken as the target damping of the wheel on the non-impacted side in each sinking driving stage.

[0316] When the vehicle sinks and passes through a sinking scenario with both sides sinking, the intermediate damping of the front axle wheels in each sinking driving stage is taken as the target damping of the front axle wheels in each sinking driving stage; and based on the actual damping of the front axle wheels and the time offset corresponding to the vehicle wheelbase, the intermediate damping of the rear axle wheels in each sinking driving stage is corrected to obtain the target damping of the rear axle wheels in each sinking driving stage.

[0317] In one alternative approach, the sinking driving stage is the stage after entering the sinking area; based on the vehicle's vertical acceleration, the wheel's suspension travel, and the PID control parameters corresponding to each sinking driving stage, the reference damping for each sinking driving stage is corrected to obtain the intermediate damping of the wheel in each sinking driving stage, including:

[0318] In the entry section after entering the sinkhole, the damping correction amount of each wheel in the entry section after entering the sinkhole is determined based on the vertical acceleration of the vehicle body, the suspension dynamic travel of the wheel, and the first PID control parameters; and the intermediate damping of each wheel in the entry section after entering the sinkhole is determined based on the damping correction amount in the entry section after entering the sinkhole and the reference damping in the stage after entering the sinkhole.

[0319] After entering the bottom of the sinkhole, the damping correction amount of each wheel is determined based on the vertical acceleration of the vehicle body, the suspension travel of the wheel, and the second PID control parameters. Then, based on the damping correction amount and the reference damping of the bottom of the sinkhole, the intermediate damping of each wheel is determined.

[0320] In one alternative approach, the sinking driving phase is the exiting sinking phase; based on the vehicle's vertical acceleration, the wheel's suspension travel, and the PID control parameters corresponding to each sinking driving phase, the reference damping for each sinking driving phase is corrected to obtain the intermediate damping of the wheel in each sinking driving phase, including:

[0321] Based on the distance between the wheels and the crest of the slope in the sinking condition, the vertical acceleration amplitude of the vehicle body, and the deviation of the suspension travel, the climbing phase and the climbing phase are divided.

[0322] In the early stage of climbing out of the sinkhole, the damping correction of each wheel in the early stage of climbing is determined based on the vertical acceleration of the vehicle body, the suspension travel of the wheel, and the third PID control parameters; and the intermediate damping of each wheel in the early stage of climbing is determined based on the damping correction in the early stage of climbing and the reference damping in the stage of exiting the sinkhole.

[0323] During the mid-climb phase after exiting the sinkhole, the damping correction amount for each wheel during the mid-climb phase is determined based on the vehicle's vertical acceleration, the wheel's suspension travel, and the fourth PID control parameters. Then, based on the damping correction amount during the mid-climb phase and the reference damping during the exit from the sinkhole phase, the intermediate damping for each wheel during the mid-climb phase is determined.

[0324] In one alternative approach, the reference damping for each stage of slumping is corrected based on the vehicle's vertical acceleration, the wheel's suspension travel, and the PID control parameters corresponding to each slumping stage, to obtain the intermediate damping of the wheel in each slumping stage, including:

[0325] For each stage of sinking and driving, the current deviation value is determined based on the vertical acceleration of the vehicle body and the suspension travel of the wheel;

[0326] Based on the current deviation value and the PID control parameters corresponding to this sinking driving stage, determine the current damping correction increment of the wheel;

[0327] The current damping correction of the wheel is obtained based on the damping correction amount at the previous moment and the current damping correction increment.

[0328] The intermediate damping of the wheel during this sinking driving phase is obtained based on the wheel's current damping correction and the wheel's reference damping during this sinking driving phase.

[0329] In one alternative approach, the method further includes:

[0330] When the vehicle sinking and passing scenario is continuous sinking, the PID integral term is cleared to zero and reinitialized after the previous sinking exit stage ends, and the target damping of the previous sinking exit stage is used as the reference damping for the next sinking entry stage.

[0331] In one alternative approach, the suspension parameters also include the target vehicle height; the method further includes:

[0332] When the subsidence condition level and / or the current vehicle status data meet the preset conditions, the target vehicle height of each wheel in each subsidence driving stage is determined according to the subsidence depth in the subsidence scenario model.

[0333] The preset conditions include any of the following:

[0334] The subsidence condition level is higher than the preset level;

[0335] The static compression of the suspension exceeds the preset static compression threshold.

[0336] The predicted suspension dynamic travel compression is greater than the preset dynamic compression threshold.

[0337] The current ground clearance is greater than the safety limit threshold;

[0338] The vehicle speed is within the preset range;

[0339] The vehicle's suspension is controlled based on the suspension parameters of each wheel during each stage of sinking and driving, including:

[0340] The vehicle's suspension is controlled in stages based on the target stiffness, target damping, and target vehicle height of each wheel during each sinking and driving phase.

[0341] In one alternative approach, the acquired cloud-based road subsidence data, vehicle perception data, and current vehicle status data are fused to obtain a subsidence scenario model corresponding to the subsidence condition, including:

[0342] By fusing cloud-based road subsidence data and vehicle-sensing data, subsidence characteristic information can be obtained.

[0343] By aligning the subsidence feature information with the current vehicle status data, a subsidence scenario model corresponding to the subsidence condition is obtained.

[0344] In one alternative approach, cloud-based road subsidence data and vehicle-sensed data are fused to obtain subsidence feature information, including:

[0345] Based on the data age of the cloud-based road subsidence data and the confidence level of the vehicle perception data, the first weighting coefficient of the cloud-based road subsidence data and the second weighting coefficient of the vehicle perception data are determined.

[0346] Based on the first and second weighting coefficients, the cloud-based road subsidence data and the vehicle perception data are weighted and fused to obtain subsidence feature information.

[0347] This invention provides a computer-readable storage medium storing at least one executable instruction that, when executed on an electronic device / vehicle control device, causes the electronic device / vehicle control device to perform the vehicle control method in any of the above-described method embodiments.

[0348] Specifically, executable instructions can be used to cause electronic devices / vehicle control units to perform the following operations:

[0349] When a road subsidence occurs ahead, the acquired cloud-based road subsidence data, vehicle perception data, and current vehicle status data are fused together to obtain a subsidence scenario model corresponding to the subsidence condition.

[0350] Based on the subsidence scenario model and current vehicle status data, the suspension parameters of each wheel in each subsidence driving stage are determined; where the subsidence driving stage represents the driving stage when the vehicle passes through the subsidence condition.

[0351] The vehicle's suspension is controlled in stages based on the suspension parameters of each wheel during each sinking driving phase.

[0352] In one alternative approach, the current vehicle state data includes vehicle speed, sprung mass, vertical acceleration of each wheel, and suspension travel of each wheel; based on the subsidence scenario model and the current vehicle state data, the suspension parameters for each wheel during each subsidence driving stage are determined, including:

[0353] Based on the subsidence scenario model, determine the subsidence condition level;

[0354] Based on the subsidence condition level, vehicle speed, and sprung mass, determine the target stiffness and reference damping of each wheel in each subsidence driving stage.

[0355] For each wheel, the reference damping of each sinking driving stage is corrected based on the wheel's vertical acceleration, the wheel's suspension travel, and the PID control parameters corresponding to each sinking driving stage, so as to obtain the intermediate damping of the wheel in each sinking driving stage.

[0356] Based on the vehicle's sinking and driving scenario, the intermediate damping of each wheel is corrected in each sinking and driving stage to obtain the target damping of each wheel in each sinking and driving stage; wherein, the suspension parameters in each sinking and driving stage include the target stiffness and the target damping in each sinking and driving stage.

[0357] In one alternative approach, the target stiffness and reference damping of each wheel during each stage of settlement travel are determined based on the settlement condition level, vehicle speed, and sprung mass, including:

[0358] The target stiffness of each wheel in each stage of settlement driving is determined based on the settlement condition level, vehicle speed and sprung mass.

[0359] Based on the subsidence condition level, vehicle speed, and target stiffness, determine the reference damping of each wheel in each subsidence driving stage.

[0360] In one alternative approach, the subsidence scenario model includes subsidence depth, subsidence edge slope, and subsidence geometric parameters; based on the subsidence scenario model, the subsidence condition level is determined, including:

[0361] The settlement condition level is determined based on the settlement depth, settlement edge slope, and settlement geometric parameters.

[0362] In one alternative approach, based on the vehicle's sinking and traffic scenario, the intermediate damping of each wheel is corrected for each sinking and traffic stage to obtain the target damping of each wheel for each sinking and traffic stage, including:

[0363] When the vehicle sinking and passing scenario is a single-sided sinking, the intermediate damping of the wheel on the impacted side in each sinking driving stage is taken as the target damping of the wheel on the impacted side in each sinking driving stage; and the reference damping of the wheel on the non-impacted side in each sinking driving stage is taken as the target damping of the wheel on the non-impacted side in each sinking driving stage.

[0364] When the vehicle sinks and passes through a sinking scenario with both sides sinking, the intermediate damping of the front axle wheels in each sinking driving stage is taken as the target damping of the front axle wheels in each sinking driving stage; and based on the actual damping of the front axle wheels and the time offset corresponding to the vehicle wheelbase, the intermediate damping of the rear axle wheels in each sinking driving stage is corrected to obtain the target damping of the rear axle wheels in each sinking driving stage.

[0365] In one alternative approach, the sinking driving stage is the stage after entering the sinking area; based on the vehicle's vertical acceleration, the wheel's suspension travel, and the PID control parameters corresponding to each sinking driving stage, the reference damping for each sinking driving stage is corrected to obtain the intermediate damping of the wheel in each sinking driving stage, including:

[0366] In the entry section after entering the sinkhole, the damping correction amount of each wheel in the entry section after entering the sinkhole is determined based on the vertical acceleration of the vehicle body, the suspension dynamic travel of the wheel, and the first PID control parameters; and the intermediate damping of each wheel in the entry section after entering the sinkhole is determined based on the damping correction amount in the entry section after entering the sinkhole and the reference damping in the stage after entering the sinkhole.

[0367] After entering the bottom of the sinkhole, the damping correction amount of each wheel is determined based on the vertical acceleration of the vehicle body, the suspension travel of the wheel, and the second PID control parameters. Then, based on the damping correction amount and the reference damping of the bottom of the sinkhole, the intermediate damping of each wheel is determined.

[0368] In one alternative approach, the sinking driving phase is the exiting sinking phase; based on the vehicle's vertical acceleration, the wheel's suspension travel, and the PID control parameters corresponding to each sinking driving phase, the reference damping for each sinking driving phase is corrected to obtain the intermediate damping of the wheel in each sinking driving phase, including:

[0369] Based on the distance between the wheels and the crest of the slope in the sinking condition, the vertical acceleration amplitude of the vehicle body, and the deviation of the suspension travel, the climbing phase and the climbing phase are divided.

[0370] In the early stage of climbing out of the sinkhole, the damping correction of each wheel in the early stage of climbing is determined based on the vertical acceleration of the vehicle body, the suspension travel of the wheel, and the third PID control parameters; and the intermediate damping of each wheel in the early stage of climbing is determined based on the damping correction in the early stage of climbing and the reference damping in the stage of exiting the sinkhole.

[0371] During the mid-climb phase after exiting the sinkhole, the damping correction amount for each wheel during the mid-climb phase is determined based on the vehicle's vertical acceleration, the wheel's suspension travel, and the fourth PID control parameters. Then, based on the damping correction amount during the mid-climb phase and the reference damping during the exit from the sinkhole phase, the intermediate damping for each wheel during the mid-climb phase is determined.

[0372] In one alternative approach, the reference damping for each stage of slumping is corrected based on the vehicle's vertical acceleration, the wheel's suspension travel, and the PID control parameters corresponding to each slumping stage, to obtain the intermediate damping of the wheel in each slumping stage, including:

[0373] For each stage of sinking and driving, the current deviation value is determined based on the vertical acceleration of the vehicle body and the suspension travel of the wheel;

[0374] Based on the current deviation value and the PID control parameters corresponding to this sinking driving stage, determine the current damping correction increment of the wheel;

[0375] The current damping correction of the wheel is obtained based on the damping correction amount at the previous moment and the current damping correction increment.

[0376] The intermediate damping of the wheel during this sinking driving phase is obtained based on the wheel's current damping correction and the wheel's reference damping during this sinking driving phase.

[0377] In one alternative approach, the method further includes:

[0378] When the vehicle sinking and passing scenario is continuous sinking, the PID integral term is cleared to zero and reinitialized after the previous sinking exit stage ends, and the target damping of the previous sinking exit stage is used as the reference damping for the next sinking entry stage.

[0379] In one alternative approach, the suspension parameters also include the target vehicle height; the method further includes:

[0380] When the subsidence condition level and / or the current vehicle status data meet the preset conditions, the target vehicle height of each wheel in each subsidence driving stage is determined according to the subsidence depth in the subsidence scenario model.

[0381] The preset conditions include any of the following:

[0382] The subsidence condition level is higher than the preset level;

[0383] The static compression of the suspension exceeds the preset static compression threshold.

[0384] The predicted suspension dynamic travel compression is greater than the preset dynamic compression threshold.

[0385] The current ground clearance is greater than the safety limit threshold;

[0386] The vehicle speed is within the preset range;

[0387] The vehicle's suspension is controlled based on the suspension parameters of each wheel during each stage of sinking and driving, including:

[0388] The vehicle's suspension is controlled in stages based on the target stiffness, target damping, and target vehicle height of each wheel during each sinking and driving phase.

[0389] In one alternative approach, the acquired cloud-based road subsidence data, vehicle perception data, and current vehicle status data are fused to obtain a subsidence scenario model corresponding to the subsidence condition, including:

[0390] By fusing cloud-based road subsidence data and vehicle-sensing data, subsidence characteristic information can be obtained.

[0391] By aligning the subsidence feature information with the current vehicle status data, a subsidence scenario model corresponding to the subsidence condition is obtained.

[0392] In one alternative approach, cloud-based road subsidence data and vehicle-sensed data are fused to obtain subsidence feature information, including:

[0393] Based on the data age of the cloud-based road subsidence data and the confidence level of the vehicle perception data, the first weighting coefficient of the cloud-based road subsidence data and the second weighting coefficient of the vehicle perception data are determined.

[0394] Based on the first and second weighting coefficients, the cloud-based road subsidence data and the vehicle perception data are weighted and fused to obtain subsidence feature information.

[0395] As can be seen from the above, the computer-readable storage medium provided in the embodiments of the present invention stores at least one executable instruction. When the executable instruction runs on the electronic device / vehicle control device, it can combine the advantages of vehicle-cloud multi-source fusion perception to make up for the defects of low accuracy and weak anti-interference ability of single visual perception, predict the sinking conditions in advance and realize differentiated and phased active control of the four wheels, effectively weaken the impact disturbance when the vehicle passes through the sinking road surface, suppress the attitude fluctuations such as body pitch and roll, and significantly improve the chassis control accuracy and control robustness under the harsh sinking conditions. While ensuring the vehicle driving stability and chassis structural safety, it significantly optimizes the vehicle driving comfort.

[0396] The algorithms or displays provided herein are not inherently related to any particular computer, virtual system, or other device. Furthermore, the embodiments of this invention are not directed to any particular programming language.

[0397] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of the invention may be practiced without these specific details. Similarly, in order to streamline the invention and aid in understanding one or more aspects of the invention, in the description of exemplary embodiments of the invention above, various features of the embodiments of the invention are sometimes grouped together in a single embodiment, figure, or description thereof.

[0398] Those skilled in the art will understand that the modules in the device of the embodiment can be adaptively changed and placed in one or more devices different from that embodiment. Modules, units, or components in the embodiment can be combined into a single module, unit, or component, and further, they can be divided into multiple sub-modules, sub-units, or sub-components, except that at least some of such features and / or processes or units are mutually exclusive.

[0399] It should be noted that the above embodiments are illustrative of the invention and not restrictive of it. The invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein. It is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. Unless otherwise specified, the steps in the above embodiments should not be construed as limiting the order of execution.

Claims

1. A vehicle control method, characterized in that, include: When a road subsidence occurs ahead, the acquired cloud-based road subsidence data, vehicle perception data, and current vehicle status data are fused together to obtain a subsidence scenario model corresponding to the subsidence condition. Based on the subsidence scenario model and the current vehicle status data, the suspension parameters of each wheel are determined in each subsidence driving stage; wherein, the subsidence driving stage represents the driving stage of the vehicle when passing through the subsidence condition. The vehicle's suspension is controlled based on the suspension parameters of each wheel during each stage of sinking.

2. The method according to claim 1, characterized in that, The current vehicle status data includes vehicle speed, sprung mass, vertical acceleration of each wheel, and suspension travel of each wheel; the determination of suspension parameters for each wheel at each stage of subsidence based on the subsidence scenario model and the current vehicle status data includes: Based on the subsidence scenario model, the subsidence condition level is determined; Based on the subsidence condition level, the vehicle speed, and the sprung mass, determine the target stiffness and reference damping of each wheel in each subsidence driving stage; For each wheel, the reference damping of each sinking driving stage is corrected based on the wheel's vertical acceleration, the wheel's suspension travel, and the PID control parameters corresponding to each sinking driving stage, so as to obtain the intermediate damping of the wheel in each sinking driving stage. Based on the vehicle's sinking and driving scenario, the intermediate damping of each wheel in each sinking and driving stage is corrected to obtain the target damping of each wheel in each sinking and driving stage; wherein, the suspension parameters of each sinking and driving stage include the target stiffness and the target damping of each sinking and driving stage.

3. The method according to claim 2, characterized in that, The determination of the target stiffness and reference damping of each wheel in each stage of subsidence driving, based on the subsidence condition level, the vehicle speed, and the sprung mass, includes: Based on the subsidence condition level, the vehicle speed, and the sprung mass, determine the target stiffness of each wheel in each subsidence driving stage; Based on the subsidence condition level, the vehicle speed, and the target stiffness, the reference damping of each wheel is determined for each subsidence driving stage.

4. The method according to claim 2, characterized in that, The subsidence scenario model includes subsidence depth, subsidence edge slope, and subsidence geometric parameters; determining the subsidence condition level based on the subsidence scenario model includes: The subsidence condition level is determined based on the subsidence depth, subsidence edge slope, and subsidence geometric parameters.

5. The method according to claim 2, characterized in that, The process involves correcting the intermediate damping of each wheel during each stage of subsidence driving, based on the vehicle's subsidence driving scenario, to obtain the target damping of each wheel during each stage of subsidence driving, including: When the sinking and passage scenario of the vehicle is a single-sided sinking, the intermediate damping of the wheel on the impact side in each sinking and passage stage is taken as the target damping of the wheel on the impact side in each sinking and passage stage; and the reference damping of the wheel on the non-impact side in each sinking and passage stage is taken as the target damping of the wheel on the non-impact side in each sinking and passage stage. When the vehicle sinks and passes through a sinking scenario with both sides sinking, the intermediate damping of the front axle wheels in each sinking driving stage is taken as the target damping of the front axle wheels in each sinking driving stage; and based on the actual damping of the front axle wheels and the timing offset corresponding to the vehicle wheelbase, the intermediate damping of the rear axle wheels in each sinking driving stage is corrected to obtain the target damping of the rear axle wheels in each sinking driving stage.

6. The method according to claim 2, characterized in that, The sinking driving stage refers to the stage after entering the sinking area; the reference damping for each sinking driving stage is corrected based on the vertical acceleration of the vehicle body, the suspension travel of the wheel, and the PID control parameters corresponding to each sinking driving stage, to obtain the intermediate damping of the wheel in each sinking driving stage, including: In the entry section after the vehicle enters the sinkhole, the damping correction amount of each wheel in the entry section after the vehicle enters the sinkhole is determined based on the vehicle's vertical acceleration, the suspension travel of the wheel, and the first PID control parameters; and the intermediate damping of each wheel in the entry section after the vehicle enters the sinkhole is determined based on the damping correction amount in the entry section after the vehicle enters the sinkhole and the reference damping in the stage after the vehicle enters the sinkhole. After entering the sinking bottom section, the damping correction amount of each wheel in the sinking bottom section is determined based on the vertical acceleration of the vehicle body of the wheel, the suspension dynamic travel of the wheel, and the second PID control parameters; and the intermediate damping of each wheel in the sinking bottom section is determined based on the damping correction amount in the sinking bottom section and the reference damping of the stage after entering the sinking.

7. The method according to claim 2, characterized in that, The sinking driving stage is the exiting sinking stage; the reference damping for each sinking driving stage is corrected based on the vehicle's vertical acceleration, the suspension travel of the wheel, and the PID control parameters corresponding to each sinking driving stage, to obtain the intermediate damping of the wheel in each sinking driving stage, including: Based on the distance between the wheels and the top of the slope in the sinking condition, the vertical acceleration amplitude of the vehicle body, and the suspension travel deviation, the early stage and middle stage of climbing are divided. In the early stage of climbing out of the sinkhole, the damping correction amount of each wheel in the early stage of climbing is determined based on the vertical acceleration of the vehicle body of the wheel, the suspension travel of the wheel, and the third PID control parameters; and the intermediate damping of each wheel in the early stage of climbing is determined based on the damping correction amount in the early stage of climbing and the reference damping in the stage of driving out of the sinkhole. During the mid-climb phase of exiting the sinkhole, the damping correction amount of each wheel during the mid-climb phase is determined based on the vertical acceleration of the vehicle body, the suspension travel of the wheel, and the fourth PID control parameter; and the intermediate damping of each wheel during the mid-climb phase is determined based on the damping correction amount during the mid-climb phase and the reference damping during the exit from the sinkhole phase.

8. The method according to claim 2, characterized in that, The method of correcting the reference damping for each sinking driving stage based on the vehicle's vertical acceleration, the wheel's suspension travel, and the PID control parameters corresponding to each sinking driving stage, to obtain the intermediate damping of the wheel in each sinking driving stage, includes: For each stage of sinking and driving, the current deviation value is determined based on the vertical acceleration of the vehicle body of the wheel and the suspension travel of the wheel; Based on the current deviation value and the PID control parameters corresponding to the sinking driving stage, determine the current damping correction increment of the wheel; The current damping correction amount of the wheel is obtained based on the damping correction amount at the previous moment and the current damping correction increment. The intermediate damping of the wheel during the sinking driving phase is obtained based on the current damping correction of the wheel and the reference damping of the wheel during the sinking driving phase.

9. A vehicle control device, characterized in that, The device includes: The fusion module is used to fuse the acquired cloud-based road subsidence data, vehicle perception data, and current vehicle status data when a subsidence condition occurs on the road ahead, in order to obtain a subsidence scenario model corresponding to the subsidence condition. The determination module is used to determine the suspension parameters of each wheel in each sinking driving stage based on the sinking scenario model and the current vehicle status data; wherein, the sinking driving stage represents the driving stage of the vehicle when passing through the sinking condition. The control module is used to control the vehicle's suspension based on the suspension parameters of each wheel during each sinking driving stage.

10. A vehicle, characterized in that, include: Vehicle body and electronic equipment; The electronic device is used to perform the operation of the vehicle control method as described in any one of claims 1-8.

Images