Vehicle door opening control method, device, vehicle, program product, and storage medium

Abstract

The application provides a vehicle door opening control method and device, a vehicle, a program product and a storage medium, relates to the technical field of vehicle safety, and aims to solve the problem that in the prior art, when a passenger opens a vehicle door in a parked state, the safety of personnel and property is seriously threatened due to negligence observation or limited field of view. The method comprises the following steps: acquiring vehicle surrounding environment data, vehicle state data and user feature data; performing multi-modal feature fusion on the vehicle surrounding environment data to generate an environment feature vector; determining risk information of opening of the vehicle door based on the environment feature vector and the vehicle state data; determining an early warning mode and an intervention strategy according to the risk information and the user feature data; controlling a warning device of the vehicle to perform early warning based on the early warning mode, and controlling opening of the vehicle door based on the intervention strategy.

Description

Technical Field

[0001] This application relates to the field of vehicle safety technology, and in particular to a door opening control method, device, vehicle, program product, and storage medium. Background Technology

[0002] With the continuous growth of car ownership and the increasing complexity of urban traffic environment, accidents (commonly known as "door-opening accidents") are frequently seen when occupants open car doors while the vehicle is parked, resulting in collisions with vehicles, pedestrians, or surrounding obstacles due to negligence in observation or limited visibility. These accidents pose a serious threat to personal safety and property. Summary of the Invention

[0003] This application provides a door opening control method, device, vehicle, program product, and storage medium, aiming to solve the problem in the prior art where, when occupants open the door while the vehicle is parked, their negligence in observation or limited field of vision poses a serious threat to personal safety and property.

[0004] In a first aspect, this application provides a door opening control method for a vehicle, the vehicle including a warning device and a door, the control method including: acquiring vehicle surrounding environment data, vehicle state data and user characteristic data; performing multimodal feature fusion on the vehicle surrounding environment data to generate an environmental feature vector; determining risk information of door opening based on the environmental feature vector and vehicle state data; determining a warning method and intervention strategy based on the risk information and user characteristic data; controlling the vehicle's warning device to issue a warning based on the warning method, and controlling the opening of the door based on the intervention strategy.

[0005] The vehicle door opening control method provided in this application comprehensively collects various information related to vehicle door opening safety by acquiring vehicle surrounding environment data, vehicle status data, and user characteristic data, providing a multi-dimensional data foundation for risk assessment. By fusing multimodal features of the vehicle surrounding environment data to generate environmental feature vectors, different types of environmental data can be organically combined to extract features that uniformly characterize the state of the environment surrounding the vehicle door, overcoming the problem of one-sided information from a single data source. Determining the risk information of door opening based on environmental feature vectors and vehicle status data allows for a comprehensive risk assessment combining environmental and vehicle status, improving the accuracy of risk assessment. Determining warning methods and intervention strategies based on risk information and user characteristic data enables personalized decision-making based on risk level and individual user characteristics, allowing safety protection strategies to adapt to the needs of different users. For example, stronger and more intuitive warnings and stricter interventions can be used for child users, clear but not overly alarming warnings can be used for elderly users, and graded responses can be implemented for ordinary adult users based on risk level. Ultimately, by implementing personalized warnings through the warning device, a complete closed loop was achieved, from environmental perception, risk prediction, user identification to personalized warning intervention, significantly improving the safety and user experience of the car door opening process.

[0006] In some possible implementations, determining door opening risk information based on environmental feature vectors and vehicle state data includes: performing predictive analysis on environmental feature vectors and vehicle state data to determine the risk level and / or the trajectory and collision time of dynamic obstacles, where the risk level includes low risk, medium risk, and high risk; and generating risk information based on the risk level and / or the trajectory and collision time of dynamic obstacles.

[0007] According to the above embodiments, by performing predictive analysis on environmental feature vectors and vehicle state data, the risk level and / or the trajectory and collision time of dynamic obstacles can be determined. This not only assesses the current risk level but also proactively predicts the trajectory and collision time of dynamic obstacles, realizing a shift from static risk assessment to dynamic risk prediction. This proactive prediction can identify potential hazards in advance, buying valuable time for preventative measures and improving the timeliness and accuracy of early warnings.

[0008] In some possible implementations, determining the early warning method and intervention strategy based on risk information and user characteristic data includes: determining the user type based on user characteristic data, where the user type includes at least one of children, the elderly, and ordinary adults; selecting the corresponding early warning method from a preset early warning method library and the corresponding intervention strategy from a preset intervention strategy library based on the user type and risk information.

[0009] According to the above embodiments, by determining user types based on user characteristic data, the individual characteristics of occupants preparing to open the vehicle door can be identified, providing a basis for personalized decision-making. Based on user type and risk information, a corresponding warning method is selected from a preset warning method library, and a corresponding intervention strategy is selected from a preset intervention strategy library. This allows for the selection of appropriate warning methods and intervention strategies based on the perception characteristics, reaction capabilities, and safety needs of different user types (such as children, the elderly, and ordinary adults), achieving personalized adaptation of safety protection, ensuring both safety and improved user experience.

[0010] In some possible implementations, acquiring vehicle surrounding environment data, vehicle status data, and user characteristic data includes: collecting vehicle surrounding environment data through multiple types of sensors, including at least one of millimeter-wave radar, high-definition camera, or ultrasonic sensor; acquiring vehicle status data through vehicle status sensors, including vehicle speed, parking posture, and door lock status; and acquiring user characteristic data through user characteristic recognition sensors, including at least one of seat weight sensor, in-vehicle camera, or voiceprint recognition module.

[0011] According to the above embodiments, by collecting environmental data around the vehicle using multiple types of sensors, multi-dimensional and multi-modal environmental data can be obtained, overcoming the shortcomings of single-sensor information being incomplete and susceptible to interference. Millimeter-wave radar excels at measuring target distance, speed, and angle information; high-definition cameras can acquire rich visual information; and ultrasonic sensors have advantages in short-range detection. The combination of multiple sensors can achieve complementary advantages. By acquiring vehicle status data through vehicle status sensors, the vehicle's own status information can be obtained, providing vehicle background for risk assessment. By acquiring user feature recognition data through user feature recognition sensors, individual user characteristics can be obtained, providing data support for personalized early warning and intervention.

[0012] In some possible implementations, multimodal feature fusion is performed on the vehicle surrounding environment data to generate an environmental feature vector, including: preprocessing the collected vehicle surrounding environment data, the preprocessing including noise reduction and data alignment; and performing cross-modal feature extraction and fusion on the preprocessed vehicle surrounding environment data to generate an environmental feature vector.

[0013] According to the above embodiments, noise reduction and data alignment processes can improve data quality and eliminate problems such as sensor noise, time asynchrony, and spatial inconsistency, providing a high-quality data foundation for subsequent feature fusion. Through cross-modal feature extraction and fusion, environmental feature vectors are generated, organically combining heterogeneous data from different types of sensors to extract feature vectors that uniformly characterize the environmental state around the vehicle door. This enables accurate perception of complex environments and provides high-quality input for subsequent risk prediction.

[0014] In some possible implementations, the warning device includes at least one of an audible warning module, a light warning module, or a tactile warning module for issuing a warning; controlling the opening of the vehicle door includes at least one of controlling the opening speed of the vehicle door, controlling the opening angle of the vehicle door, or controlling the movement state of the vehicle door.

[0015] According to the above embodiments, warnings can be issued through at least one of a sound warning module, a light warning module, or a tactile warning module, enabling a multimodal warning approach to adapt to the perceptual needs of different users in different scenarios and improve the reliability and effectiveness of the warnings. By controlling the opening speed, opening angle, or movement state of the vehicle door, intervention measures of varying intensities can be taken according to the risk level, minimizing the impact on normal use while ensuring safety, thus achieving a balance between safety and convenience.

[0016] Secondly, this application provides a control device, comprising: an acquisition module for acquiring vehicle surrounding environment data, vehicle status data, and user characteristic data; a fusion module for performing multimodal feature fusion on the vehicle surrounding environment data to generate an environmental feature vector; a first determination module for determining risk information of door opening based on the environmental feature vector and vehicle status data; a second determination module for determining a warning method and intervention strategy based on the risk information and user characteristic data; and a control module for controlling the vehicle's warning device to issue a warning based on the warning method and controlling the opening of the door based on the intervention strategy.

[0017] Thirdly, this application provides a control device, including: a memory and a processor; the memory and the processor are coupled; the memory is used to store instructions executable by the processor; when the processor executes the instructions, it implements the door opening control method of any of the above embodiments.

[0018] Fourthly, this application provides a computer-readable storage medium storing at least one piece of program code, which is loaded and executed by a processor to implement the door opening control method of any of the above embodiments.

[0019] Fifthly, this application provides a computer program product, including a computer program; when the computer program is executed, it can implement the door opening control method of any of the above embodiments.

[0020] Sixthly, this application provides a vehicle, including: a warning device for issuing a warning; a motor controller for controlling the opening of a door; and a control device according to any of the above embodiments; or a computer-readable storage medium according to any of the above embodiments; or a computer program product according to any of the above embodiments. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in this application, the accompanying drawings used in some embodiments of this application will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings.

[0022] Figure 1 A flowchart of a door opening control method provided in some embodiments of this application Figure 1 ; Figure 2 A flowchart of a door opening control method provided in some embodiments of this application Figure 2 ; Figure 3 A flowchart of a door opening control method provided in some embodiments of this application Figure 3 ; Figure 4 A flowchart of a door opening control method provided in some embodiments of this application Figure 4 ; Figure 5 Schematic diagram of the structure of the control device provided in some embodiments of this application Figure 1 ; Figure 6 Schematic diagram of the structure of the control device provided in some embodiments of this application Figure 2 . Detailed Implementation

[0023] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] It should be noted that, in this application, the terms "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or design solutions. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0025] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature.

[0026] In the description of this application, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. "And / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone. Furthermore, "at least one" means one or more, and "multiple" means two or more.

[0027] This application provides a door opening control method, which can be applied to a vehicle's domain controller, body controller, or dedicated safety control unit. This method achieves accurate risk assessment and personalized safety intervention by fusing multi-source environmental data, vehicle status data, and user characteristic data.

[0028] The door opening control method provided in the embodiments of this application will be described in detail below.

[0029] Please see Figure 1 , Figure 1 This is a flowchart illustrating a door opening control method provided in this application. Figure 1 As shown, the method includes the following steps: S110: Acquire vehicle surrounding environment data, vehicle status data, and user characteristic data.

[0030] Vehicle surrounding environment data refers to information reflecting the state of objects within the space surrounding the vehicle doors. This data can be collected using various types of sensors, including at least one of millimeter-wave radar, high-definition cameras, or ultrasonic sensors. Millimeter-wave radar can operate in all weather conditions, unaffected by light or weather, and excels at measuring the distance, speed, and angle of targets, making it particularly suitable for detecting dynamic obstacles. High-definition cameras can acquire rich visual information to identify obstacle types, outlines, textures, etc., helping to distinguish between static obstacles (such as walls and pillars) and dynamic obstacles (such as pedestrians and vehicles). Ultrasonic sensors are advantageous for short-range detection and are commonly used to detect low obstacles (such as curbs and traffic cones) near vehicle doors (e.g., within 0.1-2 meters). These sensors can reuse existing hardware in the vehicle's parking assistance system or be added specifically for door safety control. Sensors can be placed on the door body, rearview mirrors, B-pillars, or the rear of the vehicle to cover the fan-shaped area that may be swept when the door is opened.

[0031] Vehicle status data describes the vehicle's current state, including vehicle speed, parking posture, and door lock status. Vehicle speed determines whether the vehicle is stationary or coasting at low speed, a prerequisite for activating door safety controls. Parking posture information determines whether the vehicle is parked on a slope and the angle between the vehicle and the roadside, affecting the ease and risk of opening the doors. Door lock status (e.g., unlocked, locked) determines whether the door is about to be opened, triggering subsequent risk assessment procedures. Vehicle status data is typically obtained from the vehicle's CAN bus, LIN bus, or in-vehicle Ethernet.

[0032] User characteristic data is used to identify the individual characteristics of occupants preparing to open the vehicle door. This data can be acquired through user characteristic recognition sensors, including at least one of a seat weight sensor, an in-vehicle camera, or a voiceprint recognition module. A seat weight sensor, installed at the bottom of the seat, detects pressure values ​​to determine the presence of an occupant and their weight range, thus initially distinguishing between adults, children, or empty seats. An in-vehicle camera (such as one mounted above the dashboard or on the roof) captures images of the occupant and analyzes facial features such as age and gender using facial recognition algorithms to further refine the user type. A voiceprint recognition module collects the occupant's voice (such as speaking or coughing sounds) through an in-vehicle microphone and compares it with a pre-stored voiceprint model to identify specific users (such as registered elderly users). This user characteristic data forms the basis for subsequent personalized warnings and interventions.

[0033] Through the above methods, step S110 obtains comprehensive information on the three dimensions of environment, vehicle, and user, providing data support for subsequent fusion analysis and decision-making.

[0034] S120. Perform multimodal feature fusion on the vehicle's surrounding environment data to generate an environmental feature vector.

[0035] This step enables deep fusion of heterogeneous data from different types of sensors, extracting feature vectors that can uniformly characterize the state of the environment surrounding the car door.

[0036] like Figure 4 As shown, step S120 may specifically include steps S121 and S122.

[0037] S121. Preprocess the collected vehicle surrounding environment data, including noise reduction and data alignment.

[0038] Due to the inherent characteristics of sensors and environmental interference (such as rain, snow, and changes in lighting), raw data often contains noise. Noise reduction processing can improve data quality. For millimeter-wave radar data, noise reduction AI (Artificial Intelligence) algorithms based on Convolutional Neural Networks (CNNs) can be used to filter radar point clouds and remove false point clouds caused by raindrops, birds, etc. For ultrasonic data, methods such as Kalman filtering or wavelet transform can be used to remove measurement noise. For camera image data, image denoising algorithms (such as median filtering and non-local mean denoising) can be used to reduce image noise.

[0039] Data alignment can resolve issues such as inconsistent sampling times and coordinate systems between different sensors. Timestamp alignment algorithms can unify data collected by different sensors onto the same timeline, ensuring that the data used for fusion reflects the environmental state at the same moment. For example, high-precision system timestamps can be added to radar point clouds and camera image frames, and then the timestamps can be aligned to the same moment through interpolation or nearest neighbor matching.

[0040] After preprocessing, the data from different sensors were unified in terms of time, space, and quality, laying the foundation for subsequent feature fusion.

[0041] S122. Perform cross-modal feature extraction and fusion on the preprocessed vehicle surrounding environment data to generate an environmental feature vector.

[0042] Cross-modal feature extraction refers to extracting the most representative features from data of different modalities. For example, extracting distance and velocity features from millimeter-wave radar data; extracting visual features from camera image data; and extracting distance features of nearby obstacles from ultrasonic data.

[0043] Cross-modal feature fusion refers to the organic combination of features from different modalities to form a unified, high-dimensional feature representation. This can be achieved through various AI fusion models. For example, the Transformer model (a deep learning architecture based on a self-attention mechanism) can be used, which effectively captures the correlations and dependencies between features from different modalities. Specifically, radar features, visual features, and ultrasonic features can be used as different input sequences or input channels and fed into the Transformer model. The model learns the correlations between features from different modalities through multi-head attention layers and assigns higher weights to important features, ultimately outputting a unified environmental feature vector of a preset dimension (such as 256 or 512 dimensions). This feature vector condenses comprehensive information about the environment around the car door, including key information such as the location, type, and motion state of obstacles, providing high-quality input for subsequent risk prediction.

[0044] By fusing multimodal features, step S120 overcomes the shortcomings of single sensor information being one-sided and susceptible to interference, and generates environmental feature vectors that are more robust to complex environments and richer in information.

[0045] In some embodiments of this application, GNN (Graph Neural Network) can be used instead of Transformer model to improve feature fusion accuracy in complex scenes by constructing an environmental topology graph.

[0046] S130. Determine the risk information of door opening based on environmental feature vectors and vehicle status data.

[0047] This step utilizes the fused environmental feature vectors and vehicle status data to assess and predict the risks that may be encountered when opening the car door.

[0048] like Figure 2 As shown, step S130 may specifically include steps S131 and S132.

[0049] S131. Perform predictive analysis on environmental feature vectors and vehicle state data to determine the risk level and / or the trajectory and collision time of dynamic obstacles. The risk levels include low risk, medium risk, and high risk.

[0050] Predictive analytics can be achieved through AI risk prediction models. These models can employ a structure combining a Long Short-Term Memory (LSTM) network with an attention mechanism. LSTM excels at processing sequential data and can capture the movement patterns of dynamic obstacles over time. The attention mechanism helps the model focus on the features most important for risk assessment (such as the distance, speed, and direction of movement of obstacles).

[0051] The inputs to the AI ​​risk prediction model include the environmental feature vector generated in step S120 and the vehicle state data (such as vehicle speed and parking posture) obtained in step S110. The model's output may include: Risk Level: The model classifies the overall risk in the current environment and outputs a risk level as low, medium, or high. The risk level determination can consider multiple factors, such as: no obstacles or obstacles at a great distance indicate low risk; the presence of static obstacles at a manageable distance, or the presence of dynamic obstacles with a long collision time, indicates medium risk; and obstacles very close, or dynamic obstacles about to enter the collision zone, indicate high risk.

[0052] Trajectory and Time of Collision of Dynamic Obstacles: For identified dynamic obstacles (such as pedestrians, bicycles, and vehicles), the model can predict their trajectory within a future time period (e.g., 1-2 seconds). Trajectory prediction can be based on extrapolation of the obstacle's position and velocity data from the past few frames, or employ more complex trajectory prediction algorithms. Simultaneously, the model can calculate the estimated time to collision (TTC) between the dynamic obstacle and an upcoming door. TTC calculation can be estimated based on the obstacle's current position, velocity vector, and the fan-shaped area of ​​the door opening.

[0053] Through the AI ​​risk prediction model, step S131 achieves a leap from static judgment of "obstacle existence" to dynamic prediction of "risk evolution", which can identify potential dangers in advance and buy time for taking preventive measures.

[0054] S132. Generate risk information based on the risk level and / or the trajectory and collision time of dynamic obstacles.

[0055] Risk information encapsulates and transmits the results of risk assessment. It can be a structured data packet, containing at least the risk level. If the model predicts the trajectory and collision time of dynamic obstacles, these predictions can also be included in the risk information. For example, risk information can be represented as: risk level, obstacle ID, type, predicted trajectory, collision time, etc. Risk information will serve as an important input for personalized decision-making in step S140.

[0056] Through step S130, the system obtains a quantitative and forward-looking assessment of the risk of the car door opening.

[0057] S140. Determine early warning methods and intervention strategies based on risk information and user characteristic data.

[0058] This step combines risk assessment results with individual user characteristics to make personalized security decisions.

[0059] like Figure 3As shown, step S140 may specifically include steps S141 and S142.

[0060] S141. Determine the user type based on user characteristic data. The user type includes at least one of children, the elderly, and ordinary adults.

[0061] Determining the user type involves analyzing and classifying the user characteristic data obtained in step S110. The determination logic can combine data from multiple sensors for a comprehensive judgment. Child users: When the weight detected by the seat weight sensor is less than a preset weight threshold (e.g., 40kg), and the in-vehicle camera does not detect adult facial features (e.g., facial contours and proportions that match a child's), the user can be identified as a child. Child users typically have slower reaction times and weaker risk awareness, requiring the strongest, most intuitive warnings and the most stringent protection.

[0062] Elderly users: When facial feature analysis shows an age greater than or equal to a preset age threshold (e.g., 60 years old), or when voiceprint recognition matches pre-stored elderly user information, the user can be identified as elderly. Elderly users may experience hearing or vision loss or slow reaction time, requiring clear and prominent warnings, but excessive fright should be avoided.

[0063] Ordinary adult users: If the above criteria for determining child or elderly users are not met, they can be identified as ordinary adult users. For these users, standard, tiered early warning and intervention strategies can be adopted.

[0064] User type identification can be done in real time, with a judgment made before each door is opened to adapt to different scenarios of passengers getting on and off the vehicle.

[0065] S142. Based on the user type and risk information, select the corresponding early warning method from the preset early warning method library, and select the corresponding intervention strategy from the preset intervention strategy library.

[0066] The warning method library and intervention strategy library are pre-configured rule bases containing a variety of options. The options in the library are associated with different user types and risk levels (or other parameters in the risk information, such as collision time).

[0067] Warning devices may include at least one of the following: audible warning modules (such as in-vehicle speakers), light warning modules (such as door ambient lights and dashboard indicator lights), or tactile warning modules (such as seat vibration, steering wheel vibration, and door handle vibration). The warning method library defines how to combine and configure these warning modules under different combinations of user types and risk levels.

[0068] For example, for child users, regardless of the risk level, a high-intensity warning method is selected. The sound warning uses a high-volume, rapid alarm sound. The light warning uses a rapidly flashing red light. The tactile warning uses continuous high-frequency vibration. The aim is to ensure that children's attention is strongly drawn.

[0069] For elderly users, choose a medium-intensity alert mode. Audible alerts use clear, slightly slow voice prompts. Light alerts use a steady yellow or orange light or slow flashing. Tactile alerts use gentle pulsed vibrations. The aim is to provide clear cues without being overly disruptive.

[0070] For ordinary adult users, select the appropriate alert level based on the risk level: Low risk: No warning or slight alert sound, green light.

[0071] Medium risk: A medium-volume warning sound is emitted, and the light flashes yellow.

[0072] High risk: Emits a high-volume alarm sound, displays rapidly flashing red lights, and may be accompanied by tactile vibrations.

[0073] The intervention strategy library defines how to control the opening of car doors under different combinations of user types and risk levels. Controlling door opening includes at least one of the following: controlling the door opening speed, controlling the door opening angle, or controlling the door's movement state. The door's movement state includes normal opening, temporary locking, etc.

[0074] For example, for child users, regardless of the risk level, a high-intensity intervention strategy is chosen. In high-risk situations, the doors are locked directly, preventing them from opening. In medium-risk situations, the opening angle is limited to a very small safe range. Even in low-risk situations, a slow opening method is used.

[0075] For elderly users, choose a moderate-intensity intervention strategy. Lock the doors when there is high risk. Limit the opening angle when there is medium risk. Allow normal opening or slight, slow opening when there is low risk.

[0076] For ordinary adult users, the appropriate intervention strategy should be selected based on the risk level: Low risk: Allows the doors to open at normal speed and angle.

[0077] Medium risk: Control the car door to open slowly, with the opening speed limited to a preset speed range (such as 50% of the normal speed) and the opening angle limited to a preset angle range (such as 60% of the maximum opening angle).

[0078] High risk: Sends a locking command to the door motor controller to temporarily lock the door for a duration within a preset range (e.g., until the risk is eliminated or the user forcibly unlocks it).

[0079] Through step S140, the system realizes the transformation from "unified strategy" to "personalized strategy", making security protection more precise and humanized.

[0080] In some embodiments of this application, high-risk interventions may also employ a "slow reverse closing" strategy (e.g., slowly closing the door after it has been opened to 5cm) to replace temporary locking and improve safety in emergency situations. The elderly mode may add a "step-by-step opening" strategy (e.g., opening 10cm first, pausing for 1 second, and then fully opening) to balance safety and convenience.

[0081] S150, The warning device that controls the vehicle based on the warning method issues a warning, and controls the opening of the vehicle door based on the intervention strategy.

[0082] For warnings, the control device sends corresponding control commands to the warning device. For example, it controls the sound warning module to play a specified audio at a specified volume and frequency. It controls the light warning module to light up with a specified color and flashing frequency. It controls the tactile warning module to start vibration at a specified frequency and amplitude.

[0083] For door control, the control device sends control commands to the motor controller. The motor controller adjusts the output torque or current of the motor that drives the door according to the commands, thereby controlling the door opening speed (e.g., reducing torque to decelerate), controlling the door opening angle (e.g., controlling the stop position through limit switches or motor position feedback), or controlling the door movement state (e.g., outputting reverse torque to lock).

[0084] While implementing intervention strategies, the system continuously monitors the vehicle's surroundings. When the risk level is determined to be low or the predicted collision time exceeds a safety threshold, the system can automatically release the locks or angle restrictions, restoring normal opening functionality. Simultaneously, the system also provides a mechanism for manual forced unlocking, such as pressing and holding the door unlock button to allow the user to unlock the door while issuing the strongest warning, in response to emergencies.

[0085] Through step S150, personalized warning and intervention strategies are implemented, ultimately ensuring the safety of the door opening process.

[0086] In one possible implementation, this application also includes an adaptive learning optimization function. By recording user actions (such as the number of times the lock is forcibly released, and the operation of adjusting the warning volume) and actual risk scenario feedback data (such as whether close-range avoidance occurs, and user complaints), a reinforcement learning algorithm is used to continuously optimize the AI ​​risk assessment threshold, warning parameters, and intervention strategies. The state space of the reinforcement learning algorithm includes environmental feature vectors, vehicle state data, and user feature data; the action space includes warning method selection and intervention strategy selection; and the reward function is based on user satisfaction feedback and actual safety effects. When the vehicle is stationary and connected to the network, the collected data is automatically uploaded to the cloud. After iterative training of the cloud model, an update package is pushed to the onboard AI chip, enabling online model upgrades. This adaptive learning optimization mechanism allows the system to continuously learn and adapt to the usage habits of different users and the risk characteristics of different scenarios, continuously improving the accuracy and adaptability of warnings and interventions, and extending the system's lifespan.

[0087] In one possible implementation, some AI models (such as data preprocessing and simple risk assessment) can be deployed on the edge sensor chip to achieve a deployment method of "edge computing and cloud collaboration", which reduces data transmission latency and is suitable for scenarios with poor network signals.

[0088] The foregoing mainly describes the solutions of the embodiments of this application from a methodological perspective. It is understood that, in order to achieve the above functions, the control device includes at least one of the hardware structures and software modules corresponding to the execution of each function. Those skilled in the art should readily recognize that, in conjunction with the units and algorithm steps of the various examples described in the embodiments applied herein, the embodiments of this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed by hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the embodiments of this application.

[0089] This application embodiment can divide the control device into functional modules according to the above method embodiment. For example, each function can be divided into a separate functional module, or two or more functions can be integrated into one functional module. The integrated module can be implemented in hardware or software. It should be noted that the module division in this application embodiment is illustrative and only represents one logical functional division. In actual implementation, there may be other division methods. The following description uses the example of dividing each functional module according to each function.

[0090] Figure 5This is a schematic diagram of a control device 500 provided in an embodiment of this application. The control device 500 is applied to a vehicle and can execute the door opening control method provided in the above-described method embodiment. Figure 5 As shown, the control device 500 includes: an acquisition module 501, a fusion module 502, a first determination module 503, a second determination module 504, and a control module 505.

[0091] The acquisition module 501 is used to acquire vehicle surrounding environment data, vehicle status data, and user characteristic data; The fusion module 502 is used to perform multimodal feature fusion on the vehicle's surrounding environment data to generate an environmental feature vector; The first determining module 503 is used to determine the risk information of the door opening based on environmental feature vectors and vehicle status data; The second determination module 504 is used to determine the early warning method and intervention strategy based on risk information and user characteristic data; The control module 505 is used to control the vehicle's warning device to issue warnings based on the warning method and to control the opening of the vehicle doors based on the intervention strategy.

[0092] In one possible implementation, the first determining module 503 may be specifically used to: perform predictive analysis on environmental feature vectors and vehicle state data to determine the risk level and / or the trajectory and collision time of dynamic obstacles, where the risk level includes low risk, medium risk and high risk; and generate risk information based on the risk level and / or the trajectory and collision time of dynamic obstacles.

[0093] In one possible implementation, the second determining module 504 may be specifically used to: determine the user type based on user characteristic data, where the user type includes at least one of children, the elderly, and ordinary adults; select the corresponding early warning method from a preset early warning method library and the corresponding intervention strategy from a preset intervention strategy library according to the user type and risk information.

[0094] In one possible implementation, the acquisition module 501 may be specifically used to: collect environmental data around the vehicle through multiple types of sensors, including at least one of millimeter-wave radar, high-definition camera, or ultrasonic sensor; acquire vehicle status data through vehicle status sensors, including vehicle speed, parking posture, and door lock status; and acquire user feature data through user feature recognition sensors, including at least one of seat weight sensor, in-vehicle camera, or voiceprint recognition module.

[0095] In one possible implementation, the fusion module 502 can be specifically used to: preprocess the collected vehicle surrounding environment data, including noise reduction and data alignment; and perform cross-modal feature extraction and fusion on the preprocessed vehicle surrounding environment data to generate an environmental feature vector.

[0096] In one possible implementation, the warning device includes at least one of an audible warning module, a light warning module, or a tactile warning module for issuing a warning; controlling the opening of the vehicle door includes at least one of controlling the opening speed of the vehicle door, controlling the opening angle of the vehicle door, or controlling the movement state of the vehicle door.

[0097] The control device 500 described above can be implemented in hardware or as a software functional module. When implemented as a software functional module, the control device 500 may include a processor and a memory. The processor is coupled to the memory, and the memory is used to store instructions executable by the processor. When the processor executes the instructions, it can execute the door opening control method of Embodiment 1 described above.

[0098] In implementing the functions of the integrated modules described above using hardware, this application provides a possible structure for the control device 600 involved in the above embodiments. For example... Figure 6 As shown, the control device 600 includes a processor 602 and a bus 604. Optionally, the control device 600 may also include a memory 601; alternatively, the control device 600 may also include a communication interface 603.

[0099] Processor 602 may implement or execute various exemplary logic blocks, modules, and circuits described in conjunction with embodiments of this application. Processor 602 may be a central processing unit, a general-purpose processor, a digital signal processor, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It may implement or execute various exemplary logic blocks, modules, and circuits described in conjunction with embodiments of this application. Processor 602 may also be a combination that implements computing functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0100] Communication interface 603 is used to connect to other devices via a communication network. This communication network can be Ethernet, wireless access network, wireless local area network (WLAN), etc.

[0101] The memory 601 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), disk storage medium or other magnetic storage device, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but is not limited thereto.

[0102] In one possible implementation, the memory 601 can exist independently of the processor 602. The memory 601 can be connected to the processor 602 via a bus 604 and is used to store instructions or program code. When the processor 602 calls and executes the instructions or program code stored in the memory 601, it can implement the door opening control method provided in this embodiment. In another possible implementation, the memory 601 can also be integrated with the processor 602.

[0103] Bus 604 can be an extended industry standard architecture (EISA) bus, etc. Bus 604 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 6 The 604 bus is represented by a single thick line, but this does not mean that there is only one bus 604 or only one type of bus 604.

[0104] Some embodiments of this application provide a computer-readable storage medium (e.g., a non-transitory computer-readable storage medium) storing at least one piece of program code, which, when loaded and executed by a processor, causes a computer to perform a door opening control method as described in any of the above embodiments.

[0105] For example, the aforementioned computer-readable storage media may include, but are not limited to: magnetic storage devices (e.g., hard disks, floppy disks, or magnetic tapes), optical discs (e.g., compact disks (CDs), digital versatile disks (DVDs), etc.), smart cards, and flash memory devices (e.g., erasable programmable read-only memory (EPROMs), cards, sticks, or key drives, etc.). The various computer-readable storage media described in this application may represent one or more devices and / or other machine-readable storage media for storing information. The term "machine-readable storage medium" may include, but is not limited to, wireless channels and various other media capable of storing, containing, and / or carrying instructions and / or data.

[0106] This application provides a computer program product containing instructions that, when run on a computer, cause the computer to execute the door opening control method of any of the above embodiments.

[0107] This application provides a vehicle, including: a warning device for issuing a warning; a motor controller for controlling the opening of the vehicle door; and a control device as described in any of the preceding embodiments; or a computer-readable storage medium as described in any of the preceding embodiments; or a computer program product as described in any of the preceding embodiments.

[0108] The vehicle provided in this application may be a passenger vehicle or a freight vehicle, and may also be an electric vehicle or a hybrid vehicle. This application does not limit the specific purpose or power type of the vehicle, and the choice can be made according to actual needs.

[0109] A vehicle consists of a body and wheels. The body is used for passengers and for carrying goods, while the wheels are mounted underneath the body to support it and allow it to roll on the road, thus enabling the vehicle to move.

[0110] In some possible examples, the vehicle is equipped with a control system, which typically adopts a layered distributed architecture. From the bottom layer to the top layer, it can be roughly divided into a perception layer, a control layer, a coordination layer, and an interaction layer. The layers communicate with each other through an in-vehicle network.

[0111] The perception layer mainly consists of various sensors distributed inside and outside the vehicle, including but not limited to external environment cameras, millimeter-wave radar, lidar, ultrasonic sensors, in-vehicle driver monitoring cameras, microphone arrays, and various vehicle status sensors (such as wheel speed sensors, inertial measurement units, temperature sensors, etc.). The perception layer is responsible for collecting multi-dimensional data such as the vehicle's own operating status, driver behavior, and external driving environment in real time.

[0112] The control layer consists of dozens to hundreds of electronic control units (ECUs), distributed across multiple functional domains including powertrain, chassis, body, intelligent driving, and infotainment. Each ECU embeds real-time control software that performs closed-loop control of the vehicle's actuators based on preset control strategies or upper-level commands, and generates corresponding alarm signals when abnormal conditions are detected. Typical ECUs include the engine control unit, transmission control unit, brake control unit, steering control unit, vehicle stability control unit, airbag control unit, intelligent driving domain controller, and in-vehicle infotainment unit.

[0113] The coordination layer typically exists in the form of a domain controller or a central computing platform, responsible for cross-domain data fusion, global state management, and collaborative decision-making. The coordination layer centrally processes and schedules the sensing data and control commands that were originally scattered across various functional domains, connecting downwards to various electronic control units and supporting human-machine interaction functions upwards.

[0114] The interaction layer mainly includes in-cabin display devices (such as instrument panel, central control screen, head-up display), voice interaction system, haptic feedback device, etc., which are responsible for presenting vehicle status, warning information and driving suggestions to the driver in the form of visual, auditory or tactile, while receiving the driver's touch, voice and other input commands.

[0115] Data transmission and interaction between different layers are achieved through the vehicle bus network. Common vehicle bus protocols include CAN (Controller Area Network), CAN FD (CAN with Flexible Data-Rate), LIN (Local Interconnect Network), FlexRay, and vehicle Ethernet, which supports high-bandwidth data transmission. Among these, CAN and CAN FD buses are widely used for communication in real-time control domains such as powertrain and chassis, while vehicle Ethernet is gradually being applied to high-bandwidth sensor data transmission in the intelligent driving domain and multimedia interaction scenarios in the cockpit domain.

[0116] In some possible examples, the domain controller, as the core of the in-vehicle computing platform, is configured with a processor and memory to execute the door opening control method provided in the embodiments of this application.

[0117] For example, the control device of this application is arranged in a domain controller.

[0118] For example, a domain controller may have multiple functional modules deployed to perform the various method steps.

[0119] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for controlling the opening of a vehicle door, characterized in that, For use in a vehicle, the vehicle including a warning device and a door, the control method includes: Acquire data on the vehicle's surrounding environment, vehicle status, and user characteristics. Multimodal feature fusion is performed on the environmental data surrounding the vehicle to generate an environmental feature vector; Based on the environmental feature vector and the vehicle status data, determine the risk information of the door opening; The early warning method and intervention strategy are determined based on the risk information and user characteristic data; The warning device of the vehicle is controlled to issue a warning based on the warning method, and the opening of the vehicle door is controlled based on the intervention strategy.

2. The door opening control method according to claim 1, characterized in that, The process of determining door opening risk information based on the environmental feature vector and the vehicle status data includes: Predictive analysis is performed on the environmental feature vector and the vehicle state data to determine the risk level and / or the trajectory and collision time of dynamic obstacles, wherein the risk level includes low risk, medium risk and high risk. Risk information is generated based on the risk level and / or the trajectory and collision time of the dynamic obstacle.

3. The door opening control method according to claim 1, characterized in that, The step of determining the early warning method and intervention strategy based on the risk information and the user characteristic data includes: The user type is determined based on the user characteristic data, and the user type includes at least one of child, elderly, and ordinary adult; Based on the user type and the risk information, a corresponding early warning method is selected from a preset early warning method library, and a corresponding intervention strategy is selected from a preset intervention strategy library.

4. The door opening control method according to claim 1, characterized in that, The acquisition of vehicle surrounding environment data, vehicle status data, and user characteristic data includes: The vehicle's surrounding environment data is collected by multiple types of sensors, including at least one of millimeter-wave radar, high-definition camera, or ultrasonic sensor. Vehicle status data is acquired through vehicle status sensors, including vehicle speed, parking posture, and door lock status. User feature data is acquired through a user feature recognition sensor, which includes at least one of a seat weight sensor, an in-vehicle camera, or a voiceprint recognition module.

5. The door opening control method according to claim 1, characterized in that, The step of performing multimodal feature fusion on the environmental data surrounding the vehicle to generate an environmental feature vector includes: The collected data on the vehicle's surrounding environment is preprocessed, including noise reduction and data alignment. Cross-modal feature extraction and fusion are performed on the preprocessed vehicle surrounding environment data to generate the environmental feature vector.

6. The door opening control method according to any one of claims 1 to 5, characterized in that, The warning device includes at least one of a sound warning module, a light warning module, or a tactile warning module for issuing warnings. Controlling the opening of the vehicle door includes at least one of controlling the opening speed of the vehicle door, controlling the opening angle of the vehicle door, or controlling the movement state of the vehicle door.

7. A control device, characterized in that, include: The acquisition module is used to acquire data on the vehicle's surrounding environment, vehicle status data, and user characteristic data. The fusion module is used to perform multimodal feature fusion on the environmental data surrounding the vehicle to generate an environmental feature vector; The first determining module is used to determine the risk information of the door opening based on the environmental feature vector and the vehicle state data; The second determining module is used to determine the early warning method and intervention strategy based on the risk information and the user characteristic data; The control module is used to control the vehicle's warning device to issue a warning based on the warning method, and to control the opening of the vehicle door based on the intervention strategy.

8. A control device, characterized in that, include: Memory and processor; Memory and processor are coupled; The memory is used to store instructions that can be executed by the processor; When the processor executes the instruction, it performs the door opening control method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one piece of program code, which is loaded and executed by a processor to implement the door opening control method as described in any one of claims 1 to 6.

10. A computer program product, characterized in that, It includes a computer program; when the computer program is executed, it is able to implement the door opening control method as described in any one of claims 1 to 6.

11. A vehicle, characterized in that, include: Early warning device, the early warning device being used for early warning; A motor controller, used to control the opening of the vehicle door; And, the control device as described in claim 8; Alternatively, the computer-readable storage medium as claimed in claim 9; or the computer program product as claimed in claim 10.

Images