A vehicle door control method and vehicle

Abstract

This application provides a door control method and a vehicle, relating to the field of vehicle control technology. The method includes: determining user intent information and the target scenario type corresponding to the vehicle's current scenario based on vehicle-collected data; determining safety constraint information during the door control process based on the target scenario type; using a constraint satisfaction model, determining a door control strategy that satisfies the safety constraint information based on the user intent information; and controlling at least one target door to open and close according to the door control strategy. The embodiments of this application can solve the problem in the prior art that it is impossible to ensure safety during the door control process while meeting user needs.

Description

Technical Field

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

[0002] As vehicles become more intelligent, the control of opening and closing doors is gradually evolving from simple manual or remote control operations towards automated control.

[0003] Based on this, traditional door control methods predefine multiple scenario modes, each with a corresponding fixed door control strategy. When the current vehicle scenario matches a target scenario mode, the door is controlled according to the fixed strategy corresponding to that target scenario mode.

[0004] However, in reality, the actual scenarios in which vehicles operate are usually complex and varied. The static rule mapping scheme based on pattern matching mentioned above suffers from rigid control and inflexible strategies when faced with complex and varied actual vehicle scenarios, thus failing to meet user needs while ensuring safety during the door control process. Summary of the Invention

[0005] In view of this, the embodiments of this application aim to provide a door control method and vehicle to solve the problem that the prior art cannot guarantee the safety of the door control process while meeting user needs.

[0006] In a first aspect, one embodiment of this application provides a vehicle door control method, comprising: determining user intent information and target scene type corresponding to the vehicle's current scene based on vehicle collected data; determining safety constraint information in the vehicle door control process according to the target scene type; determining a vehicle door control strategy that satisfies the safety constraint information based on the user intent information using a constraint satisfaction model; and controlling at least one target vehicle door to perform opening and closing actions according to the vehicle door control strategy.

[0007] In this application, by determining the safety constraint information of the door control process in the target scenario corresponding to the vehicle's location, the safety constraint information changes dynamically according to the changes in the vehicle's location. Based on this, a constraint satisfaction model can be used to dynamically determine a door control strategy that can satisfy the dynamically changing safety constraint information based on user intent information. This allows the door control strategy to strictly follow the safety constraints of the door control process in the vehicle's location scenario while maximizing the satisfaction of user intent. By controlling the door opening and closing actions according to this dynamically determined door control strategy, the door control can adapt to complex and ever-changing actual scenarios, thereby ensuring the safety of the door control process while meeting user needs.

[0008] In conjunction with the first aspect, in some implementations of the first aspect, a constraint satisfaction model is used to determine a door control strategy that satisfies safety constraints based on user intent information. This includes: constructing constraint parameters based on safety constraints and user intent information, wherein the constraint parameters include at least one of spatial constraint parameters, interlock constraint parameters, center of gravity constraint parameters, and door control timing constraint parameters; constructing a constraint satisfaction model based on user intent information and using the constraint parameters as constraint conditions; and solving the constraint satisfaction model to obtain the door control strategy.

[0009] In this way, by defining multi-dimensional constraint parameters such as space, interlocking, center of gravity and timing based on scenario-related safety constraint information and user intent information related to user intent, a constraint satisfaction model can be constructed. This model can comprehensively and systematically express the physical limitations and safety requirements of door opening and closing control under different scenarios and user intents, thereby obtaining a more reasonable and safer door control strategy.

[0010] In conjunction with the first aspect, in certain implementations of the first aspect, solving the constraint satisfaction model to obtain the door control strategy includes: using a backtracking algorithm to solve the constraint satisfaction model and generating a multi-door cooperative action sequence, wherein the multi-door cooperative action sequence includes multiple target doors to be controlled, the door opening angle corresponding to the target doors, and the action time; if it is determined from the multi-door cooperative action sequence that at least two of the multiple target doors need to be opened and closed simultaneously, then it is determined whether there is a target risk when at least two target doors perform opening and closing actions simultaneously, wherein the target risk includes inter-door interference risk and / or vehicle center of gravity offset risk; if there is a target risk, then the door opening angle and / or action time corresponding to at least two target doors are adjusted respectively to eliminate the target risk and obtain the door control strategy.

[0011] In this way, by actively detecting the risk of inter-door interference and center of gravity shift between target doors that move simultaneously after generating a multi-door coordinated action sequence, and dynamically adjusting the door opening angle or door control timing to eliminate the risk, the safety hazards of door collisions or vehicle instability caused by the simultaneous movement of multiple doors are avoided, further improving the safety of the door control process.

[0012] In conjunction with the first aspect, in some implementations of the first aspect, user intent information is determined based on vehicle-collected data, including: determining user movement time-series characteristics based on user data collected from vehicles; and using a user intent prediction model to predict user intent based on user movement time-series characteristics to obtain user intent information.

[0013] In this way, by extracting the user's motion timing features and using a user intent prediction model for intent recognition, the user's operational intent can be proactively predicted from the user data collected from the vehicle. This eliminates the need for users to manually trigger door opening and closing operations one by one, providing accurate intent input for subsequent proactive collaborative control, thereby improving response speed and ease of operation. In addition, using a user intent prediction model for intent recognition also improves the accuracy of user intent prediction.

[0014] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: if a fault is detected in the target sensor, acquiring door trigger information input by the user, wherein the target sensor is a sensor in the vehicle used to collect user data; and determining user intent information based on the door trigger information.

[0015] In this way, when the target sensor used to collect user data fails, the user intent prediction is downgraded to obtaining the user's direct trigger information as the source of user intent information. This ensures that the user intent information can still be determined normally even when the target sensor fails, avoids the complete loss of the automatic door control function, and improves the robustness and reliability of the system.

[0016] In conjunction with the first aspect, in some implementations of the first aspect, determining the target scene type corresponding to the scene in which the vehicle is located based on vehicle-collected data includes: using a scene classification model to identify the type of the scene in which the vehicle is located based on vehicle-collected data, and obtaining a scene type identification result, wherein the scene type identification result includes at least two scene types corresponding to the scene in which the vehicle is located and the confidence levels corresponding to the at least two scene types; if the highest confidence level in the scene type identification result is greater than or equal to a preset threshold, then the scene type corresponding to the highest confidence level is determined as the target scene type; the method further includes: if the highest confidence level in the scene type identification result is less than a preset threshold, then controlling at least one target door to open or close according to a preset conservative control strategy.

[0017] In this way, by setting a confidence threshold, the scenario type with the highest confidence in the scenario classification results is used as the target scenario type when the scenario classification results are reliable, and the conservative control strategy is used when the classification results are unreliable. This avoids improper opening and closing of the car door due to inaccurate judgment of the scenario type, and improves the safety and reliability of the scenario-adaptive door control process.

[0018] In conjunction with the first aspect, in some implementations of the first aspect, during the process of controlling at least one target door to perform opening and closing actions according to the door control strategy, the method further includes: adaptively adjusting the door action parameters of at least one target door according to the target scene type; and adjusting the opening and closing actions of at least one target door based on the door action parameters.

[0019] In this way, by dynamically linking scene types with door action parameters, the door control strategy can adjust the target door's action parameters in real time according to the scene during operation. This allows for dynamic adjustment of the target door's opening and closing actions in various scenes, improving the physical adaptability and safety of the door control process to diverse environments.

[0020] In conjunction with the first aspect, in some implementations of the first aspect, during the process of controlling at least one target door to perform opening and closing actions according to the door control strategy, the method further includes: if a target anomaly is detected, determining a corresponding anomaly handling strategy based on the frequency of occurrence of the target anomaly, wherein the target anomaly includes at least one of obstacle abrupt change anomaly, door motor anomaly, and vehicle posture anomaly; and performing anomaly handling operations on the door according to the anomaly handling strategy.

[0021] In this way, by adopting a graded anomaly handling strategy based on the frequency of occurrence of target anomalies, the availability of door functions can be maintained as much as possible while avoiding danger. This avoids the inconvenience caused by overly conservative door control methods or the danger caused by aggressive door control methods, thereby improving the intelligence of anomaly handling and system resilience.

[0022] In conjunction with the first aspect, in some implementations of the first aspect, during the process of controlling at least one target door to perform opening and closing actions according to the door control strategy, the method further includes: detecting the vehicle's battery parameters; if the battery parameters meet the abnormal parameter conditions, controlling at least one target door to stop the opening and closing actions, and switching the door control mode to manual mode.

[0023] In this way, by monitoring battery parameters in real time and actively stopping the automatic control of the doors and switching to manual mode when abnormal parameters are met, adverse consequences such as insufficient driving force, loss of control of movement or door lock failure caused by abnormal power supply are avoided, thus ensuring the safety of the vehicle and personnel in the event of abnormal power supply.

[0024] Secondly, one embodiment of this application provides a vehicle door control device, including: an information determination module, used to determine user intent information and the target scene type corresponding to the vehicle's location based on vehicle collected data; a constraint determination module, used to determine safety constraint information during the vehicle door control process according to the target scene type; a strategy determination module, used to determine a vehicle door control strategy that satisfies the safety constraint information based on the user intent information using a constraint satisfaction model; and an action control module, used to control at least one target vehicle door to perform opening and closing actions according to the vehicle door control strategy.

[0025] Thirdly, one embodiment of this application provides a computer-readable storage medium storing a computer program for performing the door control method described in the first aspect.

[0026] Fourthly, one embodiment of this application provides an electronic device, which includes: a processor; a memory for storing processor-executable instructions; the processor being used to execute the door control method described in the first aspect.

[0027] Fifthly, one embodiment of this application provides a computer program product including instructions that, when executed on an electronic device, cause the electronic device to implement the door control method described in the first aspect. Attached Figure Description

[0028] The above and other objects, features, and advantages of this application will become more apparent from the more detailed description of the embodiments of this application in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the embodiments of this application to explain this application and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0029] Figure 1 The diagram shown is a schematic flowchart of a door control method provided in an embodiment of this application.

[0030] Figure 2 The diagram shown is a structural schematic of a door control system provided in an embodiment of this application.

[0031] Figure 3 The diagram shown is a structural schematic of a door control device provided in an embodiment of this application.

[0032] Figure 4 The diagram shown is a structural schematic of an electronic device provided in an embodiment of this application. Detailed Implementation

[0033] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, 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.

[0034] Furthermore, to better illustrate this application, numerous specific details are provided in the following detailed embodiments. Those skilled in the art should understand that this application can be implemented even without certain specific details. In some instances, methods and means well-known to those skilled in the art have not been described in detail in order to highlight the main points of this application.

[0035] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0036] Furthermore, the terms “first,” “second,” “third,” and “fourth” are used only for distinguishing descriptions and should not be interpreted as indicating or implying relative importance.

[0037] As the level of vehicle intelligence continues to improve, the capabilities of automobiles in perception, decision-making and execution have been significantly enhanced. The control methods for opening or closing car doors are also undergoing profound changes—from the initial stage of relying on manual operation by users or simple remote control triggering using keys, mobile phones, etc., to a new stage of highly automated intelligent control.

[0038] Based on the aforementioned development trends, traditional vehicle door control methods typically employ a static rule mapping scheme based on pattern matching. This involves pre-setting multiple scenario modes (such as "carrying mode," "rainy day mode," and "child safety mode") for several typical usage scenarios, with each scenario mode corresponding to a fixed set of door control strategies. When the vehicle detects that the current environment or user behavior matches a preset scenario mode during operation, the system directly invokes the fixed strategy bound to that mode to execute the door control operation.

[0039] However, in real-world driving and vehicle usage environments, the actual scenarios faced by vehicles are often highly complex and diverse—for example, users carrying large items getting in and out of the vehicle on a slope, multiple passengers entering and exiting simultaneously in narrow parking spaces, or users picking up and dropping off passengers in congested traffic. In this context, the traditional control schemes relying on static rule mapping and limited pattern matching, lacking dynamic perception of scenario differences and flexible response mechanisms, are prone to exhibiting rigid control logic and inflexible response strategies. This rigid control method not only struggles to accurately meet the user's real-time needs but may also lead to safety hazards in specific situations, such as rapidly opening doors in congested traffic without adequately assessing surrounding obstacles, or simultaneously opening sliding doors and the tailgate on a slope without fully considering whether the vehicle's center of gravity has shifted. Therefore, traditional door control methods cannot guarantee safety during the door control process while meeting user needs.

[0040] To address the aforementioned technical problems, this application provides a door control method and a vehicle. The door control method includes: determining user intent information and the target scenario type corresponding to the vehicle's current scenario based on vehicle-collected data; determining safety constraint information during the door control process based on the target scenario type; determining a door control strategy that satisfies the safety constraint information based on the user intent information using a constraint satisfaction model; and controlling at least one target door to open and close according to the door control strategy. In this way, by determining the safety constraint information during the door control process in the current scenario based on the target scenario type corresponding to the vehicle's current scenario, and by dynamically changing the safety constraint information according to the changes in the vehicle's current scenario, the constraint satisfaction model can dynamically determine a door control strategy that satisfies the dynamically changing safety constraint information based on the user intent information. This ensures that the door control strategy strictly adheres to the safety constraints during the door control process in the current scenario while maximizing the satisfaction of the user's intent. By controlling the door opening and closing actions according to this dynamically determined door control strategy, the door control can adapt to complex and ever-changing real-world scenarios, thereby achieving both user needs and ensuring safety during the door control process.

[0041] The following is combined Figures 1 to 2 The door control method provided in this application is described in detail.

[0042] Figure 1 The diagram shown is a schematic flowchart of a door control method provided in an embodiment of this application. This method can be applied to vehicles. Figure 1 As shown, the method may include the following steps.

[0043] S110 determines the user intent information and the target scenario type corresponding to the vehicle's location based on the vehicle's collected data.

[0044] In some examples, the vehicle data collected can be raw or processed data collected by at least one sensor on the vehicle. This vehicle data may include, for example, at least one of the following: Passive Keyless Entry (PKE) identification signals, surround-view camera image data, ultrasonic radar distance data, gyroscope attitude data, wheel speed sensor data, etc.

[0045] For example, after the system in the vehicle is started, each sensor continuously collects data. For instance, the PKE module can collect data such as user ID, Received Signal Strength Indicator (RSSI), and Angle of Arrival (AOA) to calculate the user's position coordinates relative to the vehicle. Another example is the surround-view camera, which can collect 360-degree images of the vehicle's surroundings and use semantic segmentation algorithms to extract obstacle types and locations. Yet another example is the ultrasonic radar, which can collect distance data to nearby obstacles (e.g., 0-200cm) around the vehicle, with a sampling frequency of 10Hz. A gyroscope can also collect data such as the vehicle's pitch angle (for calculating gradient), roll angle (for calculating roll angle), and yaw angle, with a sampling frequency of 20Hz. Finally, wheel speed sensors can collect the rotational speed of each wheel to calculate the vehicle's longitudinal speed.

[0046] Furthermore, if the vehicle data collected consists of multi-source heterogeneous data from multiple sensors, a Kalman filter algorithm can be used to perform spatiotemporal alignment and fusion processing on the multi-source data to generate a unified environmental state vector. This vector can then be used to determine the user's intent information and the target scene type corresponding to the vehicle's location. Specifically, the generated environmental state vector could be, for example, State = [x_user, y_user, θ_pitch, θ_roll, v_vehicle, {obs_i}, {door_j}], where x_user and y_user are the user's position coordinates, θ_pitch is the pitch angle, θ_roll is the roll angle, v_vehicle is the vehicle speed, {obs_i} is the set of obstacles, and {door_j} is the state of each door.

[0047] In some examples, user intent information can be information that represents the user's intention to open or close the door. Specifically, the user's intention to open or close the door can be divided into multiple intent types. The user intent information can be the intent type to which the user's intention to open or close the door belongs in the current scenario.

[0048] For example, the vehicle data may include user-related data, such as data collected by the PKE module and / or surround view camera. Based on this user-related data, the intent type of the user's door opening and closing intention in the current scenario can be identified as the user intent information.

[0049] To improve the accuracy of intent recognition, neural network models can be used to predict user intent. Based on this, in some embodiments, determining user intent information based on vehicle-collected data in S110 includes: determining user movement temporal characteristics based on user data collected from the vehicle; and using a user intent prediction model to predict user intent based on these movement temporal characteristics to obtain user intent information.

[0050] In some examples, the user data collected by the vehicle could be signal timing data collected by the PKE module and personnel trajectory tracking data collected by the camera. The user intent prediction model could be a trained neural network model capable of predicting user intent, such as a Long Short-Term Memory (LSTM) network model.

[0051] For example, user motion time-series features can be constructed based on the signal timing data collected by the PKE module within the last 30 seconds and the personnel trajectory tracking data collected by the camera within the last 10 seconds. These user motion time-series features can be represented in vector form, for example, Time_Feature = [loc_t-30, ..., loc_t, traj_t-10, ..., traj_t, door_state], where loc_t-30, ..., loc_t are the signal timing data collected by the PKE module within the last 30 seconds before time t, traj_t-10, ..., traj_t are the personnel trajectory tracking data collected by the camera within the last 10 seconds before time t, and door_state is the state of each door.

[0052] The user's movement time-series features are input into an LSTM model, which then predicts the user's intent and outputs the intent information. For example, if the LSTM model detects that a user is in the tailgate area and carrying items based on the user's movement time-series features, it can output intent type 1: Open the tailgate to load luggage. If the LSTM model detects that a user is in the sliding door area and multiple people are detected based on the user's movement time-series features, it can output intent type 2: Open the sliding door to pick up or drop off passengers. If the LSTM model detects that a user is moving from the sliding door area to the driver's seat based on the user's movement time-series features, it can output intent type 3: Enter the driver's seat. If the LSTM model detects that multiple users are in different door areas based on the user's movement time-series features, it can output intent type 4: Multiple people cooperating. If the LSTM model detects that a user is leaving the vehicle area based on the user's movement time-series features, it can output intent type 5: Close all doors and leave the vehicle.

[0053] In this way, by extracting the user's motion timing features and using a user intent prediction model for intent recognition, the user's operational intent can be proactively predicted from the user data collected from the vehicle. This eliminates the need for users to manually trigger door opening and closing operations one by one, providing accurate intent input for subsequent proactive collaborative control, thereby improving response speed and ease of operation. In addition, using a user intent prediction model for intent recognition also improves the accuracy of user intent prediction.

[0054] In addition, in some embodiments, the door control method may further include: if a fault is detected in the target sensor, acquiring door trigger information input by the user, wherein the target sensor is a sensor in the vehicle used to collect user data; and determining user intent information based on the door trigger information.

[0055] In some examples, the target sensor can be a sensor in the vehicle used to collect user data, such as a PKE module or a camera. Situations where the target sensor malfunctions include, but are not limited to: receiving data indicating an abnormality, the target sensor collecting abnormal data, or not receiving data from the target sensor within a preset time period. Additionally, door trigger information can be signals or information that the user manually triggers the door opening and closing control, including but not limited to PKE unlocking triggers, pressing body buttons, and remote control commands.

[0056] For example, if the data collected by the target sensor is lost for more than a preset time, it is determined that the target sensor is faulty. In this case, only the data collected by other sensors can be used, and the user intent prediction function is disabled. That is, the model no longer predicts the user intent, but instead directly reads the door trigger information manually entered by the user. The corresponding control operation information for the door triggered by the user to open or close is used as the user intent information. For example, if the data collected by the camera is lost for more than 500ms, the user intent prediction function is disabled, and the trigger signal generated when the user performs PKE unlocking is read. Based on the trigger signal, "both front doors open" is determined as the user intent information.

[0057] In this way, when the target sensor used to collect user data fails, the user intent prediction is downgraded to obtaining the user's direct trigger information as the source of user intent information. This ensures that the user intent information can still be determined normally even when the target sensor fails, avoids the complete loss of the automatic door control function, and improves the robustness and reliability of the system.

[0058] In other examples, the target scene type can be the scene type corresponding to the current location of the vehicle. Multiple scene types can be preset in advance, including but not limited to: flat open ground, uphill slope, downhill slope, narrow parking space, crosswind on a slope, and congested road section.

[0059] For example, the vehicle data may also include data related to the in-vehicle and external environment, such as data collected by surround view cameras, ultrasonic radar, gyroscopes, and / or wheel speed sensors. Based on this environment-related data, the scene type to which the current vehicle is located can be identified as the target scene type.

[0060] To improve the accuracy of scene type identification, artificial intelligence models can be used for scene type identification. In some embodiments, determining the target scene type corresponding to the scene where the vehicle is located based on vehicle-collected data in S110 includes: identifying the type of the scene where the vehicle is located using a scene classification model based on the vehicle-collected data to obtain a scene type identification result, wherein the scene type identification result includes at least two scene types corresponding to the scene where the vehicle is located and the confidence levels corresponding to the at least two scene types; if the highest confidence level in the scene type identification result is greater than or equal to a preset threshold, then the scene type corresponding to the highest confidence level is determined as the target scene type. Correspondingly, the door control method may further include: if the highest confidence level in the scene type identification result is less than a preset threshold, then controlling at least one target door to open or close according to a preset conservative control strategy.

[0061] In some examples, the scene classification model can be a trained artificial intelligence model capable of classifying scenes, such as a Support Vector Machine (SVM) classifier.

[0062] For example, a scene feature vector can be extracted from the fused data based on the environmental state vector obtained by fusing vehicle data: Feature = [θ_pitch, θ_roll, d_min, N_obs, w_space, L_person], where θ_pitch is the pitch angle, θ_roll is the roll angle, d_min is the minimum obstacle distance, N_obs is the number of obstacles, w_space is the space width, and L_person is the number of people detected.

[0063] The scene feature vector is input into an SVM classifier, which classifies the scene and outputs the scene type corresponding to the vehicle's location. For example, if the SVM classifier detects θ_pitch < 5°, d_min > 150cm, and w_space > 400cm based on the scene feature vector, the scene type can be output as "flat and open"; if the SVM classifier detects θ_pitch >= 5° and the vehicle is below a slope, the scene type can be output as "uphill"; if the SVM classifier detects θ_pitch >= 5° and the vehicle is above a slope, the scene type can be output as "downhill"; if the SVM classifier detects d_min < 80cm or w_space < 250cm based on the scene feature vector, the scene type can be output as "narrow parking space"; if the SVM classifier detects θ_roll >= 3° and detects crosswind speed based on the scene feature vector, the scene type can be output as "slope with crosswind"; if the SVM classifier detects N_obs >= 3° based on the scene feature vector, the scene type can be output as "slope with crosswind"; if the SVM classifier detects N_obs >= 3°... If 5 and L_person>=3, then the output scenario type is congested road section.

[0064] Furthermore, it should be noted that if the scene classification model outputs only one scene type, that scene type can be directly identified as the target scene type. If the scene classification model outputs multiple scene types, to further improve the accuracy of scene type identification, the confidence level corresponding to each scene type output by the model can be used to ultimately determine the target scene type.

[0065] Based on this, in some implementations, the scene type with the highest confidence level in the scene type identification results output by the model can be determined as the target scene type. Furthermore, to improve the accuracy of subsequent decisions, it can be first determined whether the highest confidence level in the scene type identification results output by the model is greater than a preset threshold (e.g., 0.85). If the highest confidence level is greater than the preset threshold, the scene type corresponding to the highest confidence level in the scene type identification results can be used as the target scene type; conversely, if the highest confidence level is less than the preset threshold, at least one target door can be controlled to open and close according to a preset conservative control strategy. The preset conservative control strategy may include at least one of the following: The opening angle is taken as the more conservative value among adjacent scene types. For example, when the system cannot determine with sufficient confidence whether the current scene is "flat open ground" or "slope," the smaller maximum opening angle upper limit of the two scene types is taken (e.g., if flat ground allows 100% and slope allows 80%, then 80% is ultimately used); the opening speed of all doors is reduced to 0.6 times the base speed, lower than the speed reduction ratio of normal scenes (e.g., 0.7 for slopes, 0.5 for congested roads), taking a middle conservative value; all door interlocking strategies are activated, regardless of the scene, and all interlocking rules are effective simultaneously (e.g., tailgate and sliding door interlocking, double sliding door interlocking, etc.), restricting multiple doors from opening simultaneously; the safety distance threshold is tightened by 10cm, such as increasing the tailgate from 25cm to 35cm and the sliding door from 15cm to 25cm; automatic coordination triggered by intent prediction is disabled, only responding to explicit user trigger commands (i.e., user-inputted door trigger information), and not automatically predicting and linking other doors. In addition, once the confidence level recovers to or above the preset threshold, the system can automatically switch back from the aforementioned conservative control mode to the normal adaptive control mode.

[0066] In this way, by setting a confidence threshold, the scenario type with the highest confidence in the scenario classification results is used as the target scenario type when the scenario classification results are reliable, and the conservative control strategy is used when the classification results are unreliable. This avoids improper opening and closing of the car door due to inaccurate judgment of the scenario type, and improves the safety and reliability of the scenario-adaptive door control process.

[0067] S120 determines the safety constraint information during the door control process based on the target scenario type.

[0068] In some examples, safety constraint information can be constraints that a vehicle should follow when controlling its doors in a scenario corresponding to the target scenario type, in order to improve the safety of people and vehicles.

[0069] For example, a scenario type-constraint mapping table can be pre-built, and then the safety constraint information corresponding to the target scenario type can be determined by querying this scenario type-constraint mapping table. For example, if the scenario type is a flat open area, the safety constraint information can be determined by looking up the table as a maximum opening angle of 100% and a safety distance of 50cm; if the scenario type is a ramp scenario, the safety constraint information can be determined by looking up the table as a maximum tailgate opening angle of 70%-80% and a 30% reduction in opening speed; if the scenario type is a narrow parking space, the safety constraint information can be determined by looking up the table as a maximum sliding door opening angle of 40%-60% and a real-time detection distance; if the scenario type is a congested road section, the safety constraint information can be determined by looking up the table as a 50% reduction in the opening speed of all doors and an expansion of the personnel detection area.

[0070] S130 utilizes a constraint satisfaction model to determine a door control strategy that satisfies safety constraints based on user intent information.

[0071] In some examples, the constraint satisfaction model can be a mathematical model that can dynamically determine the optimal strategy while satisfying the given constraints, such as the Constraint Satisfaction Problem (CSP), which is a type of mathematical problem where the goal is to find variable assignments while satisfying constraints.

[0072] For example, user intent information can be used as the objective function and safety constraint information as the constraint condition to construct a constraint satisfaction problem. When solving the problem, the optimal door action sequence that satisfies the objective function under the constraint condition can be obtained as the door control strategy.

[0073] Specifically, in some embodiments, the above-mentioned S130 may include: constructing constraint parameters based on safety constraint information and user intent information, wherein the constraint parameters include at least one of spatial constraint parameters, interlock constraint parameters, center of gravity constraint parameters, and gate control timing constraint parameters; constructing a constraint satisfaction model based on the user intent information and using the constraint parameters as constraint conditions; and solving the constraint satisfaction model to obtain the door control strategy.

[0074] In some examples, spatial constraint parameters can be constraints used to limit the door movement parameters of each door under different space sizes. For example, limiting the maximum opening angle of each door according to the scene type (e.g., in a ramp scene, the maximum opening angle of the tailgate is 70%-80%, and when θ_pitch>= 10°, the tailgate opening angle does not exceed 70%; in a narrow parking space, the maximum opening angle of the sliding door is 40%-60%, and when a lateral distance d<50cm is detected, the sliding door is prohibited from opening). Interlock constraint parameters can be constraints used to limit the door movement parameters of each door to avoid interference between doors. For example, the tailgate is interlocked with the two sliding doors (e.g., when the tailgate is opened more than 50%, the sliding door is limited to within 30%), and the two sliding doors are interlocked (e.g., only one sliding door is allowed to open fully, and the opening of both doors simultaneously is limited to within 60%). Center of gravity constraint parameters can be constraints that limit the door movement parameters of each door to avoid vehicle center of gravity shift, such as ensuring the tailgate opening angle does not exceed 70% when the slope pitch angle θ_pitch ≥ 10°. Door control timing constraint parameters can be constraints that determine the execution sequence of each door's actions based on user intent. Door movement parameters can include, for example, parameters related to door opening and closing actions such as door opening / closing status, opening / closing speed, and opening angle.

[0075] For example, the constraint parameters required for CSP can be constructed based on security constraint information and user intent information. Specifically, spatial constraint parameters, interlock constraint parameters, and centroid constraint parameters can be constructed based on scenario-related security constraint information, and temporal constraint parameters can be constructed based on user intent information.

[0076] In some examples, constructing timing constraint parameters based on user intent information may specifically include: constructing an intent-gate sequence mapping table, and determining the timing constraint parameters corresponding to the user intent information by querying the intent-gate sequence mapping table.

[0077] For example, an intent-gate sequence mapping table can be constructed as follows: Intent Category 1 - [Tailgate opens to 80%, and if passenger movement is detected after a 2-second delay, the sliding door opens]; Intent Category 2 – [Sliding door opened to 60%, tailgate opened to 50% auxiliary lighting]; Intent Category 3 - [If the tailgate is open, close the tailgate first, and open the driver's side door after a 1-second delay]; Intent Category 4 – [Open each door sequentially according to user distance priority, with a 1.5-second interval between doors]; Intent Category 5 - [Close each door in order of distance from farthest to closest, with a 1-second interval].

[0078] If the current user intent information indicates intent category 1, then the timing constraint parameter can be determined as "the tailgate opens first, and the sliding door opens if passenger movement is detected after a 2-second delay".

[0079] For example, a Control Space Parameter (CSP) is constructed using security constraint information as the constraint condition and user intent information as the objective function. For instance, the variables for constructing the CSP are Door = [D_tail, D_slide_L, D_slide_R, D_front_L, D_front_R], where D_tail is the tailgate, D_slide_L is the left sliding door, D_slide_R is the right sliding door, D_front_L is the left front door, and D_front_R is the right front door; the value range is Value = [closed, angle_10, ..., angle_100], where closed indicates closed, angle_10 indicates an opening angle of 10%, and angle_100 indicates an opening angle of 100% (i.e., fully open); the constraints are Constraint = [spatial constraint parameters, interlocking constraint parameters, centroid constraint parameters, temporal constraint parameters].

[0080] By employing a specific algorithm (such as backtracking) to solve for the CSP, a multi-door cooperative action sequence can be obtained, for example, Sequence = [(Door_1, Angle_1, T_1), (Door_2, Angle_2, T_2), ...], where: Door_i is the door number, Angle_i is the opening angle percentage, and T_i is the execution time. This dynamically solved multi-door cooperative action sequence can be used as the final door control strategy.

[0081] In this way, by defining multi-dimensional constraint parameters such as space, interlocking, center of gravity and timing based on scenario-related safety constraint information and user intent information related to user intent, a constraint satisfaction model can be constructed. This model can comprehensively and systematically express the physical limitations and safety requirements of door opening and closing control under different scenarios and user intents, thereby obtaining a more reasonable and safer door control strategy.

[0082] In addition, to further improve the safety of the door control process, in some embodiments, the above-mentioned constraint satisfaction model is used to obtain the door control strategy. Specifically, this may include: using a backtracking algorithm to solve the constraint satisfaction model to generate a multi-door cooperative action sequence, wherein the multi-door cooperative action sequence includes multiple target doors to be controlled, the door opening angle corresponding to the target doors, and the action time; if it is determined from the multi-door cooperative action sequence that at least two of the multiple target doors need to be opened and closed simultaneously, then it is determined whether there is a target risk when at least two target doors open and close simultaneously, wherein the target risk includes inter-door interference risk and / or vehicle center of gravity shift risk; if there is a target risk, then the door opening angle and / or action time corresponding to at least two target doors are adjusted respectively to eliminate the target risk and obtain the door control strategy.

[0083] For example, if at least two target doors in the solved multi-door cooperative action sequence need to be opened and closed simultaneously, it is possible to detect whether there is a risk of inter-door interference and / or a risk of vehicle center of gravity shift when the at least two target doors are opened and closed simultaneously. Specifically, an inter-door interference matrix and a center of gravity shift calculation model can be established. The inter-door interference matrix is ​​used to detect whether there is a risk of inter-door interference when the at least two target doors are opened and closed simultaneously, and the center of gravity shift calculation model is used to calculate the degree of shift of the vehicle's center of gravity when the at least two target doors are opened and closed simultaneously.

[0084] In some implementations, if there is a risk of inter-door interference between the at least two target doors, and / or the degree of shift in the vehicle's center of gravity is greater than a preset safety threshold when the at least two target doors open and close simultaneously, an interlock strategy (such as tailgate interlock with dual sliding doors, dual sliding door interlock, etc.) is automatically activated. The door opening angles of the at least two target doors in the multi-door coordinated action sequence are adjusted, and / or the action times of the at least two target doors in the multi-door coordinated action sequence are adjusted according to priority order, so that the at least two target doors perform opening and closing actions according to priority order, thereby eliminating the risk of inter-door interference and / or the risk of vehicle center of gravity shift.

[0085] In this way, by actively detecting the risk of inter-door interference and center of gravity shift between target doors that move simultaneously after generating a multi-door coordinated action sequence, and dynamically adjusting the door opening angle or door control timing to eliminate the risk, the safety hazards of door collisions or vehicle instability caused by the simultaneous movement of multiple doors are avoided, further improving the safety of the door control process.

[0086] S140, control at least one target door to open or close according to the door control strategy.

[0087] In some examples, the door control strategy may include information such as at least one target door to be controlled, the action time corresponding to the target door, and door action parameters. For example, each target door can be controlled to perform opening and closing actions at the corresponding action time according to the door action parameters (such as opening and closing speed, opening angle, etc.) of each target door in the door control strategy.

[0088] In addition, during the door control execution phase, a proportional-integral-derivative (PID) closed-loop control algorithm can be used to achieve precise door opening angle control. For example, for tailgate control, door opening angle control can be based on motor current feedback and Hall sensor position feedback; for sliding door control, door opening angle control can be based on guide rail displacement sensor and motor encoder.

[0089] In this embodiment, safety constraint information for the door control process is determined based on the target scenario type corresponding to the vehicle's location. This safety constraint information changes dynamically according to the changes in the vehicle's location. Based on this, a constraint satisfaction model can be used to dynamically determine a door control strategy that satisfies the dynamically changing safety constraint information based on user intent information. This ensures that the door control strategy strictly adheres to the safety constraints in the door control process under the vehicle's location while maximizing the satisfaction of user intent. By controlling the door opening and closing actions according to this dynamically determined door control strategy, the door control can adapt to complex and ever-changing real-world scenarios, thereby ensuring safety during the door control process while meeting user needs.

[0090] In addition, based on the above embodiments, in order to further improve the dynamic adaptability of the door control process to the scene in which the vehicle is located, in the process of controlling at least one target door to perform opening and closing actions according to the door control strategy (step S140), the door control method may further include: adaptively adjusting the door action parameters of at least one target door according to the target scene type; and adjusting the opening and closing actions of at least one target door based on the door action parameters.

[0091] For example, the door action parameters can be adaptively adjusted during the door control process to better suit different scenario types. These door action parameters may include, for example, opening angle, opening speed, and closing speed.

[0092] In some specific examples, if the target scene type is a ramp scene, then the opening angle of the tailgate is controlled by the preset opening angle × (1 - 0.2 × θ_pitch / 10°), where the preset opening angle can be the default opening angle of the tailgate set in the vehicle; the tailgate opening speed is controlled by the base speed × 0.7, and the tailgate closing speed is controlled by the base speed × 1.3, where the base speed can be the default opening and closing speed of the tailgate set in the vehicle.

[0093] In other specific examples, if the target scenario is a narrow parking space, the opening angle of the sliding door is controlled as min(preset opening angle, (d_space - 30cm) / maximum travel × 100%), where the preset opening angle can be the default opening angle of the sliding door set in the vehicle, d_space can be the real-time monitoring distance between the door and surrounding obstacles, and the maximum travel can be the travel distance corresponding to the default maximum opening angle of the sliding door set in the vehicle. Simultaneously, ultrasonic radar is monitored in real-time to determine the specific distance d between the door and surrounding obstacles; when d < 15cm, the door movement immediately stops.

[0094] In some specific examples, if the target scenario is a congested road, the opening speed of all doors is controlled to be equal to the base speed multiplied by 0.5, where the base speed can be the default door opening speed set within the vehicle. Simultaneously, the vehicle continuously monitors the area of ​​human movement during the opening process, stopping the door operation if someone approaches the door's path.

[0095] After obtaining the adaptively adjusted door action parameters, the opening and closing action of at least one target door can be dynamically adjusted based on the door action parameters (such as opening angle, opening and closing speed, etc.) during the opening and closing action of the target door.

[0096] In this way, by dynamically linking scene types with door action parameters, the door control strategy can adjust the target door's action parameters in real time according to the scene during operation. This allows for dynamic adjustment of the target door's opening and closing actions in various scenes, improving the physical adaptability and safety of the door control process to diverse environments.

[0097] In addition, during the door opening and closing process, the ultrasonic radar can scan the distance to obstacles along the door's movement path at a frequency of 20Hz. When an obstacle distance d is detected to be less than a preset distance threshold, the door movement is immediately paused and an audible and visual alarm is issued. The door will resume opening and closing only after the obstacle is removed. Different preset distance thresholds can be set for different types of doors; for example, a safety-assured preset distance threshold of 25cm can be set for the tailgate, and a safety-assured preset distance threshold of 15cm can be set for the sliding doors.

[0098] In addition, to further improve the dynamic adaptability of the door control process to abnormal situations, during the process of controlling at least one target door to perform opening and closing actions according to the door control strategy (step S140), the door control method may further include: if a target abnormality is detected, determining a corresponding abnormality handling strategy based on the frequency of occurrence of the target abnormality, wherein the target abnormality includes at least one of obstacle sudden change abnormality, door motor abnormality, and vehicle posture abnormality; and performing abnormality handling operations on the door according to the abnormality handling strategy.

[0099] In some examples, the target anomaly can be an abnormal situation that may affect the safety of the vehicle or occupants during the opening and closing of the door, such as abrupt obstacle changes, door motor malfunctions, or abnormal vehicle posture. For example, detecting abrupt obstacle changes could involve using ultrasonic radar to monitor the distance *d* of obstacles along the door's movement path; if *d* < 10 cm, an obstacle change anomaly is detected. Detecting door motor malfunctions could involve monitoring the current of the motor controlling the door; if the current exceeds a threshold of 150% for 100 ms, the motor is considered stalled, indicating a door motor malfunction. Detecting abnormal vehicle posture could involve using a gyroscope to detect the roll angle *θ_roll*; if *θ_roll* > 5°, a risk of vehicle roll is identified, indicating an abnormal vehicle posture.

[0100] For example, the corresponding anomaly handling strategy can be determined based on the frequency of occurrence of the target anomaly. Different frequencies of occurrence can correspond to different anomaly handling strategies, and different anomaly handling operations can be performed according to different anomaly handling strategies. For example, for the first anomaly within a target time period (such as within the last minute), the corresponding anomaly handling strategy may include pausing the movement of the currently moving door, issuing an alarm prompt, and waiting for user confirmation or obstacle removal; for the second anomaly within the target time period, the corresponding anomaly handling strategy may include retracting the current door to a safe position, such as closing or opening it to 30%; for the third or more anomalies within the target time period (such as consecutive occurrences), the corresponding anomaly handling strategy may include locking all automatically controlled doors, switching to manual mode, and simultaneously prompting the user of the target anomaly through the dashboard.

[0101] In this way, by adopting a graded anomaly handling strategy based on the frequency of occurrence of target anomalies, the availability of door functions can be maintained as much as possible while avoiding danger. This avoids the inconvenience caused by overly conservative door control methods or the danger caused by aggressive door control methods, thereby improving the intelligence of anomaly handling and system resilience.

[0102] In addition, to further improve the dynamic adaptability of the door control process to abnormal battery conditions, during the process of controlling at least one target door to perform opening and closing actions according to the door control strategy (step S140), the door control method may also include: detecting the battery parameters of the vehicle; if the battery parameters meet the abnormal parameter conditions, controlling at least one target door to stop the opening and closing actions, and switching the door control mode to manual mode.

[0103] In some examples, battery parameters can be parameters reflecting the battery's state, such as current, voltage, and remaining charge. Abnormal parameter conditions can be parameter conditions corresponding to battery parameters that indicate that the battery is in an abnormal condition, such as current being higher or lower than a preset current threshold, voltage being higher or lower than a preset voltage threshold, or remaining charge being lower than a preset charge threshold.

[0104] For example, if the system detects that the battery voltage in the vehicle is below 9V, it can control all moving doors to stop immediately and switch the door control mode to manual mode, such as automatically switching the door locks to mechanical locking and allowing the user to manually control the doors.

[0105] In this way, by monitoring battery parameters in real time and actively stopping the automatic control of the doors and switching to manual mode when abnormal parameters are met, adverse consequences such as insufficient driving force, loss of control of movement or door lock failure caused by abnormal power supply are avoided, thus ensuring the safety of the vehicle and personnel in the event of abnormal power supply.

[0106] Furthermore, when any sensor malfunctions, such as when data collection is lost for more than a preset period, a degraded mode can be used for door control. For example, if the camera malfunctions, only data collected by sensors such as PKE, ultrasonic, and gyroscope can be used, and the user intent prediction function can be disabled; if the ultrasonic radar malfunctions, only data collected by the camera can be used for visual ranging, and the opening angle of all doors can be limited to within 50%; if the gyroscope malfunctions, the hill-adaptive function can be disabled, and conservative parameters can be used by default, such as limiting the maximum opening angle of all doors to 70%, limiting the opening speed to 0.7 times the base speed, and limiting the closing speed to 1.3 times the base speed, while tightening the safety distance threshold.

[0107] Based on the above embodiments, the following key algorithm parameters and value ranges can be adopted in the door control process: PKE sampling frequency can be 1Hz; camera frame rate can be 15fps; ultrasonic radar sampling frequency can be 10Hz, ranging range can be 0-200cm, and accuracy can be ±2cm; gyroscope sampling frequency can be 20Hz, angle measurement range can be ±30°, and accuracy can be ±0.1°; intent prediction time window can be 30 seconds for PKE signal and 10 seconds for personnel trajectory; door movement speed range can be 5-25cm / s (tailgate) and 8-30cm / s (sliding door); safety distance threshold can be 25cm for tailgate and 15cm for sliding door; scene classification confidence threshold can be set to 0.85.

[0108] Based on the above embodiments, the following is combined with Figure 2 Describe the door control system provided in this application.

[0109] Figure 2 The diagram shown is a structural schematic of a vehicle door control system according to an embodiment of this application. This system can be configured in a vehicle. Figure 2 As shown, the door control system 200 may include a user interaction layer 201, a perception fusion layer 202, a decision planning layer 203, and an execution control layer 204.

[0110] In some examples, the user interaction layer 201 may include a PKE / Remote Keyless Entry (RKE) remote control, body buttons / touchscreen, voice control module, etc.

[0111] The perception fusion layer 202 may include a PKE module, a surround-view camera, an ultrasonic radar, a gyroscope, and wheel speed sensors. The PKE module is used to detect user identity and location, the surround-view camera is used to detect obstacle distribution, the ultrasonic radar is used to collect distance data, the gyroscope is used to detect vehicle attitude, and the wheel speed sensors are used to detect vehicle speed.

[0112] The decision planning layer 203 may include a scenario classification module, an intent prediction module, a collaborative planning module, and an interlocking strategy module. The scenario classification module determines the scenario type, the intent prediction module determines the user's intent, the collaborative planning module determines the door action sequence, and the interlocking strategy module applies interlocking safety constraints.

[0113] The execution control layer 204 may include a body control module (BCM), a tailgate actuator, a sliding door actuator, an electric door actuator, etc.

[0114] It should be noted that the parts not described in detail above can be referred to the relevant parts of the foregoing embodiments, and will not be repeated here.

[0115] The above text combined Figures 1 to 2 The embodiments of the door control method of this application are described in detail below, in conjunction with... Figure 3 This application describes in detail embodiments of the door control device. It should be understood that the descriptions of the door control method embodiments correspond to the descriptions of the door control device embodiments; therefore, any parts not described in detail can be found in the preceding method embodiments.

[0116] Figure 3 The diagram shown is a structural schematic of a door control device according to an embodiment of this application. This device can be applied to vehicles. Figure 3 As shown, the door control device 300 provided in this application embodiment includes: The information determination module 301 is used to determine the user intent information and the target scene type corresponding to the scene in which the vehicle is located based on the vehicle collected data; The constraint determination module 302 is used to determine safety constraint information during the door control process based on the target scenario type. The strategy determination module 303 is used to determine a door control strategy that satisfies safety constraints based on user intent information using a constraint satisfaction model. The motion control module 304 is used to control at least one target door to perform opening and closing actions according to the door control strategy.

[0117] In one embodiment of this application, the strategy determination module 303 is further configured to: construct constraint parameters based on safety constraint information and user intent information, wherein the constraint parameters include at least one of spatial constraint parameters, interlock constraint parameters, center of gravity constraint parameters, and gate timing constraint parameters; construct a constraint satisfaction model based on the user intent information and using the constraint parameters as constraint conditions; and solve the constraint satisfaction model to obtain the door control strategy.

[0118] In one embodiment of this application, the strategy determination module 303 is further configured to: solve the constraint satisfaction model using a backtracking algorithm to generate a multi-door cooperative action sequence, wherein the multi-door cooperative action sequence includes multiple target doors to be controlled, the door opening angle corresponding to the target doors, and the action time; if it is determined from the multi-door cooperative action sequence that at least two of the multiple target doors need to be opened and closed simultaneously, then it is determined whether there is a target risk when the at least two target doors open and close simultaneously, wherein the target risk includes inter-door interference risk and / or vehicle center of gravity offset risk; if there is a target risk, then the door opening angle and / or action time corresponding to the at least two target doors are adjusted respectively to eliminate the target risk and obtain a door control strategy.

[0119] In one embodiment of this application, the information determination module 301 is further configured to: determine the user's motion timing characteristics based on the user data collected by the vehicle; and predict the user's intent based on the user's motion timing characteristics using a user intent prediction model to obtain user intent information.

[0120] In one embodiment of this application, the information determination module 301 is further configured to: if a fault is detected in the target sensor, acquire door trigger information input by the user, wherein the target sensor is a sensor in the vehicle used to collect user data; and determine user intent information based on the door trigger information.

[0121] In one embodiment of this application, the information determination module 301 is further configured to: identify the type of scene in which the vehicle is located based on the vehicle collected data using a scene classification model, and obtain a scene type identification result, wherein the scene type identification result includes at least two scene types corresponding to the scene in which the vehicle is located and the confidence level corresponding to the at least two scene types; if the highest confidence level in the scene type identification result is greater than or equal to a preset threshold, then the scene type corresponding to the highest confidence level is determined as the target scene type; if the highest confidence level in the scene type identification result is less than the preset threshold, then at least one target door is controlled to open or close according to a preset conservative control strategy.

[0122] In one embodiment of this application, the motion control module 304 is further configured to: adaptively adjust the door motion parameters of at least one target door according to the target scene type; and adjust the opening and closing motion of at least one target door based on the door motion parameters.

[0123] In one embodiment of this application, the motion control module 304 is further configured to: if a target anomaly is detected, determine a corresponding anomaly handling strategy based on the frequency of occurrence of the target anomaly, wherein the target anomaly includes at least one of obstacle abrupt change anomaly, door motor anomaly, and vehicle posture anomaly; and perform anomaly handling operation on the door according to the anomaly handling strategy.

[0124] In one embodiment of this application, the motion control module 304 is further configured to: detect the battery parameters of the vehicle; if the battery parameters meet the abnormal parameter conditions, control at least one target door to stop opening and closing, and switch the door control mode to manual mode.

[0125] Below, for reference Figure 4 This describes an electronic device according to embodiments of the present application. Figure 4 The diagram shown is a structural schematic of an electronic device provided in an exemplary embodiment of this application.

[0126] like Figure 4 As shown, the electronic device 400 includes one or more processors 401 and memory 402.

[0127] The processor 401 may be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 400 to perform desired functions.

[0128] The memory 402 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 401 may execute the program instructions to implement the door control methods of the various embodiments of this application described above and / or other desired functions. The computer-readable storage medium may also store various content such as vehicle acquisition data, target scene type, safety constraint information, door control strategies, etc.

[0129] In one example, the electronic device 400 may also include an input device 403 and an output device 404, which are interconnected via a bus system and / or other forms of connection mechanism (not shown).

[0130] The input device 403 may include, for example, a keyboard, a mouse, etc.

[0131] The output device 404 can output various information to the outside, including vehicle data, target scene type, safety constraint information, door control strategy, etc. The output device 404 may include, for example, a display, speaker, printer, and communication network and its connected remote output devices, etc.

[0132] Of course, for the sake of simplicity, Figure 4 Only some of the components of the electronic device 400 relevant to this application are shown in this illustration; components such as buses, input / output interfaces, etc., are omitted. In addition, the electronic device 400 may include any other suitable components depending on the specific application.

[0133] In addition to the methods and devices described above, embodiments of this application may also be computer program products, which include computer program instructions that, when executed by a processor, cause the processor to perform the steps in the door control methods according to various embodiments of this application described above.

[0134] The computer program product can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of this application. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0135] Furthermore, embodiments of this application may also be computer-readable storage media storing computer program instructions thereon, which, when executed by a processor, cause the processor to perform the steps in the door control methods according to various embodiments of this application described above.

[0136] The computer-readable storage medium may be any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0137] The basic principles of this application have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this application are merely examples and not limitations, and should not be considered as essential features of each embodiment of this application. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the application to the necessity of employing the aforementioned specific details for implementation.

[0138] The block diagrams of devices, apparatuses, devices, and systems involved in this application are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.

[0139] It should also be noted that in the apparatus, equipment, and methods of this application, the components or steps can be disassembled and / or recombined. These disassemblies and / or recombinations should be considered as equivalent solutions of this application.

[0140] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of this application. Therefore, this application is not intended to be limited to the aspects shown herein, but rather to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0141] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this application to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

Claims

1. A door control method, characterized in that, include: Based on vehicle-collected data, determine the user intent information and the target scenario type corresponding to the vehicle's location. Determine the safety constraint information during the door control process based on the target scenario type; Using a constraint satisfaction model, a door control strategy that satisfies the safety constraints is determined based on the user intent information. Control at least one target door to open or close according to the door control strategy.

2. The method according to claim 1, characterized in that, The method of using a constraint satisfaction model to determine a door control strategy that satisfies the safety constraints based on the user intent information includes: Constraint parameters are constructed based on the security constraint information and the user intent information, wherein the constraint parameters include at least one of spatial constraint parameters, interlock constraint parameters, centroid constraint parameters, and gating timing constraint parameters; Based on the user intent information, a constraint satisfaction model is constructed using the constraint parameters as constraints. Solve the constraint satisfaction model to obtain the door control strategy.

3. The method according to claim 2, characterized in that, Solving the constraint-satisfying model yields the door control strategy, including: The constraint satisfaction model is solved by using a backtracking algorithm to generate a multi-door cooperative action sequence, wherein the multi-door cooperative action sequence includes multiple target doors to be controlled, the door opening angle corresponding to the target doors, and the action time; If, based on the multi-door coordinated action sequence, it is determined that at least two of the multiple target doors need to be opened and closed simultaneously, then it is determined whether there is a target risk when the at least two target doors are opened and closed simultaneously. The target risk includes door interference risk and / or vehicle center of gravity shift risk. If the target risk exists, the door opening angle and / or action time corresponding to the at least two target doors are adjusted to eliminate the target risk, thereby obtaining the door control strategy.

4. The method according to claim 1, characterized in that, Based on vehicle-collected data, user intent information is determined, including: Determine the temporal characteristics of user movement based on user data collected from vehicles; The user intent information is obtained by predicting the user intent based on the user's movement time sequence characteristics using a user intent prediction model.

5. The method according to claim 4, characterized in that, Also includes: If a fault is detected in the target sensor, the door trigger information input by the user is obtained, wherein the target sensor is the sensor in the vehicle used to collect the user data; The user intent information is determined based on the door trigger information.

6. The method according to claim 1, characterized in that, Based on the vehicle-collected data, the target scene type corresponding to the vehicle's location is determined, including: Based on vehicle-collected data, a scene classification model is used to identify the type of scene in which the vehicle is located, and a scene type identification result is obtained. The scene type identification result includes at least two scene types corresponding to the scene in which the vehicle is located and the confidence level corresponding to the at least two scene types. If the highest confidence level in the scene type identification result is greater than or equal to a preset threshold, then the scene type corresponding to the highest confidence level is determined as the target scene type; The method further includes: If the highest confidence level in the scene type identification result is less than the preset threshold, then at least one target door is controlled to open or close according to the preset conservative control strategy.

7. The method according to any one of claims 1 to 6, characterized in that, In the process of controlling at least one target door to open and close according to the door control strategy, the method further includes: The door action parameters of at least one target door are adaptively adjusted according to the target scene type; The opening and closing actions of at least one target door are adjusted based on the door action parameters.

8. The method according to any one of claims 1 to 6, characterized in that, In the process of controlling at least one target door to open and close according to the door control strategy, the method further includes: If a target anomaly is detected, a corresponding anomaly handling strategy is determined based on the frequency of occurrence of the target anomaly, wherein the target anomaly includes at least one of obstacle mutation anomaly, door motor anomaly, and vehicle posture anomaly. Perform abnormal handling operations on the car door according to the aforementioned abnormal handling strategy.

9. The method according to any one of claims 1 to 6, characterized in that, In the process of controlling at least one target door to open and close according to the door control strategy, the method further includes: Detect the battery parameters of the vehicle; If the battery parameters meet the abnormal parameter conditions, then control at least one target door to stop opening and closing, and switch the door control mode to manual mode.

10. A vehicle, characterized in that, include: processor; Memory used to store the processor's executable instructions; The processor is used to execute the door control method according to any one of claims 1 to 9.

Images