Vehicle electric tailgate unlocking method and related equipment

Abstract

The present invention provides a vehicle electric tailgate unlocking method and related equipment, the method comprising the following steps: detecting a first action performed by a target user in a preset identification area of ​​a target vehicle; after detecting that the target user performs the first action in the preset identification area of ​​the target vehicle, obtaining a mobile behavior data segment of the target user within a preset time length before performing the first action; inputting the mobile behavior data into a preset tailgate control prediction model, and outputting the behavior prediction result of the target user through the tailgate control prediction model; if the behavior prediction result is to control the tailgate behavior, projecting a first projection pattern in a preset projection area, and generating a first control instruction according to the behavior prediction result after projecting the first projection pattern; and controlling the tailgate of the target vehicle to open or close according to the first control instruction. The present invention has the effect of improving the convenience and adaptability of the electric tailgate operation method.

Description

Technical Field

[0001] The invention belongs to the technical field of vehicle control, and in particular relates to a vehicle electric tailgate unlocking method and related equipment. Background Art

[0002] With the booming development of the automobile industry and consumers' constant pursuit of quality of life, electric tailgates have become standard equipment for most models. The original intention of designing this function is to solve the many inconveniences caused by traditional manual tailgate operation. The widespread application of electric tailgate technology has greatly improved the practicality and comfort of vehicles, and plays an important role in many scenarios in daily life, such as shopping, loading and unloading luggage, and transporting large items. The traditional electric tailgate operation method mainly relies on active triggering methods such as buttons or remote controls. This operation mode requires users to free their hands to press buttons on the car or operate the remote control, which will undoubtedly cause considerable inconvenience when the user's hands are already full of things. For example, when returning from shopping, with both hands full of shopping bags, users may need to put the shopping bags on the ground before they can operate the remote control or the button on the car to open the tailgate, which not only increases the complexity of the operation, but may also affect the safety of users in some cases.

[0003] In addition, in some special cases, the traditional electric tailgate operation method may also face challenges. For example, on rainy days, users may need to hold an umbrella with one hand and hold items with the other hand, which makes it extremely difficult to operate the remote control or buttons. Or in the cold winter, users wearing heavy gloves may affect the precise operation of the buttons. These situations highlight the shortcomings of the traditional electric tailgate operation method in terms of convenience and adaptability. Summary of the invention

[0004] The present invention provides a vehicle electric tailgate unlocking method and related equipment to solve the problem that the conventional electric tailgate operation mode is insufficient in convenience and adaptability.

[0005] In a first aspect, the present invention provides a method for unlocking a vehicle electric tailgate, the method comprising the following steps:

[0006] When it is determined that the current state of the target vehicle satisfies a preset safety precondition, detecting a first action performed by the target user in a preset identification area of ​​the target vehicle;

[0007] After detecting that the target user performs a first action in a preset identification area of ​​the target vehicle, obtaining a movement behavior data segment of the target user within a preset time length before performing the first action, the movement behavior data in the movement behavior data segment is obtained through the camera and sensor of the target vehicle, and the movement behavior data includes the movement speed, movement direction, in-situ waiting position and in-situ waiting time of the target user;

[0008] Inputting the movement behavior data into a preset tailgate control prediction model, and outputting the behavior prediction result of the target user through the tailgate control prediction model;

[0009] If the behavior prediction result is to control the tailgate behavior, projecting a first projection pattern in a preset projection area, and generating a first control instruction according to the behavior prediction result after projecting the first projection pattern;

[0010] The tailgate of the target vehicle is controlled to be opened or closed according to the first control instruction.

[0011] Optionally, when it is determined that the current state of the target vehicle satisfies a preset safety precondition, detecting a first action performed by the target user in a preset identification area of ​​the target vehicle comprises the following steps:

[0012] When it is determined that the current state of the target vehicle meets the preset safety precondition, it is determined whether the vehicle Bluetooth module of the target vehicle is in a Bluetooth connection state with any target terminal, wherein the target terminal is a mobile terminal that has been pre-bidirectionally authenticated with the vehicle Bluetooth module;

[0013] If the vehicle Bluetooth module is in a Bluetooth connection state with any of the target terminals, when the target user enters the preset detection range of the target vehicle, the target vehicle's camera and sensor will continuously collect the target user's movement behavior data within the detection range until the target user is detected to perform a first action in the preset identification area of ​​the target vehicle or the target user is detected to have left the detection range for a period exceeding a preset time.

[0014] Optionally, after detecting that the target user performs a first action in a preset recognition area of ​​the target vehicle, obtaining a movement behavior data segment of the target user within a preset time length before performing the first action comprises the following steps:

[0015] If it is detected that the target user performs a first action in the preset recognition area of ​​the target vehicle, the continuously collected movement behavior data is stored in a timestamp order as a first movement behavior data set;

[0016] If the data time span of the first mobile behavior data set is less than or equal to a preset time length, the first mobile behavior data set is used as a mobile behavior data segment of the target user before performing the first action and within the time length;

[0017] If the data time span of the first movement behavior data set is greater than the time length, the movement behavior data segment is cut out from the end of the first movement behavior data set based on the time length.

[0018] Optionally, the method further comprises the following steps:

[0019] If it is detected that the target user has been away from the detection range for a time period exceeding a preset time period, the continuously collected movement behavior data are stored in a timestamp order as a second movement behavior data set;

[0020] A first label is marked for the second movement behavior data set, and the tailgate control prediction model is trained using the second movement behavior data set marked with the first label, wherein the first label indicates that the behavior prediction result is a non-control tailgate behavior.

[0021] Optionally, after controlling the tailgate of the target vehicle to open or close according to the first control instruction, the following steps are also included:

[0022] During the opening or closing process of the tailgate of the target vehicle, if it is detected that the target user performs a second action in the projection area where the first projection pattern is projected, the first control instruction is terminated, and a second control instruction is generated according to the second action;

[0023] Controlling the tailgate of the target vehicle to open or close according to the second control instruction, wherein the second control instruction controls the action of the tailgate of the target vehicle opposite to that of the first control instruction;

[0024] A second label is marked for the behavior prediction result, and the tailgate control prediction model is trained in combination with the first mobile behavior data set and the behavior prediction result marked with the second label, wherein the second label indicates that the behavior prediction result is predicted incorrectly.

[0025] Optionally, the method further comprises the following steps:

[0026] If the target user is not detected to perform a second action in the projection area where the first projection pattern is projected during the opening or closing process of the tailgate of the target vehicle, then when the tailgate of the target vehicle is completely closed or completely opened, the projection of the first projection pattern is terminated;

[0027] A third label is marked for the behavior prediction result, and the tailgate control prediction model is trained in combination with the first mobile behavior data set and the behavior prediction result marked with the third label, wherein the third label indicates that the behavior prediction result is predicted correctly.

[0028] Optionally, the method further comprises the following steps:

[0029] If the behavior prediction result is a non-controlled tailgate behavior, projecting a second projection pattern on the projection area based on a preset projection time;

[0030] If it is detected that the target user performs a third action in the projection area projected with the second projection pattern within the projection time, a first control instruction is generated according to the third action, and the tailgate of the target vehicle is controlled to be opened or closed according to the first control instruction;

[0031] A second label is marked for the behavior prediction result, and the tailgate control prediction model is trained in combination with the first mobile behavior data set and the behavior prediction result marked with the second label, wherein the second label indicates that the behavior prediction result is predicted incorrectly.

[0032] Optionally, the method further comprises the following steps:

[0033] If the target user is not detected performing a third action in the projection area where the second projection pattern is projected within the projection time, a third label is marked for the behavior prediction result, and the tailgate control prediction model is trained in combination with the first mobile behavior dataset and the behavior prediction result marked with the third label, and the third label indicates that the behavior prediction result is predicted correctly.

[0034] In a second aspect, the present invention also provides a vehicle electric tailgate unlocking system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the vehicle electric tailgate unlocking method as described in the first aspect is implemented.

[0035] In a third aspect, the present invention further provides a computer-readable storage medium having instructions stored thereon, which, when executed by a processor, configures the processor to execute the vehicle electric tailgate unlocking method described in the first aspect.

[0036] The beneficial effects of the present invention are:

[0037] The present invention greatly improves the convenience and intelligence of the user experience. It enhances the interactive experience between the user and the vehicle by intelligently sensing and predicting the user's behavioral intentions and introducing a projection feedback mechanism, and realizes automatic control without manual operation, solving the inconvenience of the user needing to free up his hands to operate buttons or remote controls in traditional methods. Secondly, the present invention effectively reduces the risk of misoperation and improves the safety and reliability of the system by introducing the determination of safety preconditions. The present invention can also more accurately predict user intentions by collecting and analyzing user's mobile behavior data, thereby providing more personalized and intelligent services. At the same time, this control method based on behavior prediction also improves the adaptability of the system and can cope with various complex usage scenarios. On the other hand, the present invention fully utilizes the existing hardware resources of the vehicle by integrating camera and sensor data, and realizes functional upgrades without increasing additional hardware costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 The process diagram of the vehicle electric tailgate unlocking method in one embodiment of the present application is as follows Figure 1 .

[0039] Figure 2 The process diagram of the vehicle electric tailgate unlocking method in one embodiment of the present application is as follows Figure 2 . DETAILED DESCRIPTION

[0040] The following will be combined with the drawings in the embodiments of the present application to clearly describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments in the present application belong to the scope of protection of this application.

[0041] The terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described here, and the objects distinguished by "first", "second", etc. are generally of one type, and the number of objects is not limited. For example, the first object can be one or more. In addition, "and / or" in the specification and claims represents at least one of the connected objects, and the character " / " generally indicates that the objects associated with each other are in an "or" relationship.

[0042] Figure 1 FIG. 1 is a flow chart of a method for unlocking a vehicle electric tailgate in one embodiment. It should be understood that although Figure 1The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1 At least part of the steps in the above method may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps. Figure 1 As shown, a vehicle electric tailgate unlocking method disclosed in the present invention specifically includes the following steps:

[0043] S101. When it is determined that the current state of the target vehicle satisfies a preset safety precondition, a first action performed by the target user in a preset recognition area of ​​the target vehicle is detected.

[0044] Among them, in this step, it is first necessary to determine whether the current state of the target vehicle meets the preset safety preconditions. These safety preconditions include that the vehicle is stationary, the gearbox is in the parking gear, the electronic handbrake is enabled, and the door is locked. The determination of these conditions is achieved through various sensors of the vehicle, such as speed sensor, gearbox position sensor, handbrake state sensor and door lock sensor. Only when all these safety conditions are met will the next step be entered. After the safety conditions are met, the system will start monitoring the preset identification area around the target vehicle. This identification area is usually set near the rear of the vehicle. The monitoring process can be carried out by combining a camera with an infrared sensor. The camera is responsible for capturing visual images, while the infrared sensor is used to detect heat sources. In this way, the first action performed by the target user in the identification area is detected, and the first action can be a preset specific action such as staying or kicking. The implementation effect of this step is to ensure that the system will only be activated in safe and appropriate circumstances, effectively avoiding misoperation and potential safety hazards. At the same time, by setting specific identification areas and triggering actions, the accuracy and reliability of the system are also improved, and the possibility of mistriggers is reduced. This design takes into account both the convenience of the user and the safety and accuracy of the operation.

[0045] S102. After detecting that the target user performs a first action in a preset recognition area of ​​the target vehicle, obtain a movement behavior data segment of the target user within a preset time length before performing the first action.

[0046] Among them, after detecting that the target user performs the first action in the preset recognition area, the system will immediately start to trace back and obtain the user's mobile behavior data for a period of time before the first action is performed. This preset time length may vary from 10 to 30 seconds, depending on the design and performance requirements of the system. The acquisition of mobile behavior data mainly relies on the cameras and various sensors equipped in the vehicle. The camera may include a 360-degree panoramic camera system around the vehicle, which can capture the entire process of the user approaching the vehicle. Sensors may include infrared sensors, ultrasonic sensors, lidar, etc., which are used to accurately measure the distance changes between the user and the vehicle. Specifically, mobile behavior data includes the following aspects:

[0047] Movement speed: By analyzing continuous image frames, the system can calculate the user's movement speed. For example, if the user moves 2 meters in 1 second, then the movement speed is 2 meters / second.

[0048] Direction of movement: Also through image analysis, the system can determine the user's movement trajectory and direction. This may be expressed as an angle value, such as the offset angle relative to the centerline of the rear of the vehicle.

[0049] Waiting position: If the user stops while approaching the vehicle, the system will record the coordinates of the location where they stopped. This coordinate may be the XY coordinate value relative to a fixed point of the vehicle (such as the center of the rear bumper).

[0050] Waiting time in place: For the user's stay, the system will also record the duration of the stay. This time can be accurate to milliseconds.

[0051] These detailed behavior data provide a rich information basis for the next step of behavior prediction, and can help the system judge the user's intention more accurately. By analyzing the user's approach speed, direction change, and stop behavior, it can be distinguished whether the approach is intended to open the tailgate, or just passing by or other unrelated behaviors. This data collection and analysis greatly improves the intelligence level and accuracy of the system, laying the foundation for subsequent automated operations.

[0052] S103. Input the mobile behavior data into a preset tailgate control prediction model, and output the behavior prediction result of the target user through the tailgate control prediction model.

[0053] Among them, after obtaining the user's mobile behavior data, these data need to be input into the preset tailgate control prediction model. This prediction model is a complex model based on a machine learning algorithm, such as a deep learning model suitable for processing time series data such as a recurrent neural network (RNN) or a long short-term memory network (LSTM). The input of the model is the mobile behavior data obtained in the previous step, including information such as the user's movement speed, movement direction, waiting position, and waiting time in situ over a period of time. These data may be organized in the form of a time series, with each time point corresponding to a set of feature values. For example, assuming one data point per second, the input data is presented in the following format:

[0054] [t1,v1,θ1,x1,y1,w1]

[0055] [t2,v2,θ2,x2,y2,w2] ...

[0056] [tn,vn,θn,xn,yn,wn]

[0057] Among them, t represents time, v represents speed, θ represents direction angle, (x, y) represents position coordinates, w represents waiting time (0 if there is no waiting at this moment), and n represents the number of data points.

[0058] The prediction model processes and analyzes these input data. The model contains multiple hidden layers, each of which extracts and transforms features from the data. For example, the model will focus on the pattern of speed changes, the frequency of direction adjustments, the characteristics of stop behavior, etc. The output of the model is the prediction result of the user's behavior, which is a probability distribution that represents the probability of different behaviors. For example: the probability of opening the tailgate: 0.85, the probability of passing by: 0.10, and the probability of other behaviors: 0.05. The training process of the model involves a large amount of historical data, including the user's behavior data in various scenarios and the final actual operation results. Through repeated training and optimization, the model can learn the relationship between various behavior patterns and the final intention. The use of the prediction model greatly improves the intelligence level of the system. It can make accurate predictions based on the user's behavior pattern, rather than relying solely on a single trigger action. This means that the user's intention can be better understood, misjudgment and unnecessary operations can be reduced, and more intelligent and humane services can be provided. At the same time, because the model can be continuously learned and updated, the performance of the system can be continuously improved with the increase of usage time, adapting to the habits and preferences of different users.

[0059] S104. If the behavior prediction result is to control the tailgate behavior, a first projection pattern is projected in a preset projection area, and a first control instruction is generated according to the behavior prediction result after the first projection pattern is projected.

[0060] Among them, when the behavior output by the tailgate control prediction model is to control the tailgate behavior, it indicates that the target user has the intention to control the tailgate. At this time, the system will project the first projection pattern in the preset projection area. This projection area is usually set on the ground at the rear of the vehicle. The projection system can use LED projection technology to project the pattern onto the ground through a projection device hidden at the rear of the vehicle. The design of the first projection pattern should be concise and clear, and be able to clearly convey information. The first projection pattern can be a dynamic arrow pattern pointing to the location of the tailgate. The display of the projection pattern is not just static, but may include animation effects. For example, the arrow may have a pulsating effect, or the text may flash and fade to attract the user's attention. The brightness of the projection will automatically adjust according to the ambient light to ensure that it can be clearly seen under various lighting conditions.

[0061] While projecting the first projection pattern, the system will generate the first control instruction based on the behavior prediction results. This control instruction is a data packet containing multiple parameters, including the following information: Operation type: open / close; Operation timing: immediate / delayed for X seconds; Opening range: fully open / partially open (such as open to 70%); Opening speed: normal / slow / fast; Sound prompt: yes / no. The generation of this control instruction will take into account multiple factors, including the user's approach speed, weather conditions (such as whether it is raining), the loading status of the vehicle, etc. For example, if it is detected that the user is approaching at a fast speed, the system may choose a faster opening speed; if it is detected that it is raining, the system may choose to partially open to reduce rainwater from entering the car.

[0062] Through the generation of projection patterns and control instructions, intuitive interaction and precise control with users are achieved. The projection patterns provide users with clear visual feedback, letting them know that the system has recognized their intentions. At the same time, refined control instructions ensure that the tailgate operation can best meet the needs of users and current environmental conditions, providing a more intelligent and humane service experience.

[0063] S105. Control the tailgate of the target vehicle to open or close according to the first control instruction.

[0064] Among them, after the first control instruction is generated, the system will control the tailgate of the target vehicle to open or close according to this instruction. First, the system will quickly check the safety conditions again to ensure that the safety situation has not changed in the short time from the generation of the control instruction to the actual execution, including checking whether the vehicle is still stationary and whether the surrounding environment is still safe. Then, the system will adjust the tailgate drive device according to the parameters in the control instruction. During the opening process, the system will also continue to monitor obstacles, which can be achieved by ultrasonic sensors or infrared sensors. If any obstacles are detected during the opening process, the system will immediately stop the opening operation and retract the tailgate slightly to ensure safety.

[0065] In one embodiment, when it is determined that the current state of the target vehicle satisfies a preset safety precondition, detecting a first action performed by the target user in a preset identification area of ​​the target vehicle includes the following steps:

[0066] When it is determined that the current state of the target vehicle meets the preset safety precondition, it is determined whether the vehicle Bluetooth module of the target vehicle is in a Bluetooth connection state with any target terminal, and the target terminal is a mobile terminal that has been pre-bidirectionally authenticated with the vehicle Bluetooth module;

[0067] If the vehicle's Bluetooth module is in a Bluetooth connection state with any target terminal, when the target user enters the preset detection range of the target vehicle, the target vehicle's camera and sensor will continuously collect the target user's movement behavior data within the detection range until the target user is detected to perform the first action in the preset identification area of ​​the target vehicle or the target user is detected to have left the detection range for longer than the preset time.

[0068] In this embodiment, after detecting that the target user performs a first action in a preset recognition area of ​​the target vehicle, obtaining a movement behavior data segment of the target user within a preset time length before performing the first action includes the following steps:

[0069] If it is detected that the target user performs a first action in the preset recognition area of ​​the target vehicle, the continuously collected movement behavior data is stored in a timestamp order as a first movement behavior data set;

[0070] If the data time span of the first mobile behavior data set is less than or equal to the preset time length, the first mobile behavior data set is used as a mobile behavior data segment of the target user before performing the first action and within the time length;

[0071] If the data time span of the first movement behavior data set is greater than the time length, a movement behavior data segment is cut out from the end of the first movement behavior data set based on the time length.

[0072] In this embodiment, first, it is determined whether the current state of the target vehicle meets the preset safety preconditions. These safety preconditions may include that the vehicle is stationary, the engine is turned off, the doors are locked, etc. Various vehicle status data are collected in real time through various sensors of the vehicle, such as speed sensors, engine status sensors, door lock status sensors, etc. The collected status data is compared with the preset safety conditions. If all conditions are met, it is determined that the current state of the vehicle meets the safety preconditions. The purpose of this is to ensure that the vehicle is in a safe and controllable state and to prevent subsequent operations under unsafe conditions, thereby improving the safety and reliability of the entire system.

[0073] Next, determine whether the vehicle Bluetooth module of the target vehicle is in a Bluetooth connection state with any target terminal. The vehicle Bluetooth module will continue to broadcast its existence and scan the surrounding area for paired target terminals. When the target terminal is detected to be within range, the vehicle Bluetooth module will try to establish a connection with it. After the connection is established, the vehicle Bluetooth module will maintain a list of connected devices, recording all target terminals currently connected to it. During the judgment process, you only need to check whether this list is empty. If there is at least one target terminal in the list, it is determined to be in a Bluetooth connection state. Target terminals refer to mobile devices that have passed the pre-set two-way authentication process, such as smartphones, smart watches, vehicle keys, etc. Two-way authentication involves steps such as exchanging encryption keys between devices and verifying unique device identifiers to ensure the security of the connection and the legitimacy of the device.

[0074] After confirming that the vehicle's Bluetooth module is connected to at least one target terminal, the system enters the continuous monitoring stage. In this stage, the camera and various sensors of the target vehicle are activated to collect the target user's mobile behavior data within the preset detection range. The detection range is a certain radius centered on the vehicle. The camera covers a 360-degree area around the vehicle and captures images and videos in real time. At the same time, the target vehicle is equipped with a variety of sensors, such as infrared sensors, ultrasonic sensors, laser radars, etc., which are used to detect human body heat, distance changes and other information. Then the data will be analyzed in real time to determine whether the user has entered the preset recognition area and whether the predefined first action has been implemented. At the same time, the system will record the time when the user enters the detection range. If the user leaves the detection range without performing effective operations within the preset time, the system will stop data collection.

[0075] When the system detects that the target user performs the first action in the preset recognition area, the mobile behavior data collected continuously before will be processed and stored. These data are organized into a complete data set in timestamp order, called the first mobile behavior data set. The accuracy of the timestamp reaches the millisecond level to ensure the time series accuracy of the data. The data set contains multi-dimensional information, such as the user's three-dimensional coordinate changes, movement speed, acceleration, body posture changes, etc. Each data record is associated with its corresponding timestamp to form a time series. This data set comprehensively records the entire process of the user from entering the detection range to performing the first action, providing rich raw data for subsequent behavior analysis. The data is stored in an efficient time series database, such as InfluxDB or TimescaleDB, to optimize the storage and query efficiency of large amounts of time series data.

[0076] Next, the system will perform a time span analysis on the first mobile behavior data set. First, the difference between the earliest and latest timestamps in the data set is calculated to obtain the total time span of the data. This time span is then compared with the preset time length. The preset time length is a reasonable value based on a large amount of user behavior data statistics. If the time span of the data set is less than or equal to the preset time length, the entire data set will be directly used as a data segment for subsequent analysis. In this case, the data set completely records all the user's behaviors within the preset time, and no interception is required. However, if the time span of the first mobile behavior data set is greater than the preset time length, data interception is required. The interception method is to start from the end of the data set, that is, the latest data, and intercept a data segment equal to the preset time length. The reason for this interception method is that the data closest to the first action usually contains the most relevant and valuable information. For example, if the preset time length is 10 seconds, and the user performs the first action 15 seconds after entering the detection range, the system will intercept the last 10 seconds of data as the data segment for analysis. This interception method ensures that the data used for analysis always has a consistent time length, which is conducive to the standardized processing of subsequent pattern recognition and feature extraction algorithms. At the same time, it also balances data integrity and computational efficiency, providing a guarantee for the real-time response of the system.

[0077] In one embodiment, if it is detected that the target user has been away from the detection range for a time period exceeding a preset time period, the continuously collected movement behavior data is stored in a timestamp order as a second movement behavior data set;

[0078] A first label is marked for the second movement behavior data set, and a tailgate control prediction model is trained using the second movement behavior data set marked with the first label, wherein the first label indicates that a behavior prediction result is a non-control tailgate behavior.

[0079] In this embodiment, the user leaving the detection range means that all these sensors can no longer detect the user's presence. When it is detected that the target user has left the detection range for longer than the preset time, all the mobile behavior data collected from the time the user enters the detection range to the last time the user's presence is detected will be processed. These data include multi-dimensional information such as the user's movement trajectory, speed change, posture change, etc. The system will sort and organize these data in the order of timestamps to form a complete time series data set, called the second mobile behavior data set. The second mobile behavior data set can be described by the following mathematical expression:

[0080] ;

[0081] in, represents the second mobile behavior dataset, is the timestamp, Is the user at time The three-dimensional coordinates of It's speed. is the acceleration, It is the body orientation. is the time when the user enters the detection range, Is the last time the user was detected.

[0082] After the second mobile behavior dataset is stored, the next step is to label the dataset with the first label. The first label here represents "non-controlled tailgate behavior", which means that this set of data reflects the behavior pattern of the user approaching the vehicle but ultimately not triggering the tailgate to open. The labeling process actually adds an additional attribute field to each data record to indicate the behavior category to which this data belongs. After labeling is completed, these labeled data will be used to train the tailgate control prediction model.

[0083] The training process can be described by the following objective function:

[0084] ;

[0085] in, are model parameters, is the model function, are input features, is the true label, is the loss function, is the sample size, is the regularization term, is the regularization coefficient.

[0086] The effect of this training method is that the model can learn the typical behavior patterns of users when they do not intend to operate the tailgate. These patterns may include the characteristics of the user's movement trajectory around the vehicle, speed change characteristics, and residence time characteristics. Through training with a large amount of such data, the model can gradually improve the recognition accuracy of non-controlled tailgate behavior, so as to better distinguish whether the user intends to open the tailgate in actual applications. This not only improves the intelligence of the system, but also optimizes the user experience and reduces the occurrence of misoperation.

[0087] In one embodiment, referring to Figure 2 , after controlling the tailgate of the target vehicle to be opened or closed according to the first control instruction, the method further includes the following steps:

[0088] S201. During the opening or closing process of the tailgate of the target vehicle, if it is detected that the target user performs a second action in the projection area where the first projection pattern is projected, the first control instruction is terminated, and a second control instruction is generated according to the second action;

[0089] S202. Controlling the tailgate of the target vehicle to open or close according to the second control instruction, the second control instruction controlling the tailgate of the target vehicle is opposite to the first control instruction;

[0090] S203. Mark the behavior prediction result with a second label, and train a tailgate control prediction model in combination with the first mobile behavior dataset and the behavior prediction result marked with the second label, wherein the second label indicates that the behavior prediction result is predicted incorrectly.

[0091] In this embodiment, if the target user is not detected to perform the second action in the projection area where the first projection pattern is projected during the opening or closing process of the tailgate of the target vehicle, the projection of the first projection pattern is terminated when the tailgate of the target vehicle is completely closed or opened.

[0092] A third label is marked for the behavior prediction result, and a tailgate control prediction model is trained by combining the first mobile behavior data set and the behavior prediction result marked with the third label, wherein the third label indicates that the behavior prediction result is predicted correctly.

[0093] In this embodiment, during the opening or closing process of the tailgate of the target vehicle, the system will continuously monitor the user behavior in the projection area where the first projection pattern is projected. Specifically, various sensors (such as cameras, infrared sensors, pressure sensors, etc.) can be used to monitor the user's actions in this area in real time. When it is detected that the target user performs a second action in this area, the system will respond immediately. The second action is a predefined footstep action (such as kicking, stepping action) or staying in the area for more than a certain time. Once the second action is detected, the system will immediately terminate the first control instruction currently being executed. For example, if the tailgate is opening, the system will immediately send a stop signal to the tailgate controller to interrupt the operation of the motor. Then, the system will generate a new control instruction, namely the second control instruction, based on the identified second action. The execution effect of the second control instruction is to make the movement direction of the tailgate opposite to the first control instruction. Specifically, if the first control instruction is to open the tailgate, then the second control instruction will cause the tailgate to start closing; conversely, if the first control instruction is to close the tailgate, then the second control instruction will cause the tailgate to start opening.

[0094] After executing the second control instruction, the system will evaluate and record the interaction process. Specifically, the system will mark the second label for this behavior prediction result. The second label indicates that the behavior prediction result is wrong, which means that the system's initial prediction (based on the first control instruction) is inconsistent with the user's actual intention (expressed by the second action). This labeling process is actually part of a self-correction and learning mechanism. Next, the system will use these data marked with the second label to update and train the tailgate control prediction model. This process is actually a method of online learning or incremental learning. The system will add the newly labeled data to the training data set and then retrain or fine-tune the existing prediction model.

[0095] The training process can be described by the following objective function:

[0096] ;

[0097] The first two items are the same as described above, and the third Indicates the newly added sample marked as a prediction error. is a weight coefficient used to adjust the degree of influence of new samples on model updates. The effect of this training method enables the model to continuously learn and improve from its mistakes. By incorporating cases with incorrect predictions into the training data, the model can gradually adjust its decision boundary and improve its ability to recognize similar situations. This continuous learning and self-correction mechanism enables the system to adapt to changes in user behavior patterns and continuously improve prediction accuracy, thereby providing increasingly accurate services in long-term use.

[0098] If the target user is not detected to perform the second action in the projection area where the first projection pattern is projected during the opening or closing process of the target vehicle's tailgate, the system will continue to execute the original tailgate control operation until it is completed. During this process, the system will continue to monitor the projection area and be ready to respond to possible user intervention at any time. The movement of the tailgate may be driven by an electric motor, and the position of the tailgate is tracked by a precise position sensor (such as a Hall sensor or a photoelectric encoder). When the tailgate reaches the predetermined end position (fully open or fully closed), the system will receive a position signal, which can come from a limit switch on the tailgate mechanism. Once it is confirmed that the tailgate has been fully opened or closed, the system will immediately send a command to terminate the projection of the first projection pattern. Termination of projection marks the end of a complete tailgate control interaction process. This not only saves energy, but also avoids unnecessary visual interference, improving the overall efficiency of the system. At the same time, it also provides a clear feedback to the user that the system has completed the intended operation and no further user input is required.

[0099] In this case, the system will label the behavior prediction result with a third label. The third label indicates that the behavior prediction result is correct, which means that the system's initial prediction based on the user's mobile behavior is consistent with what actually happened. This labeling process is part of the system's self-evaluation and continuous learning mechanism, and is critical to improving the system's long-term performance. Next, the system uses this data labeled with the third label to update and train the tailgate control prediction model. This process is a reinforcement learning method that aims to consolidate and optimize the model's ability to identify correct cases. The system adds the newly labeled data to the training dataset and then uses this data to fine-tune the existing prediction model.

[0100] The training process can be described by the following objective function:

[0101] ;

[0102] The first two items are the same as described above, and the third Indicates that the newly added mark is a sample predicted correctly. Is a weight coefficient used to adjust the influence of correctly predicted samples on model updating.

[0103] In one embodiment, the vehicle electric tailgate unlocking method further includes the following steps:

[0104] If the behavior prediction result is a non-controlled tailgate behavior, projecting a second projection pattern in the projection area based on a preset projection time;

[0105] If it is detected within the projection time that the target user performs a third action in the projection area projected with the second projection pattern, a first control instruction is generated according to the third action, and the tailgate of the target vehicle is controlled to be opened or closed according to the first control instruction;

[0106] A second label is marked for the behavior prediction result, and a tailgate control prediction model is trained by combining the first mobile behavior data set and the behavior prediction result marked with the second label, wherein the second label indicates that the behavior prediction result is predicted incorrectly.

[0107] In this embodiment, if the target user is not detected performing the third action in the projection area where the second projection pattern is projected within the projection time, a third label is marked for the behavior prediction result, and the tailgate control prediction model is trained in combination with the first mobile behavior data set and the behavior prediction result marked with the third label. The third label indicates that the behavior prediction result is predicted correctly.

[0108] In this embodiment, when the behavior prediction result is a non-control tailgate behavior, it means that the user's behavior is not intended to control the tailgate. The system will start a preset projection time timer and project a second projection pattern in the projection area. This second projection pattern is a more eye-catching or interactive image, such as a dynamic arrow indicator or a graphical interface with text prompts. The projection time is usually set to a moderate length, such as 5-10 seconds, to balance the user experience and system efficiency. This time length may be dynamically adjusted according to user habits and environmental conditions. For example, in a dimly lit environment, the projection time may be slightly extended to ensure that the user has enough time to notice and respond to the projection. The size and shape of the projection area are dynamically adjusted according to the vehicle model and the surrounding environment. The system uses environmental perception sensors (such as lidar or 3D cameras) to adjust the projection area in real time to avoid projecting onto other vehicles or pedestrians. The purpose of projecting the second projection pattern is to provide users with a clear interactive interface to guide users to perform tailgate control operations. The effect of this method is to increase the initiative of the system, and even in the case of an initial prediction of a non-control tailgate behavior, it still provides users with convenient interaction opportunities, thereby improving the flexibility and user-friendliness of the system.

[0109] During the preset projection time, the system will continuously monitor the third action that the target user may perform in the projection area where the second projection pattern is projected. The third action may be a series of predefined footsteps or body postures, such as making specific footsteps in the projection area, or staying in a specific position for a certain period of time. Once the third action is detected, the system will immediately generate the corresponding first control instruction according to the predefined mapping relationship. The generated first control instruction will be immediately sent to the tailgate control unit of the vehicle to trigger the opening or closing operation of the tailgate.

[0110] When the system detects that the target user performs the third action in the projection area where the second projection pattern is projected, and successfully generates and executes the first control instruction based on this action, it will immediately start the evaluation and marking process of the behavior prediction result. In this case, the system will mark the behavior prediction result with a second label, indicating that the initial behavior prediction result is wrong. After the marking is completed, the system will immediately start the model training process. This process combines the first mobile behavior data set and the newly labeled data to update and optimize the tailgate control prediction model. The training process adopts online learning or incremental learning methods, using optimization algorithms such as stochastic gradient descent (SGD) or adaptive moment estimation (Adam), and the training objective can be expressed as minimizing the prediction error.

[0111] If the system fails to detect that the target user performs the third action in the projection area where the second projection pattern is projected within the preset projection time, the system will enter a different processing flow. First, the system will confirm that the projection time has ended according to the projection time timer. Once the projection time is confirmed to have ended, the system will immediately terminate the projection of the second projection pattern. Next, the system will start the evaluation and marking process of the behavior prediction results. In this case, the system will mark the behavior prediction result with a third label, indicating that the initial behavior prediction result is correct. After the marking is completed, the system will immediately start the model training process. This process combines the first mobile behavior data set and the newly labeled data to update and optimize the tailgate control prediction model. The above process can realize the dynamic learning and self-verification mechanism of the tailgate control prediction model. The tailgate control prediction model can continuously learn and strengthen from the correct predictions, and gradually improve its understanding and prediction capabilities of user behavior patterns. In the long run, this method can significantly improve the prediction accuracy and stability of the system, making the tailgate control function more accurate and reliable, thereby improving the user experience and the overall performance of the system.

[0112] The present invention also discloses a vehicle electric tailgate unlocking system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the vehicle electric tailgate unlocking method described in any one of the above embodiments is implemented.

[0113] Among them, the processor can adopt a central processing unit (CPU). Of course, according to actual usage, other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. can also be adopted. The general-purpose processor can adopt a microprocessor or any conventional processor, etc., and this application does not impose any restrictions on this.

[0114] Among them, the memory can be an internal storage unit of a computer device, such as a hard disk or memory of a computer device, or an external storage device of a computer device, such as a plug-in hard disk, a smart memory card (SMC), a secure digital card (SD) or a flash memory card (FC) equipped on the computer device, and the memory can also be a combination of an internal storage unit and an external storage device of a computer device. The memory is used to store computer programs and other programs and data required by the computer device. The memory can also be used to temporarily store data that has been output or is to be output, and this application does not impose any restrictions on this.

[0115] The present invention also discloses a computer-readable storage medium, on which instructions are stored. When the instructions are executed by a processor, the processor is configured to execute the production control method of the UV platemaking machine described in any one of the above embodiments.

[0116] Among them, the computer program can be stored in a machine-readable medium, the computer program includes computer program code, the computer program code can be in the form of source code, object code, executable file or certain middleware, etc. The machine-readable medium includes any entity or device that can carry the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the machine-readable medium includes but is not limited to the above-mentioned components.

[0117] Among them, through this computer-readable storage medium, the comprehensive fault detection method for transmission lines in the above embodiment is stored in a computer-readable storage medium, and is loaded and executed on a processor to facilitate the storage and application of the above method.

[0118] A person skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of protection of the present application is limited to these examples. In line with the concept of the present application, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of different aspects of one or more embodiments of the present application as above, which are not provided in detail for the sake of simplicity.

[0119] One or more embodiments of the present application are intended to cover all such substitutions, modifications and variations that fall within the broad scope of the present application. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of one or more embodiments of the present application should be included in the protection scope of the present application.

Claims

1. A method for unlocking a vehicle electric tailgate, characterized in that: The steps include: When it is determined that the current state of the target vehicle meets the preset safety precondition, it is determined whether the vehicle Bluetooth module of the target vehicle is in a Bluetooth connection state with any target terminal, and the target terminal is a mobile terminal that has been pre-bidirectionally authenticated with the vehicle Bluetooth module; If the vehicle's Bluetooth module is in Bluetooth connection with any target terminal, when the target user enters the preset detection range of the target vehicle, the target vehicle's camera and sensor will continuously collect the target user's movement behavior data within the detection range; After detecting that the target user performs a first action in a preset identification area of ​​the target vehicle, obtaining a segment of movement behavior data of the target user within a preset time length before performing the first action, if it is detected that the target user performs the first action in the preset identification area of ​​the target vehicle, storing the continuously collected movement behavior data in timestamp order as a first movement behavior data set; If the data time span of the first mobile behavior data set is less than or equal to the preset time length, the first mobile behavior data set is used as a mobile behavior data segment of the target user before performing the first action and within the time length; If the data time span of the first mobile behavior data set is greater than the time length, a mobile behavior data segment is intercepted from the end of the first mobile behavior data set based on the time length, and the mobile behavior data in the mobile behavior data segment includes the moving speed, moving direction, waiting position and waiting time of the target user; Input the mobile behavior data into a preset tailgate control prediction model, and output the behavior prediction results of the target user through the tailgate control prediction model; If it is detected that the target user has been away from the detection range for a time period exceeding a preset time period, the continuously collected movement behavior data is stored in a timestamp order as a second movement behavior data set; marking a first label for the second movement behavior data set, and using the second movement behavior data set marked with the first label to train a tailgate control prediction model, wherein the first label indicates that the behavior prediction result is a non-control tailgate behavior; If the behavior prediction result is to control the tailgate behavior, projecting a first projection pattern in a preset projection area, and generating a first control instruction according to the behavior prediction result after projecting the first projection pattern; Controlling the tailgate of the target vehicle to open or close according to the first control instruction, and during the opening or closing process of the tailgate of the target vehicle, if it is detected that the target user performs a second action in the projection area where the first projection pattern is projected, terminating the first control instruction and generating a second control instruction according to the second action; Controlling the tailgate of the target vehicle to open or close according to a second control instruction, wherein the second control instruction controls the action of the tailgate of the target vehicle opposite to that of the first control instruction; A second label is marked for the behavior prediction result, and a tailgate control prediction model is trained by combining the first mobile behavior data set and the behavior prediction result marked with the second label, wherein the second label indicates that the behavior prediction result is predicted incorrectly.

2. The vehicle electric tailgate unlocking method according to claim 1, characterized in that: The method further comprises the steps of: If the target user is not detected to perform the second action in the projection area where the first projection pattern is projected during the opening or closing process of the tailgate of the target vehicle, then when the tailgate of the target vehicle is completely closed or completely opened, the projection of the first projection pattern is terminated; A third label is marked for the behavior prediction result, and a tailgate control prediction model is trained by combining the first mobile behavior data set and the behavior prediction result marked with the third label, wherein the third label indicates that the behavior prediction result is predicted correctly.

3. The vehicle electric tailgate unlocking method according to claim 1, characterized in that: The method further comprises the steps of: If the behavior prediction result is a non-controlled tailgate behavior, projecting a second projection pattern in the projection area based on a preset projection time; If it is detected within the projection time that the target user performs a third action in the projection area projected with the second projection pattern, a first control instruction is generated according to the third action, and the tailgate of the target vehicle is controlled to be opened or closed according to the first control instruction; A second label is marked for the behavior prediction result, and a tailgate control prediction model is trained by combining the first mobile behavior data set and the behavior prediction result marked with the second label, wherein the second label indicates that the behavior prediction result is predicted incorrectly.

4. The vehicle electric tailgate unlocking method according to claim 3, characterized in that: The method further comprises the steps of: If the target user is not detected performing the third action in the projection area where the second projection pattern is projected within the projection time, a third label is marked for the behavior prediction result, and the tailgate control prediction model is trained in combination with the first mobile behavior data set and the behavior prediction result marked with the third label. The third label indicates that the behavior prediction result is predicted correctly.

5. A vehicle electric tailgate unlocking system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the vehicle electric tailgate unlocking method as described in any one of claims 1 to 4 is implemented.

6. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instruction is executed by a processor, the processor is configured to execute the vehicle electric tailgate unlocking method according to any one of claims 1 to 4.

Images