Intelligent door lock door opening control method and system and intelligent door lock

Abstract

The invention is suitable for the technical field of intelligent door locks, and provides an intelligent door lock door opening control method and system and an intelligent door lock.The method comprises the steps that sensor data which is obtained by a sensor deployed on the intelligent door lock or around the intelligent door lock and reflects the relative state of a user and the intelligent door lock are collected; receiving user equipment data which is sent by the authorized user equipment and comprises an equipment identifier and communication signal strength; inputting the sensor data and the user equipment data into a preset intention analysis model, and calculating an intention value through the intention analysis model; comparing the calculated intention value with at least one preset intention threshold value, and selecting an authentication strategy corresponding to the intention value from a preset authentication strategy set based on a comparison result; starting a user identity verification process specified by the selected authentication strategy; and when verification succeeds, the intelligent door lock is driven to execute unlocking operation. The intelligent door lock solves the problem that an existing intelligent door lock is insufficient in intelligence.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent door locks, and particularly to an intelligent door lock opening control method, system and intelligent door lock. Background Art

[0002] With the popularization of Internet of Things technology and the concept of smart home, intelligent door locks, as important entry-level products of smart home, have been more and more widely used. Compared with traditional mechanical door locks, intelligent door locks provide diversified unlocking methods, significantly improving user convenience. Currently, the mainstream unlocking methods of intelligent door locks on the market include but are not limited to: digital password input, fingerprint recognition, NFC (Near Field Communication) card or tag induction, and Bluetooth near field induction, Wi-Fi remote control or dedicated APP operation through user intelligent devices (such as smart phones, smart watches), etc.

[0003] Although these methods provide certain convenience, there are still some inherent problems:

[0004] Fixed authentication logic: Most intelligent door locks adopt a fixed authentication process. Regardless of the user's state or scenario, the same preset authentication steps (such as having to press fingerprints, etc.) need to be executed. This "one-size-fits-all" method lacks flexibility and can cause inconvenience in certain situations (such as when the user is carrying heavy objects with both hands, it is inconvenient to operate the mobile phone in rainy days, etc.).

[0005] Inherent contradiction between security and convenience: In order to pursue convenience, some solutions (such as near field unlocking based only on Bluetooth signal strength) may sacrifice some security and are prone to false unlocking due to signal fluctuations or the user just passing by the door. On the contrary, overemphasizing security (such as forcing multiple authentications) will reduce the convenience of daily use for users. How to dynamically balance security and convenience in different scenarios is a challenge.

[0006] Lack of judgment on the user's true intention: Existing methods mainly rely on the user's active trigger (pressing, swiping the card, APP operation) or simple physical conditions (such as device proximity), and cannot effectively judge whether the user's current behavior really has the intention to open the door. The user may just stay briefly at the door, talk to someone or pass by, not intending to enter immediately, but the mechanism triggered based on distance or simple sensors may prematurely activate the authentication process or even attempt to unlock, causing interference or potential risks.

[0007] Single information source and vulnerable to interference: Some methods that rely on a single information source (such as only relying on the mobile phone Bluetooth RSSI value to judge distance and intention) are easily affected by environmental interference, signal occlusion and other factors, resulting in inaccurate judgment and unstable user experience. Summary of the Invention

[0008] Based on this, the objective of the present invention is to provide an intelligent door lock opening control method, system, and intelligent door lock, so as to fundamentally solve the problem of insufficient intelligence of existing intelligent door locks.

[0009] An intelligent door lock opening control method according to an embodiment of the present invention, the method includes:

[0010] Collect at least one type of sensor data reflecting the relative state between the user and the intelligent door lock obtained by sensors deployed on or around the intelligent door lock;

[0011] Receive user device data including a device identifier and a communication signal strength sent by an authorized user device;

[0012] Input the collected sensor data and the received user device data into a preset intention analysis model, and calculate an intention value representing the current degree of the user's willingness to open the door through the intention analysis model;

[0013] Compare the calculated intention value with at least one preset intention threshold, and based on the comparison result, select an authentication strategy corresponding to the intention value from a set of preset authentication strategies including at least two types with different interaction methods or security levels;

[0014] Start the user identity authentication process specified by the selected authentication strategy to authenticate the user;

[0015] When the user identity authentication process is successfully verified, drive the intelligent door lock to perform an unlocking operation.

[0016] In addition, an intelligent door lock opening control method according to the above embodiment of the present invention may further have the following additional technical features:

[0017] Further, the step of collecting at least one type of sensor data reflecting the relative state between the user and the intelligent door lock obtained by sensors deployed on or around the intelligent door lock includes:

[0018] Continuously monitor a predetermined area in front of the intelligent door lock using at least one low-power passive sensor;

[0019] When the passive sensor detects that a moving object enters the predetermined area, generate and send a trigger signal for indicating the start of subsequent data collection;

[0020] In response to the trigger signal, activate at least one active sensor;

[0021] When the activated active sensor is a ranging sensor, the ranging sensor is used to periodically measure and obtain the distance information and / or relative motion information between the target object and the intelligent door lock, and the distance information and / or relative motion information is used as part of the sensor data;

[0022] When the activated active sensor is a camera or a structured light sensor, the camera or the structured light sensor is used to capture an image or depth information within the predetermined area, and the captured image or depth information is processed to extract preliminary feature information, and the preliminary feature information is used as part of the sensor data.

[0023] Further, the step of processing the captured image or depth information to extract preliminary feature information includes:

[0024] Performing an object detection algorithm on the captured image or depth information to identify and locate the human object therein;

[0025] According to the identified human object, calculating and extracting target parameters as the preliminary feature information, where the target parameters include the bounding box coordinates, centroid position, motion speed vector, orientation angle, or relative distance estimate to the intelligent door lock of the human object;

[0026] Applying a pose estimation algorithm to the human object, extracting the pose information of the skeletal joint points, comparing the pose information with a preset pose template library, obtaining a judgment result on whether the user is in a specific pose, and including the judgment result in the preliminary feature information.

[0027] Further, the step of receiving user equipment data including a device identifier and a communication signal strength sent by an authorized user equipment includes:

[0028] Periodically scanning for surrounding wireless devices by using a wireless communication module built in the intelligent door lock;

[0029] Extracting the device identifier of the user equipment from the signals captured during the scanning process;

[0030] Comparing the extracted device identifier with a pre-stored list of authorized device identifiers;

[0031] When a successfully matched authorized device identifier is found in the comparison, confirming that the successfully matched user equipment is an authorized user equipment and recording the device identifier;

[0032] Measuring and obtaining the signal strength value received when communicating with the authorized user equipment by the wireless communication module;

[0033] Combining the device identifier of the confirmed and recorded authorized user equipment and the obtained signal strength value to form user equipment data.

[0034] Further, the step of inputting the collected sensor data and the received user device data into a preset intention analysis model and calculating, through the intention analysis model, an intention value representing the current degree of the user's willingness to open the door includes:

[0035] Integrating the collected sensor data and the received user device data into an original feature set;

[0036] Performing a preprocessing operation on the original feature set, where the preprocessing operation includes time series alignment and feature value normalization;

[0037] Inputting the preprocessed feature set into a preset intention analysis model;

[0038] Calculating and processing the input multi-dimensional feature vector by the intention analysis model, and outputting a value representing the probability or confidence of the user's door opening intention as the intention value.

[0039] Further, the step of comparing the calculated intention value with at least one preset intention threshold and, based on the comparison result, selecting, from a preset authentication policy set including at least two preset authentication policies with different interaction methods or security levels, an authentication policy corresponding to the intention value includes:

[0040] Mathematically comparing the calculated intention value sequentially or in parallel with one or more preset intention thresholds stored in the intelligent door lock;

[0041] Determining the intention level interval to which the intention value belongs according to the comparison result;

[0042] Searching in the preset authentication policy set according to the determined intention level interval and setting the authentication policy uniquely corresponding to the determined intention level interval as the authentication policy in the current scenario.

[0043] Further, the step of starting the user identity authentication process specified by the selected authentication policy includes:

[0044] Determining the door lock authentication hardware module to be activated according to the selected authentication policy, and sending an activation instruction to the determined door lock authentication hardware module;

[0045] Driving the user interface to output operation guidance information corresponding to the authentication policy;

[0046] Controlling the intelligent door lock to enter a state of waiting for the user to provide authentication information through the activated authentication hardware module.

[0047] Further, the method further includes:

[0048] Record relevant data during each execution of the method, where the relevant data at least includes the sensor data, the user device data, the intention value, the selected authentication policy, and the result of the user authentication process;

[0049] Utilize the recorded relevant data to periodically or when a preset trigger condition is met, update the internal parameters of the intention analysis model or adjust the preset intention threshold.

[0050] Another object of an embodiment of the present invention is to provide an intelligent door lock opening control system, the system includes:

[0051] A sensor data acquisition module, configured to acquire at least one type of sensor data reflecting the relative state between the user and the intelligent door lock obtained by sensors deployed on or around the intelligent door lock;

[0052] A user device data receiving module, configured to receive user device data including a device identifier and / or communication signal strength sent by an authorized user device;

[0053] An intention value calculation module, configured to input the acquired sensor data and the received user device data into a preset intention analysis model, and calculate an intention value representing the current degree of the user's willingness to open the door through the intention analysis model;

[0054] An authentication policy selection module, configured to compare the calculated intention value with at least one preset intention threshold, and based on the comparison result, select an authentication policy corresponding to the intention value from a set including at least two preset authentication policies with different interaction methods or security levels;

[0055] An authentication module, configured to initiate a user authentication process specified by the selected authentication policy to authenticate the user;

[0056] An unlocking module, configured to drive the intelligent door lock to perform an unlocking operation when the authentication module successfully verifies the user authentication process.

[0057] Another object of an embodiment of the present invention is to provide an intelligent door lock, including a memory, a processor, and a program stored on the memory and executable on the processor, where when the processor executes the program, it implements the intelligent door lock opening control method as described above.

[0058] The intelligent door lock opening control method provided by the embodiment of the present invention can more accurately understand the user's current true intention by fusing multi-modal data from sensors at the intelligent door lock end and user devices, and calculating the user's door opening intention value using an intention analysis model. It avoids unnecessary disturbances or misoperations when the user has no intention of opening the door (such as passing by or staying), making the response of the intelligent door lock more intelligent and user-friendly. According to the quantified user door opening intention value, different authentication strategies are adaptively selected. When the system determines that the user has a strong intention to open the door, a more convenient authentication method can be selected, greatly improving the convenience for the user when it is inconvenient to operate (such as carrying heavy objects). When the user's intention is unclear or low, a more secure standard authentication strategy is selected or maintained to ensure the security of the door lock. This dynamic balance mechanism effectively solves the problem in the prior art that it is difficult to balance security and convenience. At the same time, sensor data reflecting the user's physical state (such as distance, speed, posture) and user device state data (such as device identifier, signal strength) are utilized, with a richer information dimension, making the intention judgment result more reliable, less susceptible to interference from a single factor, and improving the stability and robustness of the control method. It solves the problem of insufficient intelligence of existing intelligent door locks. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a flowchart of the intelligent door lock opening control method in the first embodiment of the present invention;

[0060] Figure 2 It is a structural diagram of the intelligent door lock opening control system in the second embodiment of the present invention;

[0061] Figure 3 It is a structural diagram of the intelligent door lock in the third embodiment of the present invention;

[0062] The following specific embodiments will further illustrate the present invention in conjunction with the above-mentioned drawings. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0063] For ease of understanding of the present invention, the present invention will be described more comprehensively below with reference to the relevant drawings. Several embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, these embodiments are provided to make the disclosure of the present invention more thorough and comprehensive.

[0064] It should be noted that when an element is referred to as being "fixed to" another element, it can be directly on the other element or there can also be an intermediate element. When an element is considered to be "connected" to another element, it can be directly connected to the other element or there may be an intermediate element at the same time. The terms "vertical", "horizontal", "left", "right" and similar expressions used herein are for illustrative purposes only.

[0065] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which this invention belongs. The terms used in the description of this invention herein are for the purpose of describing specific embodiments only and are not intended to limit the invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0066] Embodiment 1

[0067] Please refer to Figure 1 , which shows the intelligent door lock opening control method in the first embodiment of the present invention. For the convenience of description, only the parts related to the embodiments of the present invention are shown. The intelligent door lock opening control method provided by the embodiments of the present invention includes:

[0068] Step S10, collecting at least one type of sensor data obtained by sensors deployed on or around the intelligent door lock, which reflects the relative state between the user and the intelligent door lock;

[0069] Among them, in one embodiment of the present invention, the step of collecting at least one type of sensor data obtained by sensors deployed on or around the intelligent door lock, which reflects the relative state between the user and the intelligent door lock, includes:

[0070] Continuously monitoring a predetermined area in front of the intelligent door lock by using at least one low-power passive sensor;

[0071] When the passive sensor detects that a moving object enters the predetermined area, generating and sending a trigger signal for indicating the start of subsequent data collection;

[0072] In response to the trigger signal, activating at least one active sensor;

[0073] When the activated active sensor is a ranging sensor, using the ranging sensor to periodically measure and obtain the distance information and / or relative motion information between the target object and the intelligent door lock, and taking the distance information and / or relative motion information as part of the sensor data;

[0074] When the activated active sensor is a camera or a structured light sensor, using the camera or the structured light sensor to capture the image or depth information in the predetermined area, processing the captured image or depth information to extract preliminary feature information, and taking the preliminary feature information as part of the sensor data.

[0075] Specifically, first, a low-power passive infrared (PIR) sensor or microwave sensor is used to continuously monitor a predetermined area in front of the door lock (for example, within a range of 1-3 meters), where the passive infrared sensor is configured to continuously operate in a low-power mode (or periodic wake-up monitoring mode). When the infrared sensor detects a moving object (a potential user approaching) in this area, it is determined that an object has entered, and at this time, an internal trigger signal is generated and sent. The trigger signal activates one or more active sensors on the smart door lock.

[0076] If the activated sensor is a ranging sensor, such as activating a millimeter-wave radar, a time-of-flight (ToF) sensor, or an ultrasonic ranging sensor. Then the active sensor subsequently starts to measure periodically (for example, every 100 milliseconds) (such as emitting electromagnetic waves / sound waves / light pulses and receiving echoes / reflections) and obtains the precise distance information between the front target object and the door lock, as well as the relative motion speed information calculated through the Doppler effect or continuous ranging changes. At this time, these distance and speed information constitute a part of the sensor data.

[0077] If the activated active sensor is a camera or a structured light sensor, for example, the smart door lock is equipped with a camera (such as a wide-angle low-resolution camera or a structured light camera), then the trigger signal also activates the camera, and at this time, the camera captures an image frame (pixel matrix) or a depth map (matrix of depth values for each pixel point) of the area in front of the door. Since the original image / depth data is large in volume and redundant in information, it needs to be processed to extract useful preliminary feature information. Specifically, the steps of processing the captured image or depth information to extract preliminary feature information include:

[0078] Performing an object detection algorithm on the captured image or depth information to identify and locate the human target therein;

[0079] According to the identified human target, calculating and extracting target parameters as the preliminary feature information, where the target parameters include the bounding box coordinates, centroid position, motion speed vector, orientation angle, or relative distance estimate from the smart door lock of the human target;

[0080] Applying a pose estimation algorithm to the human target, extracting the pose information of the skeletal joint points, comparing the pose information with a preset pose template library, obtaining a judgment result on whether the user is in a specific pose, and including the judgment result in the preliminary feature information.

[0081] Specifically, for preliminary feature extraction, the processor of the intelligent door lock will apply an object detection algorithm to the captured image / depth information to identify and locate the human object. Specifically, it obtains the image data captured by the camera or the depth map data obtained by the structured light sensor. Then, necessary preprocessing is carried out, such as image denoising and size adjustment to meet the input requirements of subsequent algorithms. Then, a pre-trained object detection model is loaded. The object detection model is usually based on deep learning, such as YOLO (You Only Look Once), SSD (Single Shot MultiBox Detector), or FasterR-CNN, etc., and is trained to identify various object categories including "person". For depth data, a model specifically designed for depth maps can be used or the depth information can be encoded and input into a standard model. Then, the object detection model scans and identifies whether there is a human object based on the input image or depth data. If detected, the object detection model outputs the position information of the human object in the image / depth map, usually in the form of a bounding box, for example, giving the coordinates of the upper left vertex and the lower right vertex of the bounding box, or the upper left coordinates and the width and height of the box. At the same time, the object detection model also outputs the category of the recognition result and a confidence score.

[0082] Furthermore, based on the identified human object, target parameters are calculated and extracted. Specifically, first, the result output by the object detection algorithm is directly used as the bounding box coordinates; then, the center point coordinates of the bounding box are calculated as the centroid position of the human object; then, the change in the centroid position of the same human object in consecutive multiple frames of images is tracked to calculate the motion speed vector. Specifically, first, a target tracking mechanism (such as based on Kalman filtering or simple IOU matching) is used to associate the detection boxes of the previous and current frames, and then the difference between the centroid positions of the current frame and the previous frame (or the average of several frames) is calculated and divided by the frame interval time to finally obtain the speed vector, where the motion speed vector represents the speed and direction of the human body's movement on the image plane; then, the calculation of the orientation angle is carried out. If a structured light sensor is used, the orientation can be estimated by analyzing the distribution of depth values within the human body area or fitting a three-dimensional model. If a camera is used, a pre-trained head pose estimation model can be used to estimate the orientation, or a rough inference can be made through the aspect ratio or motion direction of the human body bounding box when there is no dedicated model (for example, when a person is facing or back to the camera, the aspect ratio is larger; when side-facing, it is smaller). Then, the calculation of relative distance estimation is carried out. If a structured light sensor is used, the depth values of the pixel points within the human body bounding box are directly read, and the average or median value is taken as the distance estimation. If a camera is used, the camera calibration parameters (focal length, etc.) and the prior assumption of the average size of the human body are used to calculate the distance by measuring the height or width of its bounding box in the image. Then, these target parameters are used as preliminary feature information.

[0083] Further, for the image region of the recognized human target (i.e., the image content within the bounding box), or directly use the entire image for processing by the pose estimation algorithm. Specifically, load a pre-trained pose estimation algorithm model, which can recognize the human body in the image (or sometimes combined with depth information) and locate the two-dimensional or three-dimensional coordinates of its main skeletal joint points, such as key points like the head, neck, shoulders, elbows, wrists, hips, knees, and ankles. Then compare the extracted skeletal joint point coordinate information with a preset pose template library, where one or more "specific poses" related to the intention of opening the door or other types of specific poses are defined in the preset pose template library. For example, a pose template of "ready to open the door" can be defined, and its features may be: first, the coordinate of one-sided wrist (such as the right wrist) is located within a predetermined three-dimensional space area near the handle of the smart door lock; second, the positional relationship of this wrist relative to the same-side shoulder and hip conforms to the pattern of raising the arm and stretching forward; finally, the body trunk (such as the shoulder connection line) is generally oriented towards the door lock. At this time, parameter thresholds or reference coordinates describing these features are stored in the preset pose template library. Further, the comparison between the skeletal joint point coordinate information and the preset pose template library can specifically include: calculating the relative distance or angle between the currently extracted key joints (such as wrists, shoulders). Checking whether the key points (such as wrists) fall within the space areas defined by each pose template. Calculating the overall similarity score between the current pose and the reference poses of each pose template (such as using the cosine similarity of the key point vectors or the weighted sum of the joint angle differences). According to the comparison result, output a clear judgment result, where the judgment result can be a boolean value or a confidence score indicating the degree of matching. Finally, add the judgment result (boolean value or confidence score) obtained in the previous step to the set of target parameters (bounding box, speed, distance, etc.) calculated in the previous steps to form a complete "preliminary feature information" data packet, where the preliminary feature information contains key information such as the position, movement, distance, and whether in a specific (such as ready to open the door) pose of the human target extracted from the original image / depth information.

[0084] Finally, the distance / speed data obtained by the ranging sensor or the preliminary feature information (human body position, movement, orientation, optional pose judgment result, etc.) obtained by the camera or structured light sensor constitute the sensor data at this moment.

[0085] Step S20: Receive user equipment data including the equipment identifier and the communication signal strength sent by the authorized user equipment;

[0086] Among them, in an embodiment of the present invention, the step of receiving user equipment data including the equipment identifier and the communication signal strength sent by the authorized user equipment includes:

[0087] The built-in wireless communication module of the smart door lock periodically scans for surrounding wireless devices;

[0088] Extract the device identifier of the user device from the signals captured during the scanning process;

[0089] Compare the extracted device identifier with the pre-stored list of authorized device identifiers;

[0090] When a successful match with an authorized device identifier is found during the comparison, confirm that the successfully matched user device is an authorized user device and record the device identifier;

[0091] The wireless communication module measures and obtains the signal strength value received when communicating with the authorized user device;

[0092] Combine the device identifier of the confirmed and recorded authorized user device and the obtained signal strength value to form user device data.

[0093] Specifically, the smart door lock has a built-in wireless communication module, which is preferably a low-power Bluetooth (BLE module), but it could also be Wi-Fi or other short-range wireless communication technologies, which are set according to actual usage needs. In this invention, BLE is specifically used as an example for illustration. In parallel or after the sensor is triggered, the wireless communication module of the smart door lock executes the device discovery process. First, start the BLE scan and listen for signals broadcast by surrounding devices. Then parse the MAC address or other unique identifier of the device from the received broadcast signals. Compare this identifier with the list of authorized user device identifiers stored internally in the smart door lock (pre-set through the pairing process). If the comparison is successful, confirm that an authorized user device (such as the user's mobile phone or smartwatch) has been detected, and record the device identifier of this user device. At the same time, measure the signal strength value (RSSI) of the received user device. Finally, combine the confirmed authorized device identifier and the corresponding signal strength value to form user device data. If multiple authorized devices are scanned, the respective device identifiers and signal strength values can be recorded.

[0094] Among them, in other embodiments of the present invention, if the smart door lock has already established a connection with the user device (such as automatically connecting through proximity triggering before), the smart door lock can also actively request or passively receive additional data from the user device, such as real-time motion status and device status. The real-time motion status is the accelerometer / gyroscope readings of the user device (mobile phone or watch), which are used to determine whether the user is walking, standing, or performing a specific action. The device status is whether the mobile phone screen is lit and unlocked, and whether the specific pairing APP is running in the foreground. At this time, these additional data can be supplemented to the user device data.

[0095] Step S30: Input the collected sensor data and the received user device data into a preset intention analysis model, and calculate an intention value representing the current degree of the user's willingness to open the door through the intention analysis model;

[0096] Among them, in an embodiment of the present invention, it should be noted that the sensor data mainly reflects the status and behavior of the user or object in the physical near area of the intelligent door lock. This part of the data focuses on the physical reality and immediate behavior. The user device data is mainly used to confirm whether the person approaching the door lock is an authorized user and provide a device-based proximity status judgment (such as the Bluetooth signal strength RSSI). This part of the data focuses on identity confirmation and the associated status of the user device. When there is only sensor data, it is impossible to distinguish whether the approaching person is an authorized user, a stranger, a pet, or a passerby. Even if it is detected that someone is quickly approaching the door lock and making an opening action, if the identity cannot be confirmed by the intelligent door lock, it cannot be directly unlocked. When there is only user device data, it is impossible to distinguish whether the user's real intention is to open the door, or just because the user is moving in the area near the door at home (such as in the area near the door in the living room), or just passing by outside the door. The signal strength such as Bluetooth RSSI is greatly affected by the environment, and it is unreliable to solely rely on the user device data to judge the precise distance and intention. At this time, combining the sensor data and the user device data can form a more complete scenario portrait. For example, when it is determined according to the sensor data that someone is quickly approaching and according to the user device data that the communication signal strength is strong and the device identifier identity is confirmed, it can be determined that the user has a high intention to open the door. When it is determined according to the sensor data that someone is staying but without a specific action and according to the user device data that the communication signal strength is medium and the device identifier identity is confirmed, it can be determined that the user has a medium or low intention to open the door, which may be various other states such as the user waiting for someone. When it is determined according to the sensor data that there is no one or only a quick passing by and according to the user device data that the communication signal strength is low, it can be determined that there is no intention to open the door. When it is determined according to the sensor data that someone is approaching and according to the user device data that the communication signal strength is none or weak, it can be determined that the user is an unauthorized user or the user does not carry the device. At this time, it needs to be processed in the existing traditional manner or be vigilant.

[0097] Among them, since the original sensor data (such as a frame of pixels from a camera, point cloud from a radar) is usually high-dimensional, contains a large amount of redundancy and is not easily directly used for decision-making. At this time, it is very complex for the intention analysis model to directly process the original pixels to judge the "raising hand" gesture. The intention analysis model will be difficult to work effectively or requires an extremely complex structure to learn from the original data. The preliminary feature information is a meaningful and task-related refinement and simplification of the original data. For example, the human body bounding box, bone joint points, motion speed vector, orientation angle, etc. are extracted from the image. These are structured, low-dimensional, and higher information density representations. Therefore, the sensor data collected in the above step S10 is distance information and / or relative motion information or more specific preliminary feature information, enabling the intention analysis model to more easily learn to distinguish subtle behavior differences, thereby more accurately judging the user's true intention to calculate the final intention value. Therefore, by providing good preliminary feature information in the sensor data, the learning difficulty of the intention analysis model is greatly reduced. At this time, using pre-extracted features usually enables the model to be more accurate, require less training data, have a lighter model structure, and run more efficiently. At the same time, the processed features are usually more resistant to noise and irrelevant changes than the original data. For example, gesture judgment based on bone points is more resistant to illumination changes than direct appearance judgment based on pixels.

[0098] Further, the above step S30 specifically includes:

[0099] Integrate the collected sensor data and the received user device data into an original feature set;

[0100] Perform preprocessing operations on the original feature set, and the preprocessing operations include time series alignment and feature value normalization;

[0101] Input the preprocessed feature set into a preset intention analysis model;

[0102] The intention analysis model performs calculation processing on the input multi-dimensional feature vector and outputs a value representing the probability or confidence of the user's door-opening intention as the intention value.

[0103] Specifically, data from different sensors and user devices (such as distance, speed, human body position, posture judgment results, signal strength values, etc.) are integrated into an original feature set or a time series segment at a unified time point. Then, these features are preprocessed, specifically including time series alignment and feature value normalization. Time series alignment ensures that data from different sources correspond to the same time point or time window, and feature value normalization scales feature values with different units and ranges to similar numerical ranges (such as 0 to 1 or mean 0 and variance 1) to meet the model input requirements. Then, the preprocessed feature set (or called feature vector) is input into a preset intention analysis model. The intention analysis model is a pre-trained machine learning model, such as Support Vector Machine (SVM), Random Forest, Logistic Regression, Neural Network (NN), Gradient Boosting Tree, etc. These models have been trained with offline data and can learn the complex mapping relationship from input features to the door opening intention. At this time, the intention analysis model executes its internal calculation logic based on the input features and finally outputs a single numerical value - the intention value. The intention value is usually a probability or confidence score between 0 and 1, indicating the likelihood that the model judges the current user has a real door opening intention. For example, 0.9 indicates a very high door opening intention, and 0.1 indicates a very low door opening intention.

[0104] Furthermore, to improve accuracy, the intention analysis model can also consider the time context and the user's historical behavior during calculation. The time context is the current time (hour, day of the week), date (weekday / weekend), which can be used as additional input features for the intention analysis model. The user's historical behavior is based on the currently recognized user device data, retrieving the user's historical door opening patterns from the database (such as usually approaching the door quickly at 6 - 7 pm in the evening; going out slower on weekends in the morning, etc.), and using the similarity between the current behavior and the historical pattern as a feature to input into the model, or for dynamically adjusting the weights of other features.

[0105] Step S40, compare the calculated intention value with at least one preset intention threshold, and based on the comparison result, select an authentication strategy corresponding to the intention value from a preset set of authentication strategies that includes at least two types with different interaction methods or security levels;

[0106] Among them, in an embodiment of the present invention, the above steps specifically include:

[0107] Mathematically compare the calculated intention value with one or more preset intention thresholds stored in the intelligent door lock sequentially or in parallel;

[0108] According to the comparison result, determine the intention level interval to which the intention value belongs;

[0109] According to the determined intention level interval, search for and set the authentication policy uniquely corresponding to the determined intention level interval from the set of preset authentication policies as the authentication policy in the current scenario.

[0110] Specifically, the intelligent door lock reads one or more (usually at least two) preset intention thresholds from its storage, such as a low intention threshold (e.g., 0.3) and a high intention threshold (e.g., 0.8), etc. Then, compare the calculated intention value with these intention thresholds. Then, according to the comparison result, determine the intention level interval to which the intention value belongs (e.g., low intention, medium intention, high intention). Then, based on the determined intention level interval, select a corresponding authentication policy by looking up a preset policy mapping table or executing rules, and set the found authentication policy as the authentication policy to be executed in the current scenario.

[0111] The set of preset authentication policies includes at least two authentication policies with different interaction methods or security levels. For example, in the low intention interval, the first authentication policy is adopted, specifically "no response" or "maintain low power consumption monitoring". In the medium intention interval, the second authentication policy is adopted, specifically standard authentication, such as activating the screen and fingerprint sensor, and waiting for the user to take the initiative to operate (such as prompting the user to press the fingerprint or enter the password). In the high intention interval, the third authentication policy is adopted, specifically convenient authentication. The specific content of the high convenience authentication policy can include near-field device confirmation or touchless or single-step unlocking. Among them, near-field device confirmation is to detect that the RSSI of the authorized smart watch or mobile phone BLE signal is very strong (indicating a very short distance) and the user device is in the unlocked / active state. The intelligent door lock only sends a confirmation request to the device screen through BLE, and the user clicks to confirm to unlock. Touchless or single-step unlocking is to complete the authentication inside the intelligent door lock under the combination of very high confidence conditions (for example, intention value > 0.95 and RSSI > -50dBm and the initial fast face recognition by the camera is successful). The user only needs to directly pull the door handle, and the intelligent door lock can sense the pulling force and drive the lock tongue to retract to achieve touchless door opening.

[0112] Step S50, start the user identity authentication process specified by the selected authentication policy to authenticate the user;

[0113] Among them, in an embodiment of the present invention, the above steps specifically include:

[0114] Determine the door lock authentication hardware module that needs to be activated according to the selected authentication policy, and send an activation instruction to the determined door lock authentication hardware module;

[0115] Drive the user interface to output operation guidance information corresponding to the authentication policy;

[0116] Control the smart door lock to enter a state of waiting for the user to provide authentication information through the activated authentication hardware module.

[0117] Specifically, according to the selected authentication strategy (such as the second authentication strategy: standard fingerprint authentication), determine the door lock authentication hardware (fingerprint sensor) and user interface components (fingerprint area indicator light, prompt tone) that need to be activated, and send activation instructions to these components. Then drive the user interface (such as lighting the fingerprint indicator light and playing the voice of "Please press your fingerprint") to provide operation guidance to the user. At this time, the smart door lock enters a state of waiting for the user to perform corresponding operations (press the fingerprint) and waits to receive the authentication information provided by the user in the specified manner (press the fingerprint). If the authentication strategy involves interaction with the user device, send a request to the user device through the wireless interface and wait for its response.

[0118] Step S60, when the user authentication process is successfully verified, drive the smart door lock to perform an unlocking operation;

[0119] Among them, in an embodiment of the present invention, when the user completes the operation according to the guidance and the authentication module of the smart door lock (such as fingerprint comparison algorithm, password verification logic, NFC verification protocol, secure Bluetooth communication confirmation) determines that the authentication information provided by the user is valid, that is, when the user authentication process is successfully verified, the control unit of the door lock sends an instruction to the physical lock body mechanism to drive the motor or electromagnet to act and retract the lock tongue to complete the unlocking operation. If the verification fails or times out, the locked state is maintained, and a failure event may be recorded or an alarm may be triggered (according to the security settings).

[0120] Among them, in other embodiments of the present invention, it further includes:

[0121] Record the relevant data during each execution of the method, and the relevant data at least includes the sensor data, the user device data, the intention value, the selected authentication strategy, and the result of the user authentication process;

[0122] Utilize the recorded relevant data to periodically or when a preset trigger condition is met, update the internal parameters of the intention analysis model or adjust the preset intention threshold.

[0123] Specifically, during each interaction between the user and the door lock (regardless of whether the door opening is successful), the intelligent door lock will record relevant key data, such as the snapshot of sensor readings at that time, user device information, calculated intention values, finally selected authentication strategies, the actual authentication method used by the user (if different from the system recommendation), and the verification result (success / failure). Then, the accumulated recorded data is used to incrementally train or retrain the intention analysis model periodically (e.g., every night) or offline to improve its accuracy in predicting user intentions. At the same time, according to the statistical analysis results (e.g., a high misjudgment rate in a certain scenario), the preset intention threshold is automatically fine-tuned to better match the usage habits and security preferences of specific users. At this time, when a decrease in model performance is detected or a large number of user feedback is received, the update process is triggered. The update process can be executed locally on the intelligent door lock (if the computing power is sufficient), or the relevant data can be uploaded to the cloud server for training, and then the updated model / intention threshold is downloaded back to the intelligent door lock.

[0124] In summary, the intelligent door lock opening control method in the above embodiments of the present invention can more accurately understand the user's current true intention by fusing multi-modal data from sensors at the intelligent door lock end and user devices, and using the intention analysis model to calculate the user's door opening intention value, avoiding unnecessary disturbances or misoperations when the user has no intention of opening the door (such as passing by or staying), making the response of the intelligent door lock more intelligent and user-friendly; and adaptively selecting different authentication strategies according to the quantified user door opening intention value. When the system determines that the user has a strong intention to open the door, a more convenient authentication method can be selected, greatly improving the convenience for the user when it is inconvenient to operate (such as carrying heavy objects); when the user's intention is unclear or low, a more secure standard authentication strategy is selected or maintained to ensure the security of the door lock. This dynamic balance mechanism effectively solves the problem that it is difficult to balance security and convenience in the prior art; at the same time, it uses sensor data (such as distance, speed, posture) reflecting the user's physical state and user device status data (such as device identifier, signal strength), with a richer information dimension, making the intention judgment result more reliable and less susceptible to interference by a single factor, improving the stability and robustness of the control method. It solves the problem of insufficient intelligence of existing intelligent door locks.

[0125] Embodiment 2

[0126] Please refer to Figure 2 , which is a schematic structural diagram of an intelligent door lock opening control system provided by the second embodiment of the present invention. For the convenience of description, only the parts related to the embodiments of the present invention are shown. The intelligent door lock opening control system includes:

[0127] The sensor data acquisition module 11 is configured to acquire at least one type of sensor data reflecting the relative state between the user and the smart door lock, which is obtained by sensors deployed on or around the smart door lock;

[0128] The user device data receiving module 12 is configured to receive user device data including device identifiers and / or communication signal strengths sent by an authorized user device;

[0129] The intention value calculation module 13 is configured to input the acquired sensor data and the received user device data into a preset intention analysis model, and calculate an intention value representing the current degree of the user's willingness to open the door through the intention analysis model;

[0130] The authentication policy selection module 14 is configured to compare the calculated intention value with at least one preset intention threshold, and based on the comparison result, select an authentication policy corresponding to the intention value from a set including at least two preset authentication policies with different interaction methods or security levels;

[0131] The identity authentication module 15 is configured to initiate a user identity authentication process specified by the selected authentication policy to authenticate the user;

[0132] The unlocking module 16 is configured to drive the smart door lock to perform an unlocking operation when the user identity authentication process by the identity authentication module 15 is successfully verified.

[0133] Further, in an embodiment of the present invention, the sensor data acquisition module 11 includes:

[0134] The monitoring unit is configured to continuously monitor a predetermined area in front of the smart door lock by using at least one low-power passive sensor;

[0135] The trigger signal sending unit is configured to generate and send a trigger signal for indicating the start of subsequent data acquisition when the passive sensor detects that a moving object enters the predetermined area;

[0136] The activation unit is configured to activate at least one active sensor in response to the trigger signal;

[0137] The first sensor data acquisition unit is configured to, when the active sensor activated by the activation unit is a ranging sensor, periodically measure and acquire distance information and / or relative motion information between the target object and the smart door lock by using the ranging sensor, and use the distance information and / or relative motion information as part of the sensor data;

[0138] The second sensor data acquisition unit is configured to, when the active sensor activated by the activation unit is a camera or a structured light sensor, capture an image or depth information within the predetermined area by using the camera or the structured light sensor, process the captured image or depth information to extract preliminary feature information, and use the preliminary feature information as part of the sensor data.

[0139] Further, in an embodiment of the present invention, the second sensor data acquisition unit includes:

[0140] The human target recognition sub-unit is configured to perform a target detection algorithm on the captured image or depth information to identify and locate the human target therein;

[0141] The first preliminary feature information extraction sub-unit is configured to calculate and extract target parameters as the preliminary feature information according to the identified human target, where the target parameters include the bounding box coordinates, centroid position, motion speed vector, orientation angle of the human target, or the relative distance estimation from the intelligent door lock;

[0142] The second preliminary feature information extraction sub-unit is configured to apply a pose estimation algorithm to the human target, extract the pose information of the skeletal joint points, compare the pose information with a preset pose template library, obtain a judgment result on whether the user is in a specific pose, and include the judgment result in the preliminary feature information.

[0143] Further, in an embodiment of the present invention, the user device data receiving module 12 includes:

[0144] The scanning unit is configured to periodically scan for surrounding wireless devices by using the wireless communication module built in the intelligent door lock;

[0145] The device identifier extraction unit is configured to extract the device identifier of the user device from the signals captured during the scanning process;

[0146] The comparison unit is configured to compare the extracted device identifier with a pre-stored list of authorized device identifiers;

[0147] The device identifier recording unit is configured to, when a successfully matched authorized device identifier is found, confirm that the successfully matched user device is an authorized user device and record the device identifier;

[0148] The signal strength value measuring unit is configured to measure and obtain the signal strength value received when communicating with the authorized user device by the wireless communication module;

[0149] The user device data forming unit is configured to combine the confirmed and recorded device identifier of the authorized user device and the obtained signal strength value to form user device data.

[0150] Further, in an embodiment of the present invention, the intention value calculation module 13 includes:

[0151] A data integration unit, configured to integrate the collected sensor data and the received user device data into an original feature set;

[0152] A data preprocessing unit, configured to perform preprocessing operations on the original feature set, where the preprocessing operations include time series alignment and feature value normalization;

[0153] A data input unit, configured to input the preprocessed feature set into a preset intention analysis model;

[0154] An intention value calculation unit, configured to perform calculation processing on the input multi-dimensional feature vector by the intention analysis model, and output a numerical value representing the probability or confidence of the user's door opening intention as the intention value.

[0155] Further, in an embodiment of the present invention, the authentication policy selection module 14 includes:

[0156] A comparison unit, configured to perform a mathematical comparison of the calculated intention value with one or more preset intention thresholds stored in the intelligent door lock sequentially or in parallel;

[0157] A determination unit, configured to determine the intention level interval to which the intention value belongs according to the comparison result;

[0158] An authentication policy selection unit, configured to search and set, according to the determined intention level interval, the authentication policy uniquely corresponding to the determined intention level interval from the preset authentication policy set as the authentication policy in the current scenario.

[0159] Further, in an embodiment of the present invention, the identity authentication module 15 includes:

[0160] An activation unit, configured to determine the door lock authentication hardware module to be activated according to the selected authentication policy, and send an activation instruction to the determined door lock authentication hardware module;

[0161] An output unit, configured to drive the user interface to output operation guidance information corresponding to the authentication policy;

[0162] A waiting unit, configured to control the intelligent door lock to enter a state of waiting for the user to provide authentication information through the activated authentication hardware module.

[0163] Further, in an embodiment of the present invention, the system further includes:

[0164] A recording module, configured to record relevant data during each execution of the method, where the relevant data at least includes the sensor data, the user device data, the intention value, the selected authentication policy, and the result of the user authentication process;

[0165] An update module, configured to use the recorded relevant data to periodically or when a preset trigger condition is met, update the internal parameters of the intention analysis model or adjust the preset intention threshold.

[0166] The intelligent door lock opening control system provided by the embodiments of the present invention has the same implementation principle and technical effects as those of the foregoing method embodiments. For a brief description, for parts not mentioned in the device embodiments, reference may be made to the corresponding content in the foregoing method embodiments.

[0167] Embodiment III

[0168] On the other hand, the present invention also proposes an intelligent door lock. Please refer to Figure 3 , which shows the intelligent door lock in the third embodiment of the present invention, including a memory 20, a processor 10, and a program 30 stored on the memory 20 and executable on the processor. When the processor 10 executes the program 30, it implements the intelligent door lock opening control method as described in the above embodiments.

[0169] Among them, in some embodiments, the processor 10 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips, and is used to run the program code stored in the memory 20 or process data, such as executing an access restriction program, etc.

[0170] Among them, the memory 20 at least includes one type of readable storage medium. The readable storage medium includes flash memory, a hard disk, a multimedia card, a card-type memory (such as an SD or DX memory, etc.), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the memory 20 may be an internal storage unit of the intelligent door lock, such as the hard disk of the intelligent door lock. In other embodiments, the memory 20 may also be an external storage device of the intelligent door lock, such as a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the intelligent door lock. Further, the memory 20 may also include both the internal storage unit and the external storage device of the intelligent door lock. The memory 20 can not only be used to store application software installed on the intelligent door lock and various types of data, but also be used to temporarily store data that has been output or will be output.

[0171] It should be noted that Figure 3The structures shown do not constitute a limitation on the intelligent door lock. In other embodiments, the intelligent door lock may include fewer or more components than those shown, or combine certain components, or have different component arrangements.

[0172] In summary, for the intelligent door lock in the above embodiments of the present invention, by integrating multi-modal data from sensors at the intelligent door lock end and user devices, and using the intention analysis model to calculate the user's door opening intention value, it can more accurately understand the user's current true intention, avoiding unnecessary disturbances or misoperations when the user has no intention of opening the door (such as passing by or staying), making the response of the intelligent door lock more intelligent and user-friendly; and according to the quantified user's door opening intention value, different authentication strategies are adaptively selected. When the system determines that the user has a strong intention to open the door, a more convenient authentication method can be selected, greatly improving the convenience for the user when it is inconvenient to operate (such as carrying heavy objects); when the user's intention is unclear or low, a more secure standard authentication strategy is selected or maintained to ensure the security of the door lock. This dynamic balance mechanism effectively solves the problem in the prior art that it is difficult to balance security and convenience; at the same time, sensor data reflecting the user's physical state (such as distance, speed, posture) and user device state data (such as device identifier, signal strength) are used, with a richer information dimension, making the intention judgment result more reliable, not easily interfered by a single factor, and improving the stability and robustness of the control method. It solves the problem of insufficient intelligence of existing intelligent door locks.

[0173] The embodiment of the present invention also provides a storage medium, on which a program is stored, and when the program is executed by a processor, it implements the intelligent door lock opening control method as described in the above embodiments.

[0174] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, only the above division of each functional unit and module is used as an example. In actual applications, the above functions can be allocated to different functional units or modules according to needs, that is, the internal structure of the storage device is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiments can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of each functional unit and module are only for the convenience of mutual distinction and do not limit the protection scope of the present application.

[0175] Those skilled in the art will understand that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a definite sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any storage medium for use by an instruction execution system, apparatus or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus or device), or used in combination with these instruction execution systems, apparatus or devices. For the purposes of this specification, a "storage medium" can be any device that can contain, store, communicate, propagate or transport a program for use by or in combination with an instruction execution system, apparatus or device.

[0176] More specific examples (non-exhaustive list) of readable storage media include the following: an electrical connection part with one or more wirings (electronic device), a portable computer disk cartridge (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the storage medium can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation or other appropriate processing as necessary, and then stored in a memory.

[0177] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits with logic gate circuits for implementing logical functions on data signals, application specific integrated circuits with suitable combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0178] In the description of this specification, the description referring to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc. means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0179] The above-described embodiments merely represent several implementation manners of the present invention. The description thereof is relatively specific and detailed, but it should not be construed as a limitation on the scope of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all fall within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the appended claims.

Claims

1. An intelligent door lock opening control method, characterized in that, The method includes: Collecting at least one type of sensor data that reflects the relative state between the user and the smart door lock and is obtained by sensors deployed on or around the smart door lock; Receiving user device data including a device identifier and a communication signal strength sent by an authorized user device; Inputting the collected sensor data and the received user device data into a preset intention analysis model, and calculating an intention value representing the degree of the user's current willingness to open the door through the intention analysis model; Comparing the calculated intention value with at least one preset intention threshold, and based on the comparison result, selecting an authentication policy corresponding to the intention value from a set of preset authentication policies including at least two types with different interaction methods or security levels; Starting the user identity authentication process specified by the selected authentication policy to authenticate the user; When the user identity authentication process is successfully verified, driving the smart door lock to perform an unlocking operation.

2. The intelligent door lock opening control method according to claim 1, wherein, The step of collecting at least one type of sensor data that reflects the relative state between the user and the smart door lock and is obtained by sensors deployed on or around the smart door lock includes: Continuously monitoring a predetermined area in front of the smart door lock using at least one low-power passive sensor; When the passive sensor detects a moving object entering the predetermined area, generating and sending a trigger signal for indicating the start of subsequent data collection; In response to the trigger signal, activating at least one active sensor; When the activated active sensor is a ranging sensor, periodically measuring and obtaining distance information and / or relative motion information between the target object and the smart door lock using the ranging sensor, and taking the distance information and / or relative motion information as part of the sensor data; When the activated active sensor is a camera or a structured light sensor, capturing an image or depth information in the predetermined area using the camera or the structured light sensor, processing the captured image or depth information to extract preliminary feature information, and taking the preliminary feature information as part of the sensor data.

3. The intelligent door lock opening control method according to claim 2, wherein The step of processing the captured image or depth information to extract preliminary feature information includes: Performing an object detection algorithm on the captured image or depth information to identify and locate the human object therein; According to the identified human object, calculating and extracting target parameters as the preliminary feature information, where the target parameters include the bounding box coordinates, centroid position, motion speed vector, orientation angle, or relative distance estimate from the smart door lock of the human object; Applying a pose estimation algorithm to the human object, extracting the pose information of the skeletal joint points, comparing the pose information with a preset pose template library, obtaining a judgment result on whether the user is in a specific pose, and including the judgment result in the preliminary feature information.

4. The intelligent door lock opening control method according to claim 1, wherein The step of receiving user device data including a device identifier and a communication signal strength sent by an authorized user device includes: Periodically scanning for surrounding wireless devices using a wireless communication module built in the smart door lock; Extracting the device identifier of the user device from the signals captured during the scanning process; Compare the extracted device identifier with the pre - stored list of authorized device identifiers; When a match is successfully found with an authorized device identifier, confirm that the successfully matched user device is an authorized user device and record the device identifier; Measure and obtain the signal strength value received during communication with the authorized user device by the wireless communication module; Combine the confirmed and recorded device identifier of the authorized user device and the obtained signal strength value to form user device data.

5. The intelligent door lock opening control method according to claim 1, characterized in that, The step of inputting the collected sensor data and the received user device data into a preset intention analysis model and calculating, through the intention analysis model, an intention value representing the degree of the user's current willingness to open the door includes: Integrate the collected sensor data and the received user device data into an original feature set; Perform pre - processing operations on the original feature set, and the pre - processing operations include time - series alignment and feature value normalization; Input the pre - processed feature set into a preset intention analysis model; The intention analysis model performs calculation processing on the input multi - dimensional feature vector and outputs a numerical value representing the probability or confidence level of the user's door - opening intention as the intention value.

6. The intelligent door lock opening control method according to claim 1, wherein The step of comparing the calculated intention value with at least one preset intention threshold and, based on the comparison result, selecting, from a preset set of authentication policies containing at least two types of authentication policies with different interaction methods or security levels, an authentication policy corresponding to the intention value includes: Mathematically compare the calculated intention value sequentially or in parallel with one or more preset intention thresholds stored in the smart lock; Determine the intention level interval to which the intention value belongs according to the comparison result; Based on the determined intention level interval, search in the preset set of authentication policies and set the authentication policy uniquely corresponding to the determined intention level interval as the authentication policy in the current scenario.

7. The intelligent door lock opening control method according to claim 1, characterized in that, The step of starting the user identity authentication process specified by the selected authentication policy includes: Determine the door lock authentication hardware module that needs to be activated according to the selected authentication policy and send an activation instruction to the determined door lock authentication hardware module; Drive the user interface to output operation guidance information corresponding to the authentication policy; Control the smart lock to enter a state of waiting for the user to provide authentication information through the activated authentication hardware module.

8. The intelligent door lock opening control method according to claim 1, characterized in that The method further includes: Record the relevant data during each execution of the method, and the relevant data at least includes the sensor data, the user device data, the intention value, the selected authentication policy, and the result of the user identity authentication process; Use the recorded relevant data to periodically or when a preset trigger condition is met, update the internal parameters of the intention analysis model or adjust the preset intention threshold.

9. An intelligent door lock opening control system, characterized in that, The system includes: A sensor data acquisition module for acquiring at least one type of sensor data reflecting the relative state between the user and the smart lock obtained by sensors deployed on or around the smart lock; A user device data receiving module for receiving user device data containing a device identifier and / or communication signal strength sent by an authorized user device; An intention value calculation module, configured to input the collected sensor data and the received user device data into a preset intention analysis model, and calculate, through the intention analysis model, an intention value for representing the current degree of the user's willingness to open the door; An authentication policy selection module, configured to compare the calculated intention value with at least one preset intention threshold, and based on the comparison result, select, from a set including at least two preset authentication policies with different interaction methods or security levels, an authentication policy corresponding to the intention value; An identity authentication module, configured to initiate a user identity authentication process specified by the selected authentication policy to authenticate the user; An unlocking module, configured to, when the identity authentication module successfully verifies the user identity authentication process, drive the intelligent door lock to perform an unlocking operation.

10. An intelligent door lock, characterized in that, It includes a memory, a processor, and a program stored on the memory and executable on the processor. When the processor executes the program, it implements the intelligent door lock opening control method according to any one of claims 1-8.

Images