Automatic parking lock system and method based on Bluetooth AOA positioning

Abstract

The invention relates to an automatic parking lock system and method based on Bluetooth AOA positioning, and the system comprises a Bluetooth AOA positioning module which is used for receiving a Bluetooth signal of user equipment and feeding back positioning information; the control unit is used for receiving the positioning information uploaded by the user equipment and sending a driving signal for controlling locking and unlocking of the parking lock after position verification; and the executing mechanism is used for receiving the driving signal to realize locking and unlocking of the parking lock. Compared with the prior art, the system has the advantages that the user positioning precision and the parking lock response speed are improved, the safety and reliability of the system are enhanced, and intelligent and automatic operation is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of ground lock control, and in particular to an automatic ground lock system and method based on Bluetooth AOA positioning. Background Art

[0002] The existing methods for unlocking underground garage ground locks include manual unlocking using remote control, unlocking using RFID tags, etc., which are specifically introduced as follows:

[0003] (1) Remote control unlocking: One of the most commonly used methods for unlocking ground locks is remote control unlocking. The user sends a signal to the ground lock through the remote control to trigger the unlocking mechanism. This method is simple to operate and user-friendly.

[0004] (2) Unlocking via mobile phone APP: Many modern ground lock systems support control via dedicated mobile phone applications. The user initiates the unlocking command in the APP, which communicates with the ground lock via Bluetooth or Wi-Fi signals to complete the unlocking operation.

[0005] (3) Automatic induction unlocking: Some ground lock systems are equipped with sensors that can detect the arrival of a vehicle and automatically trigger unlocking. Such systems usually rely on geomagnetic or infrared sensors to identify the vehicle.

[0006] The existing technical solutions still inevitably require manual unlocking or the installation of sensors on the vehicle, and have the following defects:

[0007] (1) Signal interference problem: Both remote control unlocking and mobile phone APP unlocking rely on wireless signals, which may be interfered by other devices in the surrounding environment, causing the unlocking signal to be unstable or ineffective.

[0008] (2) Power consumption and maintenance issues: The automatic sensing unlocking ground lock system consumes a lot of power due to the continuous operation of the sensor, and requires frequent charging or battery replacement, which increases maintenance costs.

[0009] (3) Security issues: Especially for Bluetooth-based APP unlocking systems, if the Bluetooth protocol is not secure enough or has loopholes, it may be attacked by hackers, resulting in illegal unlocking.

[0010] (4) Inaccurate positioning: Existing automatic sensing unlocking systems mostly rely on near-field sensing technology, which has limited positioning accuracy and may not be able to accurately control the unlocking and locking of ground locks, especially in complex or crowded parking environments. Summary of the invention

[0011] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and provide an automatic ground lock system and method based on Bluetooth AOA positioning to improve positioning accuracy and response speed.

[0012] The purpose of the present invention can be achieved by the following technical solutions:

[0013] An automatic ground lock system based on Bluetooth AOA positioning, comprising:

[0014] Bluetooth AOA positioning module, used to receive Bluetooth signals from user equipment and feedback positioning information;

[0015] A control unit is used to receive the positioning information uploaded by the user device, and after position verification, send a driving signal for controlling the locking and unlocking of the ground lock;

[0016] The actuator is used to receive the driving signal to achieve locking and unlocking of the ground lock.

[0017] Furthermore, the Bluetooth AOA positioning module calculates the direction of the source of the received signal based on the angle of arrival technology to obtain the positioning information of the user equipment.

[0018] Furthermore, the Bluetooth AOA positioning module includes a Bluetooth receiving device, and an antenna arranged on the Bluetooth receiving device for receiving Bluetooth signals from user equipment.

[0019] Furthermore, the shape of the Bluetooth AOA positioning module is triangular, circular or square.

[0020] Furthermore, the control unit is also used to verify the identity of the user, and after completing the position verification and user identity verification, it sends a driving signal for controlling the locking and unlocking of the ground lock.

[0021] Furthermore, the actuator is a retractable locking rod, and the locking rod is driven by an electric motor, a hydraulic system, or a pneumatic system.

[0022] Furthermore, the identity verification process specifically includes:

[0023] Data collection steps: Collect user behavior data, historical verification records, account characteristics, and identity information modification behavior;

[0024] Feature processing step: extract features from the data collected in the data collection step to obtain time features, geographic location features, device features, and user behavior features; and pre-process the extracted features;

[0025] Model training steps: define the user's identity authentication risk level as a risk label; define corresponding risk labels for the features obtained in the feature processing step and build a data set; use the data set to train the machine learning model;

[0026] Model reasoning steps: collect the current environment data of the login device and the user's historical behavior records as input data; extract and preprocess the input data to obtain the features to be predicted; input the features to be predicted into the trained machine learning model for aggregation to obtain the prediction results of the identity authentication risk level; the identity authentication risk level includes: a low risk level with a value of 0, and the corresponding system response is to allow direct verification;

[0027] The medium risk level is assigned a value of 1, and the corresponding system response is to trigger a simple verification;

[0028] The low risk level is assigned a value of 2, and the corresponding system response is to trigger multi-factor authentication;

[0029] The low risk level is assigned a value of 3, and the corresponding system response is to deny login and notify the user to confirm;

[0030] Model learning steps: Collect user feedback and perform online updates of the machine learning model.

[0031] The present invention also provides a control method for an automatic ground lock system based on Bluetooth AOA positioning as described above, comprising the following steps:

[0032] Sending a Bluetooth signal to the Bluetooth AOA positioning module via the user equipment;

[0033] The Bluetooth AOA positioning module calculates the source direction of the signal based on the received Bluetooth signal, and then feeds back the positioning information to the user device;

[0034] The user device uploads the received positioning information to the control unit, which verifies the position. If the position verification is successful, a driving signal for controlling the locking and unlocking of the ground lock is issued;

[0035] The locking and unlocking of the ground lock is achieved by controlling the actuator through the drive signal.

[0036] Furthermore, the method further comprises:

[0037] The control unit identifies the information transmitted by the user device and verifies the user's identity. When the identity verification and the position verification are passed, the driving signal is issued.

[0038] Furthermore, the method further comprises:

[0039] Detect whether there are obstacles around the ground lock to be driven. If there are obstacles, stop sending the driving signal to the actuator.

[0040] Compared with the prior art, the present invention has the following advantages:

[0041] (1) Improve positioning accuracy and response speed: By adopting Bluetooth AOA (angle of arrival) technology, the present invention can more accurately determine the location of the user's device. Compared with the traditional strength (RSSI)-based Bluetooth positioning method, AOA technology can provide higher angular resolution, thereby greatly improving positioning accuracy. This high-precision positioning ensures that the ground lock can respond quickly and accurately to the approach of authorized users, so that the user can unlock the parking space immediately when the user arrives at the parking space, optimizing the user experience.

[0042] (2) Enhanced system security and reliability: The solution of the present invention further enhances the security of the ground lock system by integrating advanced user verification processes (optional steps), such as machine learning algorithms to identify and verify user identities. In addition, the physical design of the system also takes into account anti-tampering and anti-destruction characteristics, ensuring that the ground lock can maintain a high degree of security and reliability even in public or unmonitored environments.

[0043] (3) Realize intelligent and automated operation: The automated ground lock system can automatically unlock or lock according to the user's location, without manual intervention. This intelligent function not only improves the convenience of operation, but also reduces the physical contact between the user and the device, extending the service life of the device. The system can also be configured with the function of real-time monitoring of environmental changes, such as detecting whether there are other vehicles or obstacles around, to ensure the safety of ground lock operation.

[0044] (4) Flexibility and scalability: Since the system design takes into account the interchangeability of different actuators and data processing methods, the ground lock system can be adjusted and optimized according to specific application scenarios and cost-effectiveness. For example, electric, hydraulic or pneumatic actuators can be selected as needed, or different types of control units and storage technologies can be used to adapt to different environments and budget requirements.

[0045] (5) Supporting environmental sustainability: The design of the ground lock system allows the introduction of new features or performance improvements through remote software updates, reducing the frequency of hardware replacement and related environmental impacts. In addition, the system's automated features help optimize the use of parking space, reduce vehicle emissions caused by searching for parking spaces, and support the sustainable development of urban transportation. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 The present invention provides a schematic diagram of the structure of an automatic ground lock system based on Bluetooth AOA positioning in an embodiment of the present invention. DETAILED DESCRIPTION

[0047] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0048] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0050] Example 1

[0051] like Figure 1 As shown, this embodiment provides an automatic ground lock system based on Bluetooth AOA positioning, including:

[0052] Bluetooth AOA positioning module, used to receive Bluetooth signals from user equipment and feedback positioning information;

[0053] A control unit is used to receive the positioning information uploaded by the user device, and after position verification, send a driving signal for controlling the locking and unlocking of the ground lock;

[0054] The actuator is used to receive a driving signal to lock and unlock the ground lock.

[0055] Specifically, the Bluetooth AOA positioning module includes a Bluetooth receiving device and an antenna arranged on the Bluetooth receiving device for receiving Bluetooth signals from the user equipment.

[0056] The user device may be a smart phone, and the user may automatically control the opening or closing of the ground lock on the vehicle through the Bluetooth signal of the smart phone.

[0057] The Bluetooth AOA positioning module calculates the direction of the received signal source based on the Angle of Arrival (AOA) technology, provides accurate data for ground positioning, and obtains the positioning information of the user device.

[0058] The Bluetooth AOA positioning module preferably adopts a triangular design to optimize the signal reception angle and device stability. The shape is not limited and can also be circular, square or other polygons.

[0059] As for the control unit, it includes a processor and a memory, which are integrated on an integrated circuit board;

[0060] The control unit is used to process the positioning data and control the locking and unlocking of the ground lock.

[0061] Preferably, the control unit is also used to verify the identity of the user, and after completing the position verification and user identity verification, it sends a driving signal for controlling the locking and unlocking of the ground lock.

[0062] In addition, the specific model and processing power of the control unit can be flexibly selected according to the complexity and budget of the ground locking system.

[0063] As for the actuator, it may be a locking structure driven by an electric motor, including a retractable locking rod;

[0064] After receiving the driving signal from the control unit, locking and unlocking are achieved by physically moving the lock rod.

[0065] The actuator can also use hydraulic or pneumatic systems instead of electric motors to adapt to different environmental and cost requirements.

[0066] Preferably, the control unit is further connected to a local storage unit for recording the locking state and control log of the ground lock, so as to restore the state of the ground lock when the control unit is restarted or fails, so that the ground lock can be restored to a correct state;

[0067] The storage technology of the local storage unit can be solid-state storage or cloud storage, depending on security and cost considerations.

[0068] This embodiment also provides a control method for an automatic ground lock system based on Bluetooth AOA positioning as described above, comprising the following steps:

[0069] Sending a Bluetooth signal to the Bluetooth AOA positioning module via the user equipment;

[0070] The Bluetooth AOA positioning module calculates the source direction of the signal based on the received Bluetooth signal, and then feeds back the positioning information to the user device;

[0071] The user device uploads the received positioning information to the control unit, which verifies the position. If the position verification is successful, a driving signal for controlling the locking and unlocking of the ground lock is issued;

[0072] The locking and unlocking of the ground lock is achieved by controlling the actuator through the drive signal.

[0073] Preferably, the method further comprises:

[0074] The control unit identifies the information transmitted by the user device and verifies the user's identity. When the identity verification and the position verification are passed, a driving signal is issued.

[0075] The identity verification process uses machine learning algorithms to further verify the user's identity and dynamically adjust the strictness based on the user's risk assessment. The risk level is evaluated based on whether the verification is done remotely, whether the verification is repeated, and other behavioral assessments. The machine learning model outputs its risk indicators.

[0076] Specifically include:

[0077] Data collection steps: Collect user behavior data, historical verification records, account characteristics, and identity information modification behavior;

[0078] Feature processing step: extract features from the data collected in the data collection step to obtain time features, geographic location features, device features, and user behavior features; and pre-process the extracted features;

[0079] Model training steps: define the user's identity authentication risk level as a risk label; define corresponding risk labels for the features obtained in the feature processing step and build a data set; use the data set to train the machine learning model;

[0080] Model reasoning steps: collect the current environment data of the login device and the user's historical behavior records as input data; extract and preprocess the input data to obtain the features to be predicted; input the features to be predicted into the trained machine learning model for aggregation to obtain the prediction results of the identity authentication risk level; the identity authentication risk level includes: a low risk level with a value of 0, and the corresponding system response is to allow direct verification;

[0081] The medium risk level is assigned a value of 1, and the corresponding system response is to trigger a simple verification;

[0082] The low risk level is assigned a value of 2, and the corresponding system response is to trigger multi-factor authentication;

[0083] The low risk level is assigned a value of 3, and the corresponding system response is to deny login and notify the user to confirm;

[0084] Model learning steps: Collect user feedback and perform online updates of the machine learning model.

[0085] Preferably, the method further comprises:

[0086] Detect whether there are obstacles around the ground lock to be driven. If there are obstacles, stop sending the driving signal to the actuator.

[0087] The following is a specific implementation process:

[0088] Step A (signal reception): The user device sends a Bluetooth signal, which is received by the Bluetooth AOA positioning module of the ground lock.

[0089] Step B1 / B2 (data processing): The processor calculates the precise location of the signal and determines whether the user is an authorized user.

[0090] -Optional alternative step B3 / B4: Use machine learning algorithms to further verify user identity.

[0091] Step C1 (decision execution): Based on location and user authentication, the control unit sends instructions to the execution agency.

[0092] -Omittable: In scenarios where high security requirements are not required, the user verification step can be skipped.

[0093] Step D / E (execution and feedback): the actuator operates the lock rod, and the ground lock is unlocked or locked; the system status is updated and fed back to the user device.

[0094] - Step F can be added: real-time monitoring of environmental changes and dynamic adjustment of ground lock response.

[0095] The process of verifying user identity aims to dynamically evaluate the user's identity authentication risk level through machine learning algorithms and adjust the strictness of verification according to the risk level. The core includes five parts: data collection, feature engineering, model training, model reasoning, and verification feedback. The design and implementation process of each step is explained in detail below.

[0096] 1. Data Collection

[0097] 1.1 Data Source

[0098] User behavior data

[0099] oLogin time (working hours, non-working hours)

[0100] o Login location (geographic location, IP address, country / city / device location)

[0101] oLogin device (device ID, operating system, browser fingerprint)

[0102] oDevice network information (WiFi name, mobile network operator, proxy usage)

[0103] oSensor data (accelerometer, gyroscope, device orientation)

[0104] Historical verification records

[0105] oPast login success / failure records

[0106] oLogin IP change rate

[0107] oWhether to log in from a different location (threshold for geographic location change, such as 100km)

[0108] oMany login attempts in a short period of time (failure rate)

[0109] Other data

[0110] oUser’s account characteristics (new user, active user, VIP user)

[0111] o User identity information modification behavior (whether account information, password, email address, etc. are modified multiple times in a short period of time)

[0112] 2. Feature Engineering

[0113] 2.1 Feature Extraction

[0114] Time characteristics:

[0115] oWhether the login time is working time (0 / 1)

[0116] oThe time interval between the last login and the last login (hours, days)

[0117] oWhether the login time is at night (0 / 1)

[0118] Geographical location features:

[0119] oThe change distance of the user's geographical location (calculate the distance between the last login location and the current location, unit: km)

[0120] oWhether to log in from a different location (if the distance is greater than 100km, it is considered to be a different location)

[0121] Equipment features:

[0122] oWhether the type, operating system, and browser of the login device are the same as last time (0 / 1)

[0123] oChanges in device acceleration and velocity data (compute features such as variance, maximum, and minimum values ​​through sliding windows)

[0124] User behavior characteristics:

[0125] oThe number of times a user has attempted to log in within a certain time window

[0126] oWhether the login is performed in a VPN environment (0 / 1)

[0127] oHistorical login failure rate (failure rate within 30 days)

[0128] Account Information Features:

[0129] oThe user is a new user (0 / 1)

[0130] oWhether the user frequently changes password / binds email address (0 / 1)

[0131] 2.2 Feature Preprocessing

[0132] Missing value processing: For missing data such as GPS location and device information, fill it with the previous value.

[0133] Data normalization: Use Min-Max normalization for continuous features (such as distance, time interval).

[0134] Discrete feature encoding: One-hot encoding is used for categorical features (device type, operating system, etc.).

[0135] 3. Model training

[0136] 3.1 Tag Definition

[0137] Risk Label: Classify the user's authentication risk into multiple levels:

[0138] o 0-Low risk (common equipment and location, stable over time)

[0139] o 1-Medium risk (equipment replacement, slight location change)

[0140] o 2-High risk (remote login, abnormal time, device change)

[0141] o 3- Very high risk (multiple failures in a short period of time, extremely remote locations, using proxies)

[0142] 3.2 Dataset Division

[0143] Training set (80%): used to train the model

[0144] Validation set (10%): used for hyperparameter tuning

[0145] Test set (10%): used for model performance evaluation

[0146] 3.3 Model selection

[0147] Classic machine learning algorithms:

[0148] oLogistic regression: suitable for simple classification tasks that are linearly separable

[0149] oDecision Tree / Random Forest: handles high-dimensional data, is highly adaptable, and has feature importance interpretation

[0150] o XGBoost / LightGBM: suitable for large-scale data sets, fast speed, strong generalization ability

[0151] oDeep learning model (LSTM / GRU): If the data has strong temporal dependency, LSTM can model the sequential behavior of users.

[0152] 3.4 Training Process

[0153] 1. Data loading: Load the sample data generated by feature engineering into memory.

[0154] 2. Data augmentation: Randomly perturb the data (such as randomly changing the geographic location) to enhance the robustness of the model.

[0155] 3. Model training:

[0156] o Select the XGBoost model and define the hyperparameters (tree depth, learning rate, regularization parameter).

[0157] o K-fold cross-validation was used to evaluate model performance.

[0158] 4. Model saving: Save the optimal model as a file (such as model.pkl) to facilitate online reasoning.

[0159] 4. Model Reasoning

[0160] 4.1 Reasoning Process

[0161] 1. Input data collection:

[0162] o Current environment data of the logged-in device (time, IP, location, device ID, etc.).

[0163] oUser's historical behavior records.

[0164] 2. Feature Engineering:

[0165] oRegenerate features and extract time features, location features, device features, etc. of new data.

[0166] 3. Model loading:

[0167] o Load a trained model (such as XGBoost model model.pkl) from cache or model repository.

[0168] 4. Model prediction:

[0169] oThe model input is the feature vector X, and the predicted output is the risk level y, for example: y=f(X)y=f(X)y=f(X)where y∈{0,1,2,3}y\in\{0,1,2,3\}y∈{0,1,2,3} represents the risk level.

[0170] 5. Dynamic adjustment of risk levels

[0171] Based on the output risk level of the model, the system dynamically adjusts the strictness of the verification:

[0172] Risk Level System Response 0 (low risk) Allow direct verification 1 (Medium risk) Trigger simple verification (such as verification code) 2 (High Risk) Trigger multi-factor authentication (such as SMS verification) 3 (very high risk) Deny login and notify the user to confirm

[0173] 6. Model Feedback and Online Learning

[0174] User feedback:

[0175] If a user reports a login anomaly, their data is added to the training set as a new sample.

[0176] Online model update:

[0177] Adopt online learning mechanism and retrain the model with new data at regular intervals.

[0178] Real-time monitoring:

[0179] Monitor the distribution of each risk level to ensure that the system's sensitivity and recall meet business requirements.

[0180] 7. Performance Evaluation

[0181] Evaluation indicators:

[0182] oAccuracy: used to measure the overall accuracy of model predictions.

[0183] oRecall: The proportion of high-risk behaviors that the model can correctly capture.

[0184] o AUC-ROC curve: used to evaluate the model's ability to distinguish at different thresholds.

[0185] o Confusion matrix: shows the relationship between predicted labels and true labels, and analyzes the false positive rate and false negative rate of the model.

[0186] Model performance improvement strategy:

[0187] oIn risk prediction, high recall is better than high precision, because we would rather trigger one more verification than allow the user account to be maliciously logged in.

[0188] o Improve the classification performance of small classes by resampling the imbalanced data (such as SMOTE).

[0189] 8. System architecture design

[0190] Data Layer:

[0191] oStore raw log data and feature engineering results.

[0192] Model Serving:

[0193] oUse Flask / FastAPI as API service to provide online prediction interface.

[0194] API interface:

[0195] o / predict: Input user feature data and output risk level.

[0196] o / feedback: User-submitted feedback for model retraining.

[0197] Task Scheduling:

[0198] oSchedule online training of the model through **scheduled tasks (such as Airflow)**.

[0199] Overall, this solution achieves the following functions:

[0200] (1) High-precision positioning: Through Bluetooth AOA positioning technology, high-precision user positioning can be achieved and the operation of the ground lock can be accurately controlled.

[0201] (2) Automated operation: The car automatically unlocks when the user approaches and locks when the user leaves, providing a seamless and convenient parking experience.

[0202] (3) Enhanced security: The structural design can further consider anti-tampering and anti-destruction to enhance the overall security of the ground lock system.

[0203] The preferred specific embodiments of the present invention are described in detail above. It should be understood that a person skilled in the art can make many modifications and changes based on the concept of the present invention without creative work. Therefore, any technical solution that can be obtained by a person skilled in the art through logical analysis, reasoning or limited experiments based on the concept of the present invention on the basis of the prior art should be within the scope of protection determined by the claims.

Claims

1. An automatic ground lock system based on Bluetooth AOA positioning, characterized in that: include: Bluetooth AOA positioning module, used to receive Bluetooth signals from user equipment and feedback positioning information; A control unit is used to receive the positioning information uploaded by the user device, and after position verification, send a driving signal for controlling the locking and unlocking of the ground lock; The actuator is used to receive the driving signal to achieve locking and unlocking of the ground lock.

2. The automatic ground lock system based on Bluetooth AOA positioning according to claim 1 is characterized in that: The Bluetooth AOA positioning module calculates the direction of the received signal source based on the angle of arrival technology to obtain the positioning information of the user equipment.

3. The automatic ground lock system based on Bluetooth AOA positioning according to claim 1 is characterized in that: The Bluetooth AOA positioning module includes a Bluetooth receiving device, and an antenna arranged on the Bluetooth receiving device for receiving a Bluetooth signal from a user device; The shape of the Bluetooth AOA positioning module is triangular, circular or square.

4. The automatic ground lock system based on Bluetooth AOA positioning according to claim 1 is characterized in that: The control unit is also used to verify the identity of the user, and after completing the position verification and user identity verification, it sends a driving signal for controlling the locking and unlocking of the ground lock.

5. The automatic ground lock system based on Bluetooth AOA positioning according to claim 1 is characterized in that: The actuator is a retractable locking rod, which is driven by an electric motor, a hydraulic system, or a pneumatic system.

6. The automatic ground lock system based on Bluetooth AOA positioning according to claim 1, characterized in that: The control unit is also connected to a local storage unit for recording the locking state and control log of the ground lock, so as to restore the state of the ground lock when the control unit is restarted or fails; The local storage unit is solid state storage or cloud storage.

7. A control method for an automatic ground lock system based on Bluetooth AOA positioning according to any one of claims 1 to 6, characterized in that: The following steps are involved: Sending a Bluetooth signal to the Bluetooth AOA positioning module via the user equipment; The Bluetooth AOA positioning module calculates the source direction of the signal based on the received Bluetooth signal, and then feeds back the positioning information to the user device; The user device uploads the received positioning information to the control unit, which verifies the position. If the position verification is successful, a driving signal for controlling the locking and unlocking of the ground lock is issued; The locking and unlocking of the ground lock is achieved by controlling the actuator through the drive signal.

8. The method according to claim 7, characterized in that The method further comprises: The control unit identifies the information transmitted by the user device and verifies the user's identity. When the identity verification and the position verification are passed, the driving signal is issued.

9. The method according to claim 8, characterized in that The identity verification process specifically includes: Data collection steps: Collect user behavior data, historical verification records, account characteristics, and identity information modification behavior; Feature processing step: extract features from the data collected in the data collection step to obtain time features, geographic location features, device features, and user behavior features; and pre-process the extracted features; Model training steps: define the user's identity authentication risk level as a risk label; define corresponding risk labels for the features obtained in the feature processing step and build a data set; use the data set to train the machine learning model; Model reasoning steps: collect the current environment data of the login device and the user's historical behavior records as input data; extract and preprocess the input data to obtain the features to be predicted; input the features to be predicted into the trained machine learning model for aggregation to obtain the prediction results of the identity authentication risk level; the identity authentication risk level includes: a low risk level with a value of 0, and the corresponding system response is to allow direct verification; The medium risk level is assigned a value of 1, and the corresponding system response is to trigger a simple verification; The low risk level is assigned a value of 2, and the corresponding system response is to trigger multi-factor authentication; The low risk level is assigned a value of 3, and the corresponding system response is to deny login and notify the user to confirm; Model learning steps: Collect user feedback and perform online updates of the machine learning model.

10. The method according to claim 7, characterized in that The method further comprises: Detect whether there are obstacles around the ground lock to be driven. If there are obstacles, stop sending the driving signal to the actuator.

Images