Intelligent security method and device, electronic equipment and storage medium

Abstract

This application relates to the field of security technology, and in particular to an intelligent security method, device, electronic device, and storage medium. By responding to user configuration requests, it completes device binding and rule settings for various security modes, generates security mode policies, uploads these policies to the cloud, compiles them into a standardized instruction set adapted to the security gateway, and synchronizes them to the gateway's local storage. Upon receiving security mode instructions from users or automated triggers, it monitors the network status in real time, distinguishing between network-enabled and network-free scenarios. When network is available, it sends instructions from the cloud to the security gateway; when network is unavailable, it directly sends instructions to the gateway, which then retrieves the local standardized instruction set to control the security devices to perform corresponding security operations. This application supports refined, scenario-based security configuration, simplifies operation, enhances privacy protection capabilities, and simultaneously achieves dual-link control (cloud and local), effectively reducing strong dependence on the network and improving the operational stability and scenario adaptability of the intelligent security system.

Description

Technical Field

[0001] This application relates to the field of security technology, and in particular to an intelligent security method, device, electronic device, and storage medium. Background Technology

[0002] Existing smart security systems are generally equipped with user terminal APP, cloud server, security gateway, and various sensing devices such as door magnets, infrared sensors, and cameras. The typical workflow is that the user issues an arming or disarming control command through the user terminal APP. This command is forwarded to the security gateway through the cloud server, and then the security gateway controls the various sensing devices to switch to the arming or disarming working state, thereby realizing the home security monitoring function.

[0003] Specifically, existing smart security systems operate in a rigid manner, offering only two basic operational states: arming and disarming. This fails to meet the diverse and nuanced security needs of users in various daily living scenarios, such as when they are away from home, at home, or asleep. Starting, stopping, and switching states for different scenarios all require manual operation by the user, resulting in a cumbersome process and a poor user experience. Furthermore, privacy protection mechanisms are inadequate. In private scenarios such as nighttime rest or home activities, cameras and other image-capturing devices typically remain continuously operational, posing a significant risk of privacy leaks. While users can manually turn off the devices to mitigate this risk, manual operation is not only inconvenient but also prone to forgetting to perform actions, leading to insufficient reliability of privacy protection. Additionally, their core control logic and data processing heavily rely on cloud servers. The parsing, distribution, and device status adjustment of various control commands are all handled by the cloud, resulting in a strong network dependency. When the home network experiences interruptions, delays, or fluctuations, the communication link between the cloud and the local security gateway and sensors is lost, causing the system to be unable to perform core operations such as arming and disarming, effectively paralyzing the entire security system and resulting in low operational stability and emergency reliability. Summary of the Invention

[0004] To overcome the shortcomings of existing technologies, this invention provides an intelligent security method, device, electronic device, and storage medium that supports refined and scenario-based security configuration, simplifies operation, enhances privacy protection capabilities, and enables dual-link control via cloud and local connections. This effectively eliminates strong dependence on networks and improves the operational stability and scenario adaptability of the intelligent security system.

[0005] A first aspect of this application provides an intelligent security method, the method comprising: In response to the user's configuration request for each security mode, the system binds security devices and sets security rules for each security mode according to the configuration request, and generates a security mode policy. The security mode policy is uploaded to the cloud server so that the cloud server compiles the security mode policy into a standardized instruction set adapted to the security gateway and synchronizes the standardized instruction set to the local storage of the security gateway. In response to the security mode command triggered by the user or automation, detect the network status of the current operating environment; When the current operating environment is determined to be a network-enabled scenario based on the network status, the security mode instruction is sent to the security gateway through the cloud server, so that the security gateway reads the standardized instruction set and controls the security device to perform security operations according to the standardized instruction set; When the current operating environment is determined to be a network-free scenario based on the network status, the security mode command is sent to the security gateway, causing the security gateway to read the standardized instruction set and control the security device to perform security operations according to the standardized instruction set.

[0006] In an optional implementation, the method further includes: Acquire historical behavior data and generate security configuration optimization suggestions based on the historical behavior data; In response to the update confirmation command triggered by the user, the standardized instruction set is updated according to the security configuration optimization suggestions.

[0007] In an optional implementation, generating security configuration optimization suggestions based on the historical behavior data includes: When the amount of historical behavior data is sufficient, the historical behavior data is transformed into a user behavior feature vector; Based on the behavioral feature vector, personal habit mining is performed through unsupervised learning to identify the user's stable usage habits in different scenarios and obtain the user's corresponding personal habit pattern cluster. The security configuration optimization suggestions are generated based on the individual habit pattern clusters.

[0008] In an optional implementation, the step of generating security configuration optimization suggestions based on the historical behavior data further includes: When the amount of historical behavior data is insufficient, obtain a large amount of anonymized multi-user historical behavior data. Based on the multi-user historical behavior data, group learning is performed to establish a group habit knowledge base in different scenarios and generate group habit profiles corresponding to different user types. Based on the group habit knowledge base and group habit profile, transfer learning is performed to generate the security configuration optimization suggestions.

[0009] In an optional implementation, the step of binding security devices and setting security rules for each security mode according to the security mode request, and generating a security mode policy, includes: Scan and obtain all security devices within the local area network, and create a list of configurable devices; According to the configuration request, the security devices in the configurable device list are associated with or excluded from the corresponding security modes to obtain the security device group corresponding to each security mode; According to the configuration request, configure the corresponding security control parameters and execution strategies for each security mode; Each security mode, along with the corresponding security device group, security control parameters, and execution strategy, is integrated to generate the security mode strategy.

[0010] In an optional implementation, the method further includes: Real-time acquisition of multi-source scene perception data; The privacy risk level of the current scene is calculated based on the multi-source scene perception data; Based on the installation area and equipment type of the security equipment, it is divided into private area security equipment and public area security equipment; When the privacy risk level is higher than a preset threshold, the privacy protection strategy of the security device is dynamically adjusted based on the privacy risk level, and privacy protection operations are performed on the security device in the private area. When the privacy risk level is lower than or equal to the preset threshold, the standard security functions of the private area security device and the public area security device are restored; wherein, the adjustment of the privacy protection strategy is independent of the security mode strategy.

[0011] In an optional implementation, calculating the privacy risk level of the current scene based on the multi-source scene perception data includes: The privacy contribution value of each scene perception data in the multi-source scene perception data is quantified to obtain the single privacy contribution value corresponding to each scene perception data. The individual privacy contribution values ​​are weighted and fused based on preset weighting coefficients to obtain the total privacy risk score. The total privacy risk score is mapped to the corresponding privacy risk level.

[0012] A second aspect of this application provides an intelligent security device, the device comprising: The strategy configuration generation module is used to respond to the user's configuration request for each security mode, bind security devices and set security rules for each security mode according to the configuration request, and generate a security mode strategy. The policy synchronization storage module is used to upload the security mode policy to the cloud server, so that the cloud server can compile the security mode policy into a standardized instruction set adapted to the security gateway, and synchronize the standardized instruction set to the local storage of the security gateway. The network status detection module is used to detect the network status of the current operating environment in response to the security mode command triggered by the user or by automation. The first security execution module is used to send the security mode instruction to the security gateway through the cloud server when the current operating environment is determined to be a network scenario based on the network status, so that the security gateway reads the standardized instruction set and controls the security device to perform security operations according to the standardized instruction set; The second security execution module is used to send the security mode instruction to the security gateway when the current operating environment is determined to be a network-free scenario based on the network status, so that the security gateway reads the standardized instruction set and controls the security device to perform security operations according to the standardized instruction set.

[0013] A third aspect of this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the intelligent security method.

[0014] A fourth aspect of this application provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the above-described intelligent security method.

[0015] In summary, the intelligent security method, device, electronic device, and storage medium provided in this application have at least one of the following beneficial effects: 1. Supports users to customize corresponding security modes for various usage scenarios. It can complete the binding of security devices and the customization of security rules under different security modes according to the user's configuration request, and generate exclusive security mode strategies. It can accurately adapt to the diverse and refined daily residential security needs of users when they are away from home, at home, or sleeping. Users do not need to manually start, stop, or switch the status of various security devices one by one for different scenarios, which greatly simplifies the operation process and effectively improves the user experience. 2. Differentiated device operation rules can be configured based on different security modes, which can selectively turn off the working status of camera and other image acquisition devices in private scenarios such as nighttime rest and home activities. This mechanism avoids the risk of privacy information leakage in private scenarios without requiring users to manually turn off the devices, thus improving the convenience and reliability of privacy protection in intelligent security systems. 3. Upload the generated security mode strategy to a cloud server, which compiles the strategy into a standardized instruction set adapted for the security gateway and synchronously stores the instruction set locally on the security gateway, so that the local security gateway has independent security control execution capability. When responding to a security mode instruction, the system detects the current network status in real time and executes a dual-path control logic: in a networked scenario, instructions can be normally issued through the cloud, and the security gateway calls the local standardized instruction set to complete device control, ensuring normal linked operation in conventional scenarios; in non-network or abnormal network scenarios such as network interruption, delay and fluctuation, mode instructions can be issued directly to the security gateway without relying on the cloud server, and the regulation operation of security devices can be completed independently relying on the standardized instruction set stored locally on the security gateway. This effectively breaks the system's strong dependence on the network, avoids the problems that the security system is paralyzed and core security operations cannot be performed caused by network abnormalities, and improves the overall operation stability and emergency protection reliability of the intelligent security system. Description of Drawings

[0016] Figure 1 is a schematic structural diagram of an intelligent security system shown in an embodiment of the present application; Figure 2 is a schematic flow diagram of an intelligent security method shown in an embodiment of the present application; Figure 3 is another schematic flow diagram of an intelligent security method shown in an embodiment of the present application; Figure 4 is a schematic flow diagram of an intelligent configuration optimization method driven by an AI learning engine shown in an embodiment of the present application; Figure 5 is another schematic flow diagram of an intelligent configuration optimization method driven by an AI learning engine shown in an embodiment of the present application; Figure 6 is a module schematic diagram of an intelligent security device shown in an embodiment of the present application; Figure 7 is a structural schematic diagram of an electronic device shown in an embodiment of the present application. Detailed Description of the Embodiments

[0017] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0018] The following will clearly and completely describe the concept, specific structure, and technical effects of the present invention in conjunction with embodiments and accompanying drawings, so as to fully understand the purpose, features, and effects of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are all within the scope of protection of the present invention. Furthermore, all connections / linkages involved in the patent do not simply refer to direct contact between components, but rather to the ability to form a better connection structure by adding or reducing connecting accessories according to specific implementation conditions. The various technical features in this invention can be combined interactively without contradicting each other.

[0019] Furthermore, existing intelligent security systems lack sufficient intelligence, failing to possess autonomous learning and parameter optimization capabilities. Core operational parameters such as deployment delay time still require manual debugging and repeated calibration by users, making it impossible to autonomously iterate and optimize based on historical user behavior data and usage habits. Consequently, their adaptability and intelligence level fail to meet users' personalized needs. Simultaneously, the scene mode switching of existing similar technologies heavily relies on cloud-based linkage, and the corresponding device combinations for each mode are fixed and singular, lacking support for user-defined configurations. This results in poor functional scalability and scene adaptability, further limiting the practical application effectiveness and scope of application of intelligent security systems.

[0020] Reference Figure 1 The diagram shown is an architectural schematic of an intelligent security system according to an embodiment of this application. The intelligent security system includes an APP interaction module (set in the user's electronic device), a cloud configuration server, a security gateway, a device execution module (security devices, such as sensors / cameras / lights), and a newly added AI learning engine.

[0021] In some embodiments, users configure security modes through the APP interaction module, including binding security devices, setting arming rules, configuring alarm linkage and privacy policies, etc. After configuration, the APP interaction module uploads the user configuration data to the cloud configuration server. After receiving the configuration data, the cloud configuration server compiles the user-configured security mode policy into a standardized instruction set that the security gateway can directly parse and execute, and synchronizes the standardized instruction set to the local storage of the security gateway to complete the policy pre-synchronization. When a user triggers a security mode (such as the away-from-home mode) through the APP interaction module, or when the system automatically triggers a security mode according to preset conditions, the system first determines the network status of the current operating environment. When the network status is normal, the trigger signal is sent to the security gateway through the cloud configuration server. When the network status is abnormal (no network scenario), the APP interaction module can directly send a trigger command to the security gateway through Bluetooth, Wi-Fi Direct, or the security gateway can automatically trigger the security mode according to a preset timed task. After receiving the trigger signal, the security gateway does not need to request the cloud configuration server again. It directly reads the standardized instruction set corresponding to the current security mode from the local storage and controls each security device in the device execution module to perform preset operations according to the standardized instruction set, including arming or disarming, turning lights on or off, enabling or disabling cameras, etc., so as to realize the local independent execution of the security mode and ensure the availability of security functions in the event of network outage.

[0022] Meanwhile, the AI ​​learning engine running on the cloud configuration server continuously collects user behavior data, including user adjustment records of security mode parameters, alarm feedback operations, scene context data, etc. By constructing a two-layer model to mine personal habits and conduct group transfer learning, it generates security configuration optimization suggestions that conform to user habits and pushes them to the APP interaction module. When the user confirms the adoption of the optimization suggestions, the cloud configuration server automatically updates the security mode policy and recompiles and generates a new standardized instruction set, which is synchronized to the local storage of the security gateway, realizing dynamic updates and self-optimization of security configuration, forming a complete intelligent security closed loop.

[0023] Reference Figure 2 and Figure 3 The diagram shown is a flowchart illustrating an intelligent security method according to an embodiment of this application. The intelligent security method includes the following steps.

[0024] S21, in response to the user's configuration request for each security mode, perform security device binding and security rule settings for each security mode according to the configuration request, and generate a security mode policy.

[0025] In some embodiments, by performing refined device selection, rule customization, and structured policy integration for multiple security modes, the limitations of traditional security operation modes, such as rigidity, single device configuration, and weak scene adaptability, are overcome. This enables differentiated and customized configuration of security devices, alarm logic, and privacy policies in different life scenarios, laying a standardized policy data foundation for subsequent cloud-based compilation of instruction sets and independent offline execution by local gateways. Specifically, security mode policies can be constructed through device discovery, mode association, parameter configuration, and policy integration.

[0026] In an optional implementation, the step of binding security devices and setting security rules for each security mode according to the security mode request, and generating a security mode policy, includes: Scan and obtain all security devices within the local area network, and create a list of configurable devices; According to the configuration request, the security devices in the configurable device list are associated with or excluded from the corresponding security modes to obtain the security device group corresponding to each security mode; According to the configuration request, configure the corresponding security control parameters and execution strategies for each security mode; Each security mode, along with the corresponding security device group, security control parameters, and execution strategy, is integrated to generate the security mode strategy.

[0027] In some embodiments, the system scans security devices within a local area network (LAN) / wireless LAN, including door sensors, infrared sensors, smoke detectors, gas detectors, water leak detectors, cameras, sound and light alarms, and lights. All accessible devices are added to a detection device list for users to select and bind. Users can also create various security modes via the app's interactive module: Away from Home, At Home, Sleep, and Monitor, and select which devices to bind or exclude for each mode. For example, Away from Home mode binds all devices; At Home mode binds only environmental sensors and excludes indoor infrared / cameras; Sleep mode binds door sensors and vibration sensors and disables indoor cameras, etc.

[0028] For each security mode, arming delay parameters, alarm response rules (alarm volume / duration, alarm linkage actions), detection sensitivity parameters (trigger sensitivity), composite trigger conditions, effective time periods, and camera privacy control policies (camera on / off, recording on / off, microphone permissions), as well as spatiotemporal constraints (automatic mode activation time, effective area), are set to generate a security mode policy. The security mode policy includes at least the following: Device set and status rules: Which sensors / actuators are bound, and their expected states (e.g., Camera A - motion detection is on but recording is off; Light B - turns on in conjunction with the camera; Curtain C - closes).

[0029] Trigger-Response Rules: Detailed alarm rules are preset for each bound sensor, including: sensitivity parameters (pixel change threshold for motion detection, decibel threshold for sound), delay parameters (the delay time from triggering to alarm execution (e.g., arming delay in away-from-home mode)), composite conditions (multiple sensors must trigger sequentially within a specific time window to constitute a valid alarm (e.g., "door sensor first, then indoor infrared")), linked actions (after triggering, in addition to pushing the alarm, which actuators should be controlled (e.g., turning on all lights in the house, playing an alarm sound)), and spatiotemporal constraint rules (the effective time and spatial range of the configuration strategy. For example, the "sleep mode" configuration can be set to automatically take effect from 22:00 to 6:00 the next day, and only apply to the bedroom and entrance area).

[0030] Additionally, a pattern configuration knowledge base can be maintained in the cloud. Each pattern (such as "Standard Away from Home," "Late Night at Home," and "Home Theater") is a configuration template. These templates can be generated through expert rule bases and optimal configuration mining. Specifically, there are recommended configurations pre-set by security experts that conform to security theory (e.g., night mode should cover all external entrance sensors and increase their sensitivity). Alternatively, the system can anonymize and analyze massive amounts of user data, using clustering algorithms to find stable configuration schemes with the lowest false alarm rate adopted by most users in specific scenarios (such as "similar apartment layouts" and "families with pets"), and then consolidate these into optional recommended templates. When a user creates or modifies a pattern configuration, local simulation or conflict detection based on historical data can be performed. For example, the system might prompt the user, "You have bound the infrared sensor for 'frequent pet activity area' to 'Away from Home mode,' which may lead to high-frequency false alarms. It is recommended to adjust its sensitivity to 'low' or set it as an exclusion device." or "Your linkage action for 'fire alarm' includes 'shutting down the fresh air system,' which complies with safety regulations and has been automatically added for you." After the security mode policy is generated, it is uploaded to the cloud configuration server as a structured file (called the mode configuration policy file, such as JSON / XML format).

[0031] S22, the security mode policy is uploaded to the cloud server, so that the cloud server compiles the security mode policy into a standardized instruction set adapted to the security gateway, and synchronizes the standardized instruction set to the local storage of the security gateway.

[0032] In some embodiments, upon obtaining the pattern configuration policy file, the APP uploads the file to the cloud. The cloud server, through its compilation engine, translates (compiles) the pattern configuration policy file into a low-code instruction set (i.e., a standardized instruction set) that the security gateway's local CPU can directly and efficiently parse and execute. The standardized instruction set is essentially a list of sequential or conditionally jump-based instructions. For example, [ {"cmd": "DEVICE_SET", "id": "sensor_door", "state": "ARMED", "delay":30}, {"cmd": "DEVICE_SET", "id": "camera_living_room", "mode": "DETECT_ON_RECORD_OFF"}, {"cmd": "RULE_ADD", "trigger": "sensor_door:ALARM", "action": "siren:ON", "immediate": true}, {"cmd": "RULE_ADD", "trigger": "sensor_motion:ALARM", "condition":"time BETWEEN 22:00 AND 06:00", "action": ["light_corridor:ON","notification:USER"]} ] By encoding the mode configuration policy file into a standardized instruction set, the security control instructions are atomically split and the logic is reconstructed in a flattened manner. This removes the complex business judgment logic that originally relied on cloud servers, thereby enabling the security gateway to stably and independently execute security control operations even in network interruption and low computing power hardware environments.

[0033] Furthermore, the cloud server distributes the compiled standardized instruction set to the designated local security gateway, which can persistently store the standardized instruction set in local storage.

[0034] S23, in response to the security mode command triggered by the user or by automation, detect the network status of the current operating environment.

[0035] In some embodiments, when a user triggers a security mode (such as "away mode") via an app or automated scenario (such as a geofence), a corresponding security mode command is generated. In response to the security mode command, the network status of the current operating environment is monitored in real time to determine whether a normal communication connection is maintained between the security gateway and the cloud server.

[0036] S24, when the current operating environment is determined to be a network scenario based on the network status, the security mode instruction is sent to the security gateway through the cloud server, so that the security gateway reads the standardized instruction set and controls the security device to perform security operations according to the standardized instruction set.

[0037] When the detection results determine that the current operating environment is a network scenario, the APP sends a security mode command to the cloud server. The cloud server verifies the command and then forwards it to the corresponding security gateway. After receiving the security mode command from the cloud server, the security gateway does not need to request configuration information or execute logic from the cloud server. Instead, it directly reads the command content that matches the current security mode from the standardized command set stored locally and controls the corresponding security devices to perform preset security operations according to the standardized command set, including activating the corresponding sensor arming, turning off the specified camera, turning off the specified light, and activating alarm rules.

[0038] S25, when the current operating environment is determined to be a network-free scenario based on the network status, the security mode instruction is sent to the security gateway, so that the security gateway reads the standardized instruction set and controls the security device to perform security operations according to the standardized instruction set.

[0039] When the detection results determine that the current operating environment is a network-free scenario, the APP establishes a local connection with the security gateway directly through short-range communication methods such as Bluetooth or Wi-Fi Direct, and sends the security mode command directly to the security gateway; or the security gateway automatically triggers the corresponding security mode command according to a preset timed task. After receiving the command, the security gateway directly reads the standardized command set stored locally, and independently controls each security device to enter the preset working state according to the predefined device control logic, delay rules, linkage strategies, and privacy protection rules in the standardized command set, realizing stable arming, disarming, and scene linkage operations in a network-free environment.

[0040] When performing security operations, the security gateway does not need to request the cloud server. Instead, it directly loads the corresponding "away mode instruction set" locally and interprets and executes it line by line to control devices and rules. Specifically, It should be noted that, regardless of whether there is a network or not, the security gateway does not rely on real-time cloud computing and real-time command issuance. It completes all security control actions based solely on a locally pre-stored standardized instruction set, thereby achieving the technical effect of not losing connection when the network is disconnected and maintaining stable operation even with low computing power.

[0041] In an optional implementation, the method further includes: Obtain the user's historical behavior data and generate security configuration optimization suggestions based on the historical behavior data; In response to the update confirmation command triggered by the user, the standardized instruction set is updated according to the security configuration optimization suggestions.

[0042] Refer to together Figure 3 and Figure 4In some embodiments, historical adjustment records (i.e., historical behavior data) of all security mode parameters by the user on the APP are collected. This historical behavior data is multi-source, including: Explicit feedback data: Users manually modify specific parameter values ​​in a certain mode (e.g., change the arming delay of "away mode" from 30 seconds to 45 seconds).

[0043] Implicit feedback data: Recording user "cancel" or "execute immediately" actions during the period when a parameter's default value is in effect (such as manually skipping before the countdown ends) can be regarded as a signal of dissatisfaction with the default value.

[0044] Scene metadata: The mode, time, day of the week, and home device status (such as whether anyone is home when the adjustment is triggered) corresponding to each adjustment.

[0045] Data structure: Each record contains {User ID, Pattern Type, Parameter Name, Original Value, Adjusted Value, Timestamp, Scene Label}.

[0046] In addition, a two-layer model is constructed for learning from individuals and groups, including an individual habit model and a group transfer learning model.

[0047] (1) Personal habit model (unsupervised learning - clustering).

[0048] After acquiring historical behavior data, preprocessing and feature engineering are performed. First, the historical behavior data is cleaned to filter out invalid operations such as accidental triggers and repeated adjustments within short time intervals, retaining valid parameter adjustment records and alarm feedback records. Based on the cleaned valid data, time features are extracted, extracting periodic features such as time period type, weekday attribute, and weekday / rest day attribute from the data timestamps. Then, using user, security mode, and security parameters as combination dimensions, the historical adjustment data is grouped and aggregated according to the extracted time features to construct a dynamic feature vector for the "user-mode-parameter" combination. For example, user U, away from home mode, arming delay: {weekday_morning: 40 seconds, weekend_morning: 60 seconds, daily_night: 30 seconds, ...}.

[0049] For example, for user A's "Away Mode - Arming Delay" parameter, instead of recording a bunch of isolated values, a feature vector is generated as follows: Feature Vector_User A Away Delay = { "Time Period_Weekday_07:00-09:00": {"Number of Adjustments": 15, "Adjusted Value_Median": 35}, "Time Period_Weekend_09:00-12:00": {"Number of Adjustments": 8, "Adjusted Value_Median": 60}, "Time Period_Daily_2000-2400": {"Number of Adjustments": 20, "Adjusted Value_Median": 45, "Constantly Linked Devices": ["Corridor Lights"]}, "Context_Rainy Day": {"Number of Adjustments": 5, "Adjusted Value_Offset": +10} / / Rainy days tend to prolong the delay. } When constructing a personal habit model, Gaussian mixture model or spectral clustering is used as an unsupervised learning algorithm. Compared with conventional K-means clustering, it can better adapt to scenarios with complex and highly overlapping user security usage habits, improving the accuracy and stability of habit recognition. Specifically, the behavioral feature vector containing the user's various security mode parameters is input into the personal habit model. The model then performs cluster analysis on the user's parameter usage patterns in different time periods and scenarios, automatically identifying and dividing the user into multiple habit pattern clusters. Each cluster center corresponds to a stable user habit, and the model outputs the probability that the user's historical behavior data belongs to each habit pattern cluster. For example, for the arming delay parameter in the user's away-from-home mode, the model can obtain three habit clusters through clustering: Cluster A represents the habit of leaving home in a hurry, corresponding to the morning hours on weekdays, with a shorter deployment delay; Cluster B represents individuals who are accustomed to leaving home at a leisurely pace, corresponding to weekend mornings, resulting in a longer deployment delay. Cluster C is designed for enhanced protection at night, corresponding to the evening hours each day, with a moderate deployment delay and coordinated with the corridor lights.

[0050] (2) Group transfer learning model (group knowledge base and transfer learning).

[0051] When new user data is insufficient, intelligent initial recommendations are provided through a group transfer learning model. In the cloud, based on massive amounts of anonymized multi-user historical behavior data, group habit learning is conducted. Statistical analysis and pattern mining are performed on different user attributes, family structures, apartment layouts, and usage habits to establish a group habit knowledge base covering multiple scenarios and create group habit profiles corresponding to different user types. For example, Image P1: For families with large pets, the optimized stable value of the "sensitivity" parameter for indoor motion detection is generally distributed in the "low" to "medium" range.

[0052] Image P2: The "delay of leaving home" for "9-to-5 commuting families" shows a bimodal distribution on weekdays with a morning peak (short delay) and an evening peak (medium-to-long delay).

[0053] For newly registered users or users with insufficient historical behavior data to effectively build a personal habit model, transfer learning is performed based on the aforementioned group habit knowledge base and group habit profile. The optimal group-wide configuration matching the current user's attributes is then transferred to that user, generating cold-start security configuration optimization suggestions suitable for that user. This achieves intelligent initial configuration recommendations for new users in the data-free phase. For example, group learning can identify: Households with pets typically set indoor motion detection sensitivity to low to medium. Users who work 9-to-5 tend to use shorter arming delays when leaving home on weekday mornings.

[0054] When a new user registers and selects the corresponding family attribute or usage scenario, the above-mentioned common patterns of the group are directly used as the initial recommended configuration through transfer learning, so that the security configuration can be adapted to the user's usage habits without the need for manual adjustment.

[0055] Furthermore, by detecting new user behaviors, when a new adjustment occurs, the matching degree between this new adjustment and each cluster in the personal habit model is calculated to verify pattern stability. Under specific time characteristics, if N consecutive adjustments (e.g., 5 times) in similar contexts consistently deviate from the current default value and highly match a specific habit cluster, i.e., consistently fall within the range of a certain cluster, it is determined to be a "stable new habit," and the confidence level of the stable new habit is calculated (based on frequency, consistency, and difference from the old habit). When the confidence level exceeds a set threshold (e.g., 90%), the security configuration optimization suggestion generation process is triggered. That is, the AI ​​learning engine generates natural language suggestions with accompanying explanations to enhance credibility and user experience. For example, "In the past two weeks, when you activated 'Away from Home Mode' on weekday evenings, you manually changed the arming delay from the default 30 seconds to 45 seconds 8 times, and simultaneously turned on the 'linked corridor lights'; do you want to save this setting as a new habit of 'Away from Home on weekday nights'?" Once a user confirms and accepts the security configuration optimization suggestion, the parameter configuration file for that user's corresponding mode (e.g., Away Mode) is updated in the cloud. For example, a time-based overlay rule is added to the configuration: "Time range: Weekdays 20:00-24:00, Arming delay: 45 seconds, Linked device: Corridor light." The cloud configuration server immediately recompiles this updated, more complex configuration strategy into a new low-code instruction set specifically for "Away Mode," which is the standardized instruction set updated before the security mode strategy (including the security configuration optimization suggestion) was implemented. This new instruction set includes logic such as time condition checks, and its format is as follows: [ {"cmd": "CHECK_TIME", "range": "20:00-24:00", "weekdays": "1-5"}, {"cmd": "SET_PARAM", "param": "arm_delay", "value": 45}, {"cmd": "DEVICE_CTRL", "id": "light_corridor", "action": "ON", "after": "arm"} ] The new standardized instruction set is proactively distributed to the user's local security gateway via a configured synchronization channel. The gateway then updates its locally stored instruction set. Subsequently, when a user triggers the "away mode" during a weekday evening, the gateway will automatically execute the new logic of a 45-second delay and turning on the lights. The entire process requires no real-time cloud intervention, achieving controllable operation even when the network is offline.

[0056] Finally, a closed-loop learning mechanism is formed based on user feedback on security configuration optimization suggestions. User actions such as adopting, ignoring, or adopting modified suggestions are used as reinforcement learning signals fed back to the AI ​​learning engine to achieve iterative model optimization. Specifically, when a user adopts an optimization suggestion, the weight and confidence of the corresponding habit cluster are strengthened; when a user ignores an optimization suggestion, the weight of the corresponding habit cluster is reduced and the feature judgment rules are adjusted; when a user adopts an optimization suggestion after modification, new adjustment records are generated based on the modified parameters and added to the training dataset to retrain and update the learning model. Simultaneously, the application effect of parameter optimization is monitored in real time, and the false alarm rate and frequency of user manual intervention in corresponding scenarios are statistically analyzed. These indicators are used as evaluation criteria for optimization effectiveness, and the confidence threshold for pattern recognition is dynamically adjusted based on the evaluation results to continuously improve the accuracy and rationality of habit recognition and suggestion generation. When frequent adjustments to a certain security parameter are detected and a stable usage pattern is formed, corresponding parameter optimization suggestions are pushed to the user's terminal; after user confirmation, the configuration parameters of the corresponding security mode are automatically updated, achieving adaptive optimization of the security strategy. For example, when the AI ​​learning engine detects that a user frequently adjusts a certain parameter and forms a pattern, it will push optimization suggestions to the user through the APP (such as "It has been detected that you usually set the home arming delay to 35 seconds. Do you want to set this as the default value?"). After the user confirms, the configuration parameters of that mode will be automatically updated.

[0057] Through the aforementioned optional implementation methods, by collecting multi-source historical user behavior data and constructing personal habit models and group transfer learning models, and relying on data preprocessing, feature engineering, cluster analysis, and transfer learning mechanisms, the system accurately mines users' personalized security usage habits and general configuration patterns, effectively solving the shortcomings of traditional security systems that rely on manual parameter adjustment and lack autonomous adaptation capabilities. Through habit stability verification and confidence level determination, intelligent configuration optimization suggestions are generated. After user confirmation, the system automatically updates cloud configuration strategies and local standardized instruction sets, achieving adaptive iterative optimization of security parameters. Simultaneously, a reinforcement learning closed loop is constructed based on user feedback, dynamically adjusting model judgment thresholds to continuously reduce the system's false alarm rate and the frequency of manual user intervention. This balances cold start adaptation for new users and personalized adaptation for existing users, significantly improving the intelligence level, scenario adaptability, and overall user experience of the security system.

[0058] In an optional implementation, the method further includes: Real-time acquisition of multi-source scene perception data; The privacy risk level of the current scene is calculated based on the multi-source scene perception data; Based on the installation area and equipment type of the security equipment, it is divided into private area security equipment and public area security equipment; When the privacy risk level is higher than a preset threshold, the privacy protection strategy of the security device is dynamically adjusted based on the privacy risk level, and privacy protection operations are performed on the security device in the private area. When the privacy risk level is lower than or equal to the preset threshold, the standard security functions of the private area security device and the public area security device are restored; wherein, the adjustment of the privacy protection strategy is independent of the security mode strategy.

[0059] In some embodiments, during the multi-dimensional scene perception stage, electronic devices can collect scene data in real time from multiple information sources, including device status data, human presence and behavior data, environmental audio and video metadata, and user-defined tag data. Device status data includes the deadbolt status of smart door locks, the status of master bedroom lights, and electrical appliance power consumption data. Human presence and behavior data uses millimeter-wave radar in non-camera mode to detect the presence and static / active states of people in private areas such as bedrooms and bathrooms, determining whether the user is in a private scene such as sleeping or resting. Environmental audio and video metadata uses microphones to collect background sound levels, distinguishing different sound scenes such as silence, conversation, and media playback. Simultaneously, it uses the camera's low-power mode to analyze only the rate of change in the image, identifying private activity scenes such as changing clothes or washing up corresponding to high rates of change. User-defined tag data combines event information such as "private time" and "in a meeting" marked by the user in the calendar to assist in scene judgment.

[0060] Next, each type of scene perception data in the multi-source scene perception data is quantified and assigned a value, and the individual privacy contribution value of each type of data is calculated according to the scene privacy level corresponding to each type of data. Among them, different types of perception data are matched with different levels of individual privacy contribution values ​​based on their correlation with the scene privacy level. For example, for device status data, such as a smart door lock being "locked from the inside", it corresponds to a higher privacy contribution value; "unlocked" corresponds to a lower privacy contribution value; for human presence and behavior data, such as millimeter-wave radar detecting a stationary human body (sleeping state) in the bedroom area, it corresponds to a higher privacy contribution value; only detecting brief activity corresponds to a lower privacy contribution value; for environmental audio and video metadata, such as a microphone detecting an ambient background sound of "silence" or "low conversation", it corresponds to a higher privacy contribution value; detecting strong background sounds such as TV / music corresponds to a lower privacy contribution value; when the camera image change rate is in a high range, it corresponds to a higher privacy contribution value; for user-defined tag data, such as the current time being in the user-marked "private time" or "meeting", an additional privacy contribution value is added. Furthermore, based on preset or user-configured weighting coefficients, the privacy contribution values ​​of each feature are weighted and summed to obtain the total privacy risk score for the current scenario. The weighting coefficients can be differentiated according to scenario type (e.g., bedroom / living room) and user privacy preferences; for example, human detection data in the bedroom area has a higher weight than in the living room area. Finally, based on pre-defined score intervals and a one-to-one mapping relationship with gradient privacy risk levels, the total privacy risk score is mapped to discrete privacy risk levels (e.g., L1-L5, where L1 is low risk and L5 is high risk). The privacy risk level for the current scenario is periodically recalculated and updated based on real-time updates of multi-source data to ensure synchronization between the level and the scenario state. For example, when it is detected that someone is in the master bedroom late at night and is stationary, the ambient background noise is silent, and the smart door lock is locked from the inside, the privacy risk level for the current scenario is determined to be the highest level, L5. When the user leaves the bedroom and enters the living room, the door lock is released, and the ambient background noise returns to normal conversation, the privacy risk level will dynamically decrease to L1 or L2.

[0061] While executing preset security mode commands, the security gateway receives privacy risk level signals from the system in real time and dynamically overrides the original configuration policy based on the privacy risk level: when the privacy risk level is higher than a preset threshold (such as L3 and above), it automatically performs privacy protection operations on cameras in designated areas, including turning off video streams, enabling image blurring, or retaining only contour detection functions, while disabling microphone recording; when the privacy risk level drops below the preset threshold, it automatically restores the normal video capture and microphone recording functions of the cameras, ensuring the complete security capabilities of security devices in non-private scenarios. Simultaneously, users can set dynamic privacy protection sensitivity parameters through the APP interaction module, adjust the threshold for judging privacy risk levels, and also set whitelists for specific security devices to ensure that some devices never perform privacy protection operations, balancing privacy protection and security needs.

[0062] Through the above optional implementation methods, and through the scene-adaptive privacy protection mechanism of multi-source information fusion, the system can dynamically adjust the privacy protection strategy according to the micro-scene in which the user is actually located. This solves the problem that privacy protection and security functions are difficult to balance in the traditional static binding security mode, and realizes intelligent privacy protection that is more in line with the user's actual life scenario.

[0063] To facilitate understanding of the inventive concept, let's take the user configuring and using "Sleep Mode" as an example. During the configuration phase, the user accesses the "Sleep Mode" configuration page on the app. When device binding is required, in the device list, "Door / Window Sensor" and "Vibration Sensor" are selected, while "Bedroom Camera" is unselected. Simultaneously, "Bedroom Main Light" is selected to turn off the light. The arming delay time is set to "30 seconds," and the alarm sound is set to a "gentle beep" for 10 seconds. The configured security mode policy corresponding to "Sleep Mode" is uploaded to the cloud. The cloud server compiles "Sleep Mode" into an instruction set: "30-second delay → activate door / vibration sensor alarm → deactivate bedroom camera → deactivate bedroom main light," and immediately sends it to the security gateway for storage. During the execution phase, before going to sleep, the user clicks "Sleep Mode" on the app. The gateway receives the instruction, waits 30 seconds, and then automatically executes the local instruction set. After 30 seconds, the door / vibration sensor enters armed mode, the bedroom camera is powered off (for privacy protection), and the bedroom main light automatically turns off. At this point, even if the gateway network cable is disconnected, the system can still be armed normally. During the optimization phase, it was discovered that users adjusted the arming delay of "Sleep Mode" from the default 30 seconds to 40 seconds five times within a week. After analysis by the AI ​​learning engine, a push notification was sent to the app on the eighth day: "You seem to prefer the 40-second arming delay. Would you like to set it as the new default value for 'Sleep Mode'?" The user clicked "Yes," and Sleep Mode was subsequently triggered, with the delay time automatically changing to 40 seconds.

[0064] This application utilizes an End-Cloud-Gateway Collaborative Adaptive Intelligent Security Architecture (ECGA-ISA), a layered architecture that leverages customizable terminal app policy configurations, pre-synchronized cloud-based instruction compilation, and local offline execution at the gateway. Combined with a dual-model AI intelligent learning optimization mechanism and a multi-source scenario privacy risk dynamic control mechanism, it addresses the problems of network dependence, rigid configurations, poor adaptability, and inflexible privacy protection in traditional intelligent security systems. On one hand, by compiling user-defined security mode policies into standardized executable instruction sets and pre-synchronizing them to the security gateway, it enables independent and stable local operation of security functions in both network-connected and offline scenarios, resolving issues such as security function failure in offline environments and lag on low-computing-power devices. On the other hand, it constructs a dual-layer AI learning system consisting of a personal habit clustering learning model and a group transfer learning model to mine personalized user usage patterns and general security configuration characteristics of the group, forming a continuously iterative self-optimizing closed loop for security configuration. Simultaneously, based on multi-source scenario perception data, it dynamically quantifies privacy risk levels, achieving independent and adaptive control of security policies and privacy protection policies. This solution effectively improves the scene adaptability, operational stability, and intelligent autonomous iteration capability of the intelligent security system, while taking into account the reliability of device linkage security, personalized user needs, and scene-based privacy and security protection, greatly optimizing the user experience of whole-house intelligent security.

[0065] Reference Figure 6 The diagram shown is a functional block diagram of an intelligent security device according to an embodiment of this application.

[0066] In some embodiments, the intelligent security device 60 may include multiple functional modules composed of computer program segments. The computer programs for each program segment of the intelligent security device 60 may be stored in the memory of the electronic device and executed by at least one processor to perform (see details). Figure 2 (Description) The functions of intelligent security. Based on the functions it performs, it can be divided into multiple functional modules. These functional modules may include: a policy configuration generation module 601, a policy synchronization storage module 602, a network status detection module 603, a first security execution module 604, and a second security execution module 605. The module referred to in this application is a series of computer program segments that can be executed by at least one processor and perform a fixed function, stored in memory. In this embodiment, the functions of each module will be detailed in subsequent embodiments.

[0067] The strategy configuration generation module 601 is used to respond to the user's configuration request for each security mode, bind security devices and set security rules for each security mode according to the configuration request, and generate a security mode strategy.

[0068] The policy synchronization storage module 602 is used to upload the security mode policy to the cloud server, so that the cloud server compiles the security mode policy into a standardized instruction set adapted to the security gateway, and synchronizes the standardized instruction set to the local storage of the security gateway.

[0069] The network status detection module 603 is used to detect the network status of the current operating environment in response to the security mode command triggered by the user or by automation.

[0070] The first security execution module 604 is used to send the security mode instruction to the security gateway through the cloud server when the current operating environment is determined to be a network scenario based on the network status, so that the security gateway reads the standardized instruction set and controls the security device to perform security operations according to the standardized instruction set.

[0071] The second security execution module 605 is used to send the security mode instruction to the security gateway when the current operating environment is determined to be a no-network scenario based on the network status, so that the security gateway reads the standardized instruction set and controls the security device to perform security operations according to the standardized instruction set.

[0072] It should be understood that the various variations and specific embodiments of the intelligent security method provided in the above embodiments are also applicable to the intelligent security device of this embodiment. Through the foregoing detailed description of the intelligent security method, those skilled in the art can clearly understand the implementation method of the intelligent security device in this embodiment. For the sake of brevity, it will not be described in detail here.

[0073] See Figure 7 The diagram shown is a schematic representation of the structure of an electronic device according to an embodiment of this application. In a preferred embodiment of this application, the electronic device 7 includes a memory 71, at least one processor 72, and at least one communication bus 73.

[0074] Those skilled in the art should understand that Figure 7 The structure of the electronic device shown does not constitute a limitation of the embodiments of this application. It can be a bus structure or a star structure. The electronic device 7 may also include more or fewer other hardware or software than shown, or different component arrangements.

[0075] In some embodiments, the electronic device 7 is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions. Its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), programmable gate arrays (FPGAs), digital processors, and embedded devices. The electronic device 7 may also include user equipment, which includes, but is not limited to, any electronic product capable of human-computer interaction with a user via a keyboard, mouse, remote control, touchpad, or voice control device, such as a personal computer, tablet computer, smartphone, or digital camera.

[0076] In the embodiments provided in this application, it should be understood that the disclosed methods, apparatuses, computer-readable storage media, and electronic devices can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple components or modules may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices, components, or modules may be electrical, mechanical, or other forms.

[0077] The components described as separate parts may or may not be physically separate. The components shown as components may or may not be physical modules; that is, they may be located in one place or distributed across multiple network modules. Some or all of the components can be selected to achieve the purpose of this embodiment according to actual needs.

[0078] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing module, or each component can exist physically separately, or two or more modules can be integrated into one module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0079] If the integrated module is implemented as a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: USB flash drive, portable hard drive, read-only memory (ROM). Various media that can store program code, such as only memory, random access memory (RAM), magnetic disks or optical disks.

[0080] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0081] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0082] The above is a detailed description of the preferred embodiments of the present invention. However, the present invention is not limited to the embodiments described. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. All such equivalent modifications or substitutions are included within the scope defined by the claims of this application.

Claims

1. An intelligent security method, characterized in that, The method includes: In response to the user's configuration request for each security mode, the system binds security devices and sets security rules for each security mode according to the configuration request, and generates a security mode policy. The security mode policy is uploaded to the cloud server so that the cloud server compiles the security mode policy into a standardized instruction set adapted to the security gateway and synchronizes the standardized instruction set to the local storage of the security gateway. In response to the security mode command triggered by the user or automation, detect the network status of the current operating environment; When the current operating environment is determined to be a network-enabled scenario based on the network status, the security mode instruction is sent to the security gateway through the cloud server, so that the security gateway reads the standardized instruction set and controls the security device to perform security operations according to the standardized instruction set; When the current operating environment is determined to be a network-free scenario based on the network status, the security mode command is sent to the security gateway, causing the security gateway to read the standardized instruction set and control the security device to perform security operations according to the standardized instruction set.

2. The intelligent security method according to claim 1, characterized in that, The method further includes: Acquire historical behavior data and generate security configuration optimization suggestions based on the historical behavior data; In response to the update confirmation command triggered by the user, the standardized instruction set is updated according to the security configuration optimization suggestions.

3. The intelligent security method according to claim 2, characterized in that, The method of generating security configuration optimization suggestions based on the historical behavior data includes: When the amount of historical behavior data is sufficient, the historical behavior data is transformed into a user behavior feature vector; Based on the behavioral feature vector, personal habit mining is performed through unsupervised learning to identify the user's stable usage habits in different scenarios and obtain the user's corresponding personal habit pattern cluster. The security configuration optimization suggestions are generated based on the individual habit pattern clusters.

4. The intelligent security method according to claim 2, characterized in that, The method of generating security configuration optimization suggestions based on the historical behavior data also includes: When the amount of historical behavior data is insufficient, obtain a large amount of anonymized multi-user historical behavior data. Based on the multi-user historical behavior data, group learning is performed to establish a group habit knowledge base in different scenarios and generate group habit profiles corresponding to different user types. Based on the group habit knowledge base and group habit profile, transfer learning is performed to generate the security configuration optimization suggestions.

5. The intelligent security method according to claim 1, characterized in that, The step of binding security devices and setting security rules for each security mode according to the security mode request, and generating a security mode strategy, includes: Scan and obtain all security devices within the local area network, and create a list of configurable devices; According to the configuration request, the security devices in the configurable device list are associated with or excluded from the corresponding security modes to obtain the security device group corresponding to each security mode; According to the configuration request, configure the corresponding security control parameters and execution strategies for each security mode; Each security mode, along with the corresponding security device group, security control parameters, and execution strategy, is integrated to generate the security mode strategy.

6. The intelligent security method according to any one of claims 1 to 5, characterized in that, The method further includes: Real-time acquisition of multi-source scene perception data; The privacy risk level of the current scene is calculated based on the multi-source scene perception data; Based on the installation area and equipment type of the security equipment, it is divided into private area security equipment and public area security equipment; When the privacy risk level is higher than a preset threshold, the privacy protection strategy of the security device is dynamically adjusted based on the privacy risk level, and privacy protection operations are performed on the security device in the private area. When the privacy risk level is lower than or equal to the preset threshold, the standard security functions of the private area security device and the public area security device are restored; wherein, the adjustment of the privacy protection strategy is independent of the security mode strategy.

7. The intelligent security method according to claim 6, characterized in that, The calculation of the privacy risk level of the current scene based on the multi-source scene perception data includes: The privacy contribution value of each scene perception data in the multi-source scene perception data is quantified to obtain the single privacy contribution value corresponding to each scene perception data. The individual privacy contribution values ​​are weighted and fused based on preset weighting coefficients to obtain the total privacy risk score. The total privacy risk score is mapped to the corresponding privacy risk level.

8. An intelligent security device, characterized in that, The device includes: The strategy configuration generation module is used to respond to the user's configuration request for each security mode, bind security devices and set security rules for each security mode according to the configuration request, and generate a security mode strategy. The policy synchronization storage module is used to upload the security mode policy to the cloud server, so that the cloud server can compile the security mode policy into a standardized instruction set adapted to the security gateway, and synchronize the standardized instruction set to the local storage of the security gateway. The network status detection module is used to detect the network status of the current operating environment in response to the security mode command triggered by the user or by automation. The first security execution module is used to send the security mode instruction to the security gateway through the cloud server when the current operating environment is determined to be a network scenario based on the network status, so that the security gateway reads the standardized instruction set and controls the security device to perform security operations according to the standardized instruction set; The second security execution module is used to send the security mode instruction to the security gateway when the current operating environment is determined to be a network-free scenario based on the network status, so that the security gateway reads the standardized instruction set and controls the security device to perform security operations according to the standardized instruction set.

9. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the intelligent security method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the intelligent security method according to any one of claims 1 to 7.

Images