A control method and device of an intelligent door and window management system

Abstract

This application provides a control method and device for an intelligent door and window management system, applied in the field of intelligent door and window control technology. By collecting ambient light information, user schedule information, indoor sound characteristic information, and other intelligent device status information, it comprehensively analyzes and identifies the user scenario context, obtains door and window control decisions based on the scenario context, and monitors user manual operations to generate user preference correction instructions. Through user interaction and feedback mechanisms, it optimizes control decisions, which has the beneficial effects of automatically identifying user scenario context and intelligently controlling doors and windows, while learning user preferences through user interaction and feedback mechanisms, thereby improving the degree of automation and personalized adaptability.

Description

Technical Field

[0001] This application relates to the field of intelligent door and window control technology, and in particular to a control method and device for an intelligent door and window management system. Background Technology

[0002] With the rapid development of smart home technology, smart doors and windows, as an important component for improving living comfort and convenience, have gradually become widespread. However, existing smart door and window systems still face many technical challenges and user experience bottlenecks in practical applications.

[0003] On the one hand, current smart door and window products on the market generally suffer from insufficient automation. They often rely excessively on frequent manual operation or remote control by users, making it difficult to intelligently and flexibly adjust to complex changes in the indoor and outdoor environment and users' personalized lifestyles. For example, when the weather changes frequently or user activity scenarios change, users still need to actively intervene, which not only increases the operational burden but also prevents the intelligent advantages of smart doors and windows from being fully realized.

[0004] On the other hand, the sensing capabilities of existing smart door and window systems are relatively limited. Significant differences in communication protocols between different brands and models of smart doors and windows make it difficult to use interchangeable sensor components, thus limiting the system's ability to acquire diverse environmental information. Furthermore, the types of sensors currently equipped in smart doors and windows are relatively limited, mainly focusing on a few environmental factors such as wind and rain, failing to comprehensively perceive richer, multi-dimensional information such as ambient light, user schedules, indoor sound characteristics, and the status of other smart devices. This limitation in sensing capabilities prevents the system from accurately identifying the user's context, thus hindering the making of refined and personalized door and window control decisions.

[0005] Furthermore, existing systems also have shortcomings in user interaction and preference learning. When the system performs automated control, there is a lack of effective mechanisms to obtain immediate user feedback or learn long-term user preferences, which may lead to automated decisions that do not meet the actual needs of users, thereby affecting users' trust and satisfaction with the system. After users manually intervene in the system's automated control, the system often fails to transform this intervention into learning experience to optimize future control strategies.

[0006] In conclusion, existing smart door and window systems still have significant room for improvement in terms of automation, environmental perception, personalization, and user interaction, and are far from meeting users' ideal expectations for smart living.

[0007] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0008] In view of the shortcomings of the prior art, this application provides a control method and device for an intelligent door and window management system, which can automatically identify the user scenario context and intelligently control doors and windows, while learning user preferences through user interaction and feedback mechanisms to improve the degree of automation and personalized adaptability.

[0009] Firstly, a control method for an intelligent door and window management system, the method comprising the following steps:

[0010] S1: Collects ambient light information, user schedule information, indoor sound characteristics information, and other smart device status information;

[0011] S2: By comprehensively analyzing the ambient light information, user schedule information, indoor sound feature information, and other smart device status information, the user's scene context is identified;

[0012] S3: Based on the scene context, obtain door and window control decisions from the stored scene-related preset behavior rules;

[0013] S4: Monitor user manual operations on doors and windows. If manual operations are detected, generate user preference correction instructions associated with the scene context and execute the user preference correction instructions first.

[0014] S5: Before executing the door and window control decision or the user preference correction instruction, issue a suggestion or inquiry to the user and set a preset response time window;

[0015] S6: Listen for user feedback within the preset response time window. If user feedback is present, the door and window control decision is corrected immediately based on the feedback. If no user feedback is present, the door and window actuator is driven to complete the corresponding action according to the target door and window state determined by the door and window control decision or the user preference correction instruction, or the current door and window state is kept unchanged.

[0016] Furthermore, step S2 includes:

[0017] S21: Set the user schedule information and the ambient light information as the first priority information, and set the indoor sound feature information and other smart device status information as the second priority information; other smart devices include at least a smart thermostat, an air purifier, a smart TV, and a smart speaker;

[0018] S22: Based on the first priority information, perform a first-level scene classification, which includes at least sleep-related scenes, work-related scenes, and daily activity scenes.

[0019] S23: Combine the second priority information and the result of the first-level scene classification to perform a second-level scene judgment and obtain the scene context.

[0020] Furthermore, step S23 includes:

[0021] S231: Under the first-level scene classification result of the sleep-related scene, if the indoor sound feature information is identified as quiet and the status information of other smart devices shows that the device is in the off state, then the scene context is finally identified as a quiet sleep scene.

[0022] S232: Under the first-level scene classification result of the work-related scene, if the indoor sound feature information is identified as a slight human voice but no obvious entertainment noise, and the status information of other smart devices shows that the device is in the off state, then the scene context is finally identified as a focused work scene.

[0023] S233: Under the first-level scene classification result of the daily activity scenario, if the indoor sound feature information is identified as a sound pattern with high loudness and wide frequency distribution, and the status information of other smart devices shows that the smart TV or smart speaker is on, then the scene context is finally identified as a home entertainment scenario.

[0024] Furthermore, step S3 includes:

[0025] S31: Identify all preset behavior rules associated with the scene context and doors and windows;

[0026] S32: Evaluate the conditions of the preset behavior rules and determine the set of preset behavior rules that meet the conditions;

[0027] S33: If there are multiple preset behavior rules in the set of preset behavior rules that meet the conditions, and their corresponding door and window control decisions conflict with each other, then conflict resolution is performed. The conflict resolution includes:

[0028] Based on the preset rule priority order, select the door and window control decision corresponding to the rule with the higher priority from the conflicting door and window control decisions;

[0029] If the conditions of multiple conflicting rules with the same priority are all met, the preset default door and window control decision is selected.

[0030] The door and window control decision determined after conflict resolution or in the absence of conflict is taken as the door and window control decision.

[0031] Furthermore, step S4 includes:

[0032] S41: When the doors and windows move automatically, detect the user's manual operation on the doors and windows;

[0033] S42: If a manual operation is detected, obtain the current position of the door / window when the manual operation begins, the direction of the manual operation, and the final position of the door / window when the manual operation ends.

[0034] S43: Obtain the target position for the automatic movement of the doors and windows, and compare the final position with the target position;

[0035] S44: If the direction of the manual operation is opposite to the direction of the automatic movement, or if the deviation between the final position and the target position exceeds a preset threshold, a first type of user preference correction instruction is generated.

[0036] S45: If the deviation between the final position and the target position does not exceed a preset threshold, then generate a second type of user preference correction instruction;

[0037] S46: Associate the first type of user preference correction instruction or the second type of user preference correction instruction with the scenario context, and execute the first type of user preference correction instruction or the second type of user preference correction instruction first.

[0038] Furthermore, step S46 includes:

[0039] S461: Store the first type of user preference correction instruction or the second type of user preference correction instruction and the scene context as key-value pairs in the rule database;

[0040] S462: Set the priority flag for the correction instructions in the stored key-value pairs;

[0041] S464: When making decisions on door and window control, retrieve correction instructions with priority identifiers from the rule database based on the current scene context;

[0042] S464: Apply the retrieved correction instruction with priority identifier to execute either the first type of user preference correction instruction or the second type of user preference correction instruction.

[0043] Furthermore, step S5 includes:

[0044] S51: Before executing the door and window control decision or the user preference correction instruction, obtain the urgency information of the current scenario and the user's historical response pattern information;

[0045] S52: Based on the current scene context, determine a set of interaction channels for issuing suggestions or inquiries to the user, wherein the set of interaction channels includes at least one of smart display screen, voice broadcast and mobile application push;

[0046] S53: Calculate the duration of the preset response time window based on the urgency information of the current scenario and the user's historical response pattern information.

[0047] Furthermore, within the preset response time window, user feedback is monitored. If user feedback is received, the door and window control decision is adjusted in real time based on the feedback, including:

[0048] S61: Detects the force, direction, and duration of manual pushing and pulling by the user;

[0049] S62: Determine the type and degree of correction of the user feedback based on the force, direction and duration;

[0050] S63: Convert the determined correction type and correction degree into incremental or absolute position adjustment instructions for doors and windows, and execute the adjustment instructions.

[0051] Furthermore, in step S6, if there is no user feedback, the door and window actuator is driven to complete the corresponding action according to the target door and window state determined by the door and window control decision or the user preference correction instruction, or the current door and window state is kept unchanged, including:

[0052] S64: Obtain the target door / window status determined by the door / window control decision or the user preference correction instruction;

[0053] S65: Determine the execution priority of the door and window control decision or the user preference correction instruction based on the type of the door and window control decision or the user preference correction instruction;

[0054] S66: Based on the execution priority, adjust the operating speed of the door and window actuator, and drive the door and window to reach the target door and window state or keep the current door and window state unchanged.

[0055] Secondly, a control device for an intelligent door and window management system, used to implement any of the methods described above, the device comprising:

[0056] Data acquisition module: Collects ambient light information, user schedule information, indoor sound characteristics information, and other smart device status information;

[0057] Recognition module: By comprehensively analyzing the ambient light information, user schedule information, indoor sound feature information, and other smart device status information, the recognition module identifies the user's scene context.

[0058] Acquisition module: Based on the scene context, acquire door and window control decisions from the stored scene-related preset behavior rules;

[0059] Generation module: Monitors user manual operations on doors and windows. If manual operations are detected, it generates user preference correction instructions associated with the scene context and executes the user preference correction instructions first.

[0060] Inquiry module: Before executing the door and window control decision or the user preference correction instruction, it issues suggestions or inquiries to the user and sets a preset response time window;

[0061] Execution module: Listens for user feedback within the preset response time window. If user feedback is present, the door and window control decision is corrected in real time based on the feedback. If no user feedback is present, the door and window actuator is driven to complete the corresponding action according to the target door and window state determined by the door and window control decision or the user preference correction instruction, or the current door and window state is kept unchanged.

[0062] Beneficial Effects: The control method and device for an intelligent door and window management system proposed in this application collects ambient light information, user schedule information, indoor sound characteristic information, and other intelligent device status information, comprehensively analyzes and identifies the user scenario context, obtains door and window control decisions based on the scenario context, monitors user manual operations to generate user preference correction instructions, and optimizes control decisions through user interaction and feedback mechanisms. It has the beneficial effects of automatically identifying user scenario context and intelligently controlling doors and windows, while learning user preferences through user interaction and feedback mechanisms, thereby improving the degree of automation and personalized adaptability. Attached Figure Description

[0063] Figure 1 This is a flowchart of a control method for an intelligent door and window management system proposed in this application.

[0064] Figure 2 This is a structural diagram of the control device for an intelligent door and window management system proposed in this application.

[0065] Figure 3 This is a structural diagram of an intelligent door and window management system proposed in this application.

[0066] Labeling Explanation: 201, Acquisition Module; 202, Identification Module; 203, Acquisition Module; 204, Generation Module; 205, Inquiry Module; 206, Execution Module. Detailed Implementation

[0067] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. The components of the embodiments of this application described and marked in the accompanying drawings can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0068] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0069] Please refer to Figure 1 A control method for an intelligent door and window management system, the method comprising the following steps:

[0070] S1: Collects ambient light information, user schedule information, indoor sound characteristics information, and other smart device status information;

[0071] S2: Comprehensively analyze ambient light information, user schedule information, indoor sound characteristics information, and other smart device status information to identify the user's scene context;

[0072] S3: Based on the scene context, obtain door and window control decisions from the stored scene-related preset behavior rules;

[0073] S4: Monitor user manual operations on doors and windows. If manual operations are detected, generate user preference correction instructions that are associated with the scene context and execute the user preference correction instructions first.

[0074] S5: Before executing door and window control decisions or user preference correction instructions, issue suggestions or inquiries to the user and set a preset response time window;

[0075] S6: Listen for user feedback within the preset response time window. If user feedback is received, the door and window control decision is corrected in real time based on the feedback. If no user feedback is received, the door and window actuator is driven to complete the corresponding action according to the target door and window state determined by the door and window control decision or user preference correction instruction, or the current door and window state is kept unchanged.

[0076] Please refer to Figure 3 The intelligent door and window management system of this application relies on a general signal platform as the core hub, connects various environmental sensors and controls doors and windows of different brands, and realizes intelligent linkage control across devices.

[0077] The general signal platform acquires ambient light information through indoor light sensors. This information includes not only light intensity but also potentially color temperature, used to determine whether it is daytime, dusk, or nighttime, and whether it is sunny or cloudy. Simultaneously, the platform connects to the user's electronic calendar or schedule management application to obtain schedule information such as scheduled meeting times, rest periods, or outings. To gain a deeper understanding of indoor activities, the platform also uses a microphone array to collect indoor sound characteristics. By analyzing the loudness, frequency distribution, and patterns of sound, it determines whether the room is quiet, whether people are talking, or whether music or a movie is playing. Finally, the platform connects to other smart devices via the home network, such as smart thermostats, air purifiers, smart TVs, and smart speakers, to obtain their status information, such as whether the air conditioner is on or the TV is playing a program. This collection of four types of information forms the basis for all subsequent intelligent decisions. By constructing a multi-dimensional information input, the smart door and window management system can break free from its reliance on a single sensor, thus providing a more comprehensive and accurate picture of the user's current environment and activities.

[0078] After acquiring rich environmental and status information, the intelligent door and window management system integrates and analyzes the four types of information collected to identify the specific context of the user's current situation. Context is a highly generalized description of the user's current state and environment, such as a quiet sleep scenario, a focused work scenario, a home theater scenario, or a scenario of being away from home. This identification process is not a simple listing of information, but an intelligent judgment process based on logical reasoning and pattern matching. For example, when a general signal platform detects that the time is midnight, the light intensity is extremely low, the schedule shows sleep time, the indoor sound is quiet, and all smart lights and the television are turned off, it can determine with high confidence that the current context is quiet sleep. This scene recognition capability based on multi-source information fusion is key to achieving advanced automation, enabling the intelligent door and window management system to truly understand the user's situation and potential needs, rather than simply reacting to isolated physical parameters.

[0079] Once the current scene context is identified, the General Signal platform maintains a rule database containing a large number of pre-stored behavioral rules. Each rule associates a specific scene context with one or more door and window control actions. For example, one rule might be: if the scene context is a focused work scenario, adjust the study's curtains to a half-open state to ensure sufficient natural light while avoiding screen glare. Another rule might be: if the scene context is an away-from-home scenario, close all windows and draw all curtains to ensure security and privacy. When the intelligent door and window management system identifies the current scene, it queries this rule database, finds the rule matching the current scene, and extracts the defined door and window control decisions. This decision is an initial action plan automatically generated by the General Signal platform.

[0080] Furthermore, this intelligent door and window management system possesses the ability to learn and adapt to user personalities, specifically by monitoring user manual operations and generating preference corrections. The system continuously monitors the status of doors and windows through a universal signaling platform, particularly when users manually intervene in doors and windows that are automatically executing or already in a certain state. For example, if the system opens the curtains to 50% according to the rules for a focused work scenario, and the user manually pulls the curtains to 70%, the universal signaling platform will detect this action. The platform will record this manual action and generate a user preference correction instruction. This instruction will be associated with the current scenario context, namely the focused work scenario. More importantly, this user preference-generated correction instruction will have higher priority in future decisions. This means that the next time the intelligent door and window management system identifies a focused work scenario, it will prioritize executing the user-preferred instruction to open the curtains to 70%, rather than the original 50% as defined in the rules. Through this learning mechanism, the intelligent door and window management system can continuously fine-tune its control logic based on the user's actual behavior, making the automated behavior of doors and windows increasingly aligned with the user's personal habits and preferences.

[0081] The specific technical means for generating user preference correction instructions employs a lookup table method and incremental updates. The system internally maintains a non-relational database based on JSON format. When it detects that a user manually adjusts the curtains from an automatically set 50% to 80% while focused on work, the system generates the following key-value pair record:

[0082] {"Scene":"Work_Focus","Device":"Curtain_Main","Override_Value":80,"Timestamp":Now()}. Here, Scene represents a unique identifier for the scene context; Work_Focus represents a focused work scene; Device represents the unique hardware ID or logical name of the specific smart door / window device manually operated by the user; Curtain_Main refers to the main curtains in the study; Override_Value is a correction value, representing the final target state parameter determined by the user's manual operation, i.e., the quantified value of the user's preference. In this example, 80 indicates that the user adjusted the curtain opening degree to 80%; Timestamp is a timestamp, representing the time point when the manual operation occurred or the instruction was generated, obtained by the Now() function; Now() is a function to obtain the current time point.

[0083] The next time the system enters the scene, it reads the key-value pair and calculates the weighted moving average: This allows for the gradual learning and approximation of preferences. Among these, The updated target locations for doors and windows. This refers to the historical preference positions stored in the database (i.e., the target value learned last time). This is the final position of the user's manual operation.

[0084] To strike a balance between automation and user control, and to prevent the smart window and door management system from making unexpected operations that users do not expect, this method involves interaction before making a decision. The system does not act immediately before executing an automatically acquired window and door control decision or a corrective instruction generated based on learned user preferences. Instead, it first sends a suggestion or question to the user through a universal signaling platform. For example, it might announce via a smart speaker: "We've detected you're working; would you like to adjust the study curtains to a brighter position?" or push a message to the user's mobile app. Simultaneously with issuing the question, the universal signaling platform initiates a preset response time window, such as thirty seconds. This design gives the user the final veto or confirmation power over the system's decisions, significantly increasing user trust and acceptance of the smart window and door management system.

[0085] If the user provides clear feedback via voice, mobile app, or direct manual operation of doors and windows—such as answering "okay" or manually adjusting the doors and windows—the smart door and window management system will immediately correct its original control decisions and execute them based on this feedback. For example, if the user answers "make it brighter," the system will further open the curtains based on the original decision. If the user does not provide any feedback within this time window, the smart door and window management system will assume that the user has tacitly agreed to the upcoming operation. At this point, the smart door and window management system will drive the door and window actuators, such as motors, according to the original door and window control decisions or user preference correction instructions to complete the corresponding opening, closing, or angle adjustment actions, or maintain the current judgment state. This complete closed-loop process, from information collection, scene recognition, decision generation, user preference learning, to interaction confirmation and final execution, constitutes a smart door and window management system that can proactively serve, continuously learn, and respect user wishes.

[0086] Furthermore, step S2 includes:

[0087] S21: Set user schedule information and ambient light information as the first priority information, and set indoor sound characteristic information and other smart device status information as the second priority information; other smart devices include at least smart thermostats, air purifiers, smart TVs, and smart speakers;

[0088] S22: Classify the first-level scenarios based on the first priority information. The first-level scenario classification includes at least sleep-related scenarios, work-related scenarios, and daily activity scenarios.

[0089] S23: Combine the second priority information and the results of the first-level scene classification to make a second-level scene judgment and obtain the scene context.

[0090] First, the general signal platform categorizes various information into two priorities. User schedule information, such as sleep or work meeting entries in the calendar, and ambient light information, such as the presence and intensity of sunlight, are set as the first priority. This is because these two types of information typically provide macroscopic, stable, and decisive evidence regarding the user's main activity status and basic time periods of the day. For example, sleep time on the schedule and actual darkness in the environment are strong evidence that the user has entered a sleep state. Subtle changes in indoor sound or the on / off status of other smart devices are set as the second priority information. These other smart devices include at least the aforementioned smart thermostat, air purifier, smart TV, and smart speaker. While this information is more dynamic and detailed, it may be ambiguous when viewed individually. Using it as an auxiliary basis for judgment can avoid misjudgment of the scene due to temporary noise or changes in device status.

[0091] After prioritizing information, the intelligent door and window management system uses a general signal platform to perform a first-level scene classification based on the highest priority information. This is a coarse-grained classification process, designed to quickly categorize the current state into several broad categories. For example, if the general signal platform obtains schedule information from 11 PM to 7 AM the next morning, and the ambient light sensor detects very weak indoor and outdoor light, the intelligent door and window management system will initially classify the current scene as a sleep-related scenario. Similarly, if the schedule information shows working hours on a weekday and the lighting is normal, it may be classified as a work-related scenario. Other time periods may be broadly classified as daily activity scenarios. This first-level classification provides a clear framework and direction for subsequent refined judgments, greatly narrowing the search scope and improving processing efficiency.

[0092] After completing the first-level scene classification, the intelligent door and window management system proceeds to the second-level scene judgment. Combining second-priority information—indoor sound characteristics and the status of other smart devices—it further confirms and refines the first-level classification results. In this way, the intelligent door and window management system can identify more specific and precise scene contexts from macro-level scene categories. For example, even within the first-level classification of daily activities, if the second-priority information indicates high indoor sound volume, a wide frequency distribution, and the smart TV is on, the intelligent door and window management system can further refine the scene context to a home entertainment scene. Conversely, if the indoor sound is quiet, with only the slight sound of turning pages and only a desk lamp on, it might be judged as a reading and learning scene. This layer-by-layer judgment logic from macro to micro ensures that the final scene context is both accurate and detailed, providing a high-quality basis for subsequent door and window control decisions.

[0093] Furthermore, step S23 includes:

[0094] S231: In the first-level scene classification result of sleep-related scenarios, if the indoor sound feature information is identified as quiet and the status information of other smart devices shows that the devices are in the off state, then the final scene context is identified as a quiet sleep scenario.

[0095] S232: In the first-level scene classification result of work-related scenarios, if the indoor sound feature information is identified as a slight human voice but no obvious entertainment noise, and the status information of other smart devices shows that the devices are in the off state, then the final scene context is identified as a focused work scenario.

[0096] S233: In the first-level scene classification results of daily activity scenarios, if the indoor sound feature information is identified as a sound pattern with high loudness and wide frequency distribution, and the status information of other smart devices shows that the smart TV or smart speaker is on, then the final identified scene context is a home entertainment scene.

[0097] To more clearly illustrate the specific logic of the second-level scene determination, several typical examples can be listed. The steps for determining the second-level scene by combining the second priority information and the first-level scene classification results can include the following situations:

[0098] In the first-level scene classification results for sleep-related scenarios, if the indoor sound characteristics are identified as quiet, and the status information of other smart devices shows that the devices are turned off, then the final scene context is identified as a quiet sleep scenario. In this embodiment, it is assumed that the smart door and window management system has initially classified the current scene as a sleep-related scenario by analyzing the user's schedule and ambient light information. At this time, the smart door and window management system will further analyze the second priority information. The microphone deployed in the bedroom detects that the decibel value of the ambient sound is consistently below a preset quiet threshold, such as 35 decibels, and the sound spectrum analysis shows that there is no obvious human voice or device operating noise. At the same time, the smart door and window management system queries the general signal platform and finds that the smart TV, smart speaker, reading light, and other devices in the bedroom are all in a turned-off or standby state. Combining this information, the smart door and window management system can make a final judgment and accurately identify the current scene context as a quiet sleep scenario. For this scenario, the preset door and window control decision may be to completely close the windows to isolate external noise and draw the blackout curtains to create the most ideal dark sleep environment. If, in a sleep-related scenario, the indoor sound characteristics are identified as not quiet (≥35 dB or with obvious human voices / music), or any non-window smart device is turned on, it is identified as a light sleep / sleep-aid scenario. The corresponding window and door decisions could be: close the curtains 80% and leave the windows with a 5% gap for ventilation.

[0099] In the first-level scenario classification of work-related scenarios, if the indoor sound characteristics are identified as faint human voices but without significant entertainment noise, and other smart device status information indicates that the devices are off, then the final scenario context is identified as a focused work scenario. In this embodiment, the smart door and window management system has determined that the user is in a work-related scenario, such as a home office, through the first-level classification. Next, the smart door and window management system analyzes the second priority information. Low-decibel, discontinuous human voices are identified in the sound signals collected by the microphone, which may correspond to the user making a voice call or talking to themselves. However, the analyzer used in the general signal platform to identify music or movie sound patterns is not triggered, indicating the absence of significant entertainment noise. Simultaneously, the smart door and window management system confirms that entertainment devices such as smart TVs and game consoles in the room are off, while the user's computer or desk lamp may be on. Based on this refined information, the smart door and window management system can refine the scenario context from broad work-related to a focused work scenario. In this scenario, the intelligent door and window management system might make the following door and window control decisions: adjust the curtains to an angle that prevents direct sunlight from hitting the computer screen, while moderately opening the windows to maintain air circulation, creating a bright yet undisturbed working environment for the user. However, if the indoor sound characteristics are identified as entertainment noise (music, movie dialogue, etc.) or entertainment devices (TV, game console, speakers) are on in a work-related scenario, it is identified as a leisure office environment. The corresponding door and window decisions might be: lower the curtains to 30% to reduce screen glare, and keep the windows closed to reduce external noise.

[0100] In the first-level scene classification of daily activity scenarios, if the indoor sound feature information is identified as a high-loudness sound pattern with a wide frequency distribution, and the status information of other smart devices shows that the smart TV or smart speaker is on, then the final scene context is identified as a home entertainment scenario. In this embodiment, the first-level classification categorizes the scene as a daily activity scenario. The smart door and window management system then analyzes the second priority information and finds that the average loudness of the indoor sound is high, and the spectrum analysis shows that the sound signal covers a wide range from low to high frequencies, which is highly consistent with the sound effects of movies, music, or games. At the same time, the smart door and window management system finds that the smart TV or high-powered smart speaker in the living room is in working condition. Combining these two pieces of information, the smart door and window management system can accurately determine that the user is engaged in entertainment activities, thus ultimately identifying the scene context as a home entertainment scenario. For this scenario, the smart door and window management system may automatically draw the curtains to reduce the impact of ambient light on the screen viewing effect, and may fine-tune the opening and closing of the windows according to the sound intensity to achieve a balance between ensuring ventilation and optimizing indoor acoustics. If, in a daily activity scenario, the indoor sound characteristics are identified as low volume and concentrated frequency (such as only human voices talking), and entertainment devices are turned off, then it is identified as a home leisure-conversation scenario. The corresponding door and window decisions could be: open the curtains 60% to ensure natural light, and open the windows 20% to promote air circulation.

[0101] The scenarios described above are merely examples. In practical applications, the final recognizable scenario context can also include afternoon home leisure scenarios, dining and ventilation scenarios, and inclement weather avoidance scenarios, among others. Taking the afternoon home leisure scenario as an example, if the system detects strong natural light of 30,000–60,000 lux and a color temperature of 5,500–6,500 K through a light sensor, combined with the user's calendar showing no meetings or outings between 12:00 and 15:00, and the microphone array picking up light music or page-turning sounds at 40–55 dB without significant audio effects, while the smart TV is in standby mode, the smart speaker is playing light music, and the air conditioner is set to 26°C, then the scenario context can be refined to an afternoon home leisure scenario, triggering a ventilation and lighting strategy of opening the curtains by 70% and slightly opening the windows by 15%. Similarly, ventilation scenarios during group meals can be identified by a combination of features, including a meal schedule of 18:00–20:30, 60–75 dB of conversation and tableware sounds, the presence of a range hood and kitchen lights, and ambient lighting in the living room. Severe weather avoidance scenarios are determined by a combination of factors, including rainstorm / strong wind / dust warnings from meteorological APIs, triggering of wind and rain sensors, a sudden drop in light intensity below 1000 lux, and automatic acceleration of air purifiers. The system then immediately closes and locks all doors and windows, closes all curtains, and switches the air conditioner to internal circulation, achieving rapid avoidance. The above feature combinations are only typical examples; in engineering implementation, they can be flexibly expanded using rule bases or machine learning models to cover more personalized scenarios.

[0102] Furthermore, step S3 includes:

[0103] S31: Identify all preset behavior rules associated with scene context and doors and windows;

[0104] S32: Evaluate the conditions of the preset behavior rules and determine the set of preset behavior rules that meet the conditions;

[0105] S33: If there are multiple preset behavior rules in the set of preset behavior rules that meet the conditions, and their corresponding door and window control decisions conflict with each other, then conflict resolution is performed. Conflict resolution includes:

[0106] Based on the preset rule priority order, select the door and window control decision corresponding to the rule with the higher priority from the conflicting door and window control decisions;

[0107] If the conditions of multiple conflicting rules with the same priority are all met, the preset default door and window control decision is selected.

[0108] The window and door control decisions determined after conflict resolution or in the absence of conflict shall be used as the window and door control decisions.

[0109] Taking an afternoon leisure scenario at home as an example, once the scenario context is identified, the smart door and window management system will search its rule database (a structured knowledge base within the smart door and window management system used to store, index, and call scenario-behavior mapping rules) to find all preset behavior rules associated with this scenario and the doors and windows that need to be controlled, such as the living room's floor-to-ceiling windows. This may find multiple rules, such as rule A: open the living room windows for ventilation during an afternoon leisure time; rule B: close all windows when the outdoor air quality index is higher than 150; and rule C: open windows to assist in cooling when the indoor temperature is higher than 28 degrees Celsius.

[0110] The intelligent door and window management system then checks whether the triggering conditions for each rule are met in the current environment. Assuming the current outdoor air quality index is 180 and the indoor temperature is 29 degrees Celsius, then the condition for afternoon leisure at home is met; the condition for outdoor air quality index above 150 is also met; and the condition for indoor temperature above 28 degrees Celsius is also met. Therefore, these three rules together constitute a set of preset behavioral rules that satisfy the conditions.

[0111] At this point, the intelligent window and door management system faces a decision conflict: rules A and C require opening the window, while rule B requires closing the window. To resolve this conflict, the intelligent window and door management system activates a conflict resolution mechanism. The core of this mechanism is rule priority. At the initial design stage of the intelligent window and door management system, each rule is assigned a priority value. Typically, rules related to health and safety have the highest priority. In this example, rule B, which ensures indoor air quality, has a higher priority than rules A and C, which aim at comfort. Therefore, in the first step of conflict resolution, the intelligent window and door management system, based on the preset rule priority order, selects the decision corresponding to rule B, which has the highest priority, from the two conflicting decisions of opening and closing the window—that is, closing the window.

[0112] In some situations, multiple rules with the same priority may conflict. For example, suppose there is another rule, D: close the window when rain is detected outside, and rule D has the same priority as rule B. If it happens to be raining outside and the air quality is poor, then the conditions of both rule B and rule D are met, but they both point to the same decision to close the window, so there is no conflict. However, if there is another rule, E, with the same priority as rules A and C: close the window when the air conditioner is on for energy saving, then on a summer afternoon when the air conditioner is on, rules A, C, and E may conflict due to their equal priority. In this case, the intelligent window and door management system will adopt a second conflict resolution strategy: select a preset default window and door control decision. This default decision is usually the most conservative option set based on safety or energy saving considerations, such as keeping the window closed or setting the window opening degree to 10%.

[0113] Through the aforementioned identification, assessment, and conflict resolution processes, no matter how complex the initial situation, the intelligent door and window management system can ultimately arrive at a unique and clear door and window control decision. This decision will then be transmitted to subsequent execution stages.

[0114] Furthermore, step S4 includes:

[0115] S41: When doors and windows move automatically, detect the user's manual operation on the doors and windows;

[0116] S42: If manual operation is detected, obtain the current position of the door and window when the manual operation starts, the direction of the manual operation, and the final position of the door and window when the manual operation ends.

[0117] S43: Obtain the target position for automatic movement of doors and windows, and compare the final position with the target position;

[0118] S44: If the direction of manual operation is opposite to the direction of automatic movement, or the deviation between the final position and the target position exceeds a preset threshold, a first type of user preference correction instruction is generated.

[0119] S45: If the deviation between the final position and the target position does not exceed the preset threshold, then generate a second type of user preference correction instruction;

[0120] S46: Associate the first type of user preference correction instruction or the second type of user preference correction instruction with the scene context, and execute the first type of user preference correction instruction or the second type of user preference correction instruction first.

[0121] To enable the intelligent door and window management system to accurately learn and adapt to users' personalized preferences, sensors installed on the drive motor or door / window structure continuously monitor for external force when the system is executing an automated command, such as opening curtains from fully closed to a 50% target position. Once a manual pushing or pulling operation is detected, the intelligent door and window management system immediately records the event.

[0122] Then, the intelligent door and window management system collects key parameters of this manual operation: the instantaneous position of the curtains at the start of the operation, such as 20%; the direction of the user's operation, whether to continue opening or close in the opposite direction; and the final position of the curtains when the user stops the operation, such as when the user has pulled them to 80%. This data is a quantitative description of user behavior and is the foundation for understanding the user's true intentions.

[0123] Subsequently, the intelligent door and window management system compares the user's operation result with its original goal. The system retrieves the original automatic movement target position (50%) and compares it with the final position reached by the user (80%). Based on the comparison result, the intelligent door and window management system categorizes user preference adjustments into two types.

[0124] The first scenario is when the user's manual operation is completely opposite to the system's automatic movement direction; for example, the system is opening a window, but the user forcibly closes it. Alternatively, the user's final position deviates significantly from the system's target position, exceeding a preset threshold, such as a deviation of over 30%. In this case, the intelligent door and window management system will determine that the user's behavior strongly negates or significantly corrects the system's decision, and therefore will generate a Type I user preference correction instruction. This instruction represents a user preference that requires serious attention.

[0125] The second scenario involves the user manually operating in the same direction as the intelligent window and door management system's automatic movement, with the final position deviating from the target position within a preset threshold. For example, the intelligent window and door management system might aim to open the door 50%, but the user stops at 60%. This indicates that the user generally accepts the system's decision direction but desires some fine-tuning. In this case, the intelligent window and door management system generates a second type of user preference correction instruction. This instruction represents the user's optimization and personalized adjustment of the system's decision. By distinguishing between these two types of correction instructions, the intelligent window and door management system can more precisely understand the strength and nature of user preferences, thereby adopting different strategies in subsequent learning and application.

[0126] Finally, regardless of whether it's the first or second type of user preference correction instruction, the intelligent door and window management system will firmly associate it with the context in which the instruction was generated and store it. For example, the intelligent door and window management system will record that when users are focused on work, they tend to open the curtains to 80%. Furthermore, these correction instructions generated by direct user behavior will be given priority in future decision-making processes, with a higher priority than the system's preset general behavior rules.

[0127] Furthermore, step S46 includes:

[0128] S461: Store the first type of user preference correction instruction or the second type of user preference correction instruction and the scene context as key-value pairs in the rule database;

[0129] S462: Set the priority flag for the correction instructions in the stored key-value pairs;

[0130] S464: When making decisions on door and window control, retrieve correction instructions with priority identifiers from the rule database based on the current scene context;

[0131] S464: Apply the retrieved correction instruction with priority identifier to execute either the first type of user preference correction instruction or the second type of user preference correction instruction.

[0132] When a user preference correction instruction is generated, the intelligent door and window management system packages it into a key-value pair along with the current scene context. In this key-value pair, the key is a unique identifier for the scene context, such as: "Focus on work scene_Study_Weekday_Afternoon," and the value is the specific correction instruction, such as: "Set curtain position to 80%." This key-value pair is then stored in a dedicated user preference area of ​​the general information platform's rule database. This storage method ensures that every effective user intervention is transformed into an empirical rule that can be recalled in the future and is associated with a specific context.

[0133] To reflect the priority of user preferences during decision-making, the intelligent door and window management system assigns a special priority identifier to each correction instruction while storing the key-value pair. This identifier distinguishes it from ordinary, preset behavioral rules. For example, all user-generated correction instructions can be assigned the highest priority level or a special user-defined label. First-type user preference correction instructions, representing strong user intentions, can be assigned a higher priority identifier than second-type instructions, thus gaining greater weight in the learning model.

[0134] In the subsequent door and window control decision-making process, when the intelligent door and window management system identifies a new scene context, before querying the general preset behavior rule library, it will first use the current scene context as the search key to query whether there are matching items with priority indicators in the user preference database.

[0135] If a matching correction instruction with a priority identifier is found, the intelligent window and door management system will directly apply this instruction to generate window and door control decisions. This means that the user's personal preferences will override or replace general preset rules. For example, even if the preset rules suggest opening the curtains to 50% in a focused work scenario, the intelligent window and door management system will ultimately decide to open the curtains to 80% because it has retrieved a correction instruction generated by the user in that scenario. Through this complete storage, identification, retrieval, and application mechanism, the intelligent window and door management system achieves the ability to dynamically learn from user behavior and continuously optimize its own behavior, thus making the automated control of intelligent windows and doors increasingly aligned with the user's personal habits.

[0136] Furthermore, step S5 includes:

[0137] S51: Before executing door and window control decisions or user preference correction instructions, obtain information on the urgency level of the current scenario and the user's historical response pattern information;

[0138] S52: Based on the current scenario context, determine a set of interaction channels for making suggestions or inquiries to the user. The set of interaction channels includes at least one of smart display screen, voice broadcast, and mobile application push.

[0139] S53: Calculate the duration of the preset response time window based on the urgency information of the current scenario and the user's historical response pattern information.

[0140] To further refine the interaction process, the smart window and door management system assesses the urgency of the current scenario before issuing a query. For example, if the system is connected to a weather warning service and receives an alert for an impending rainstorm, then the decision to close the windows is highly urgent. Conversely, a routine curtain adjustment to improve indoor lighting is less urgent. Simultaneously, the smart window and door management system analyzes users' historical response patterns, derived through long-term observation and recording of user interactions. For instance, the system might detect that one user prefers to interact via voice assistant, while another uses mobile applications more frequently.

[0141] Based on this information, the intelligent door and window management system intelligently selects the most appropriate interaction channel. That is, it determines one or more interaction channels most likely to effectively reach the user based on the current scene context. For example, if it detects that the user is watching a smart TV in the living room, the intelligent door and window management system might choose to pop up a non-intrusive notification window in the corner of the TV screen to ask a question, rather than interrupting the user's viewing experience with voice. If it detects that the user has left home but needs to confirm the status of the doors and windows, then mobile app push notifications become the only option. If the user is in the room and not engaged in any specific activity, then voice broadcasting may be the most direct and effective method. This multi-channel, scene-based intelligent selection ensures the effectiveness of information delivery and the comfort of the user experience.

[0142] Furthermore, the duration of the preset response time window is no longer a fixed value, but is dynamically calculated. The intelligent door and window management system comprehensively considers the urgency of the current scenario and the user's historical response patterns. For highly urgent decisions, such as closing windows during a rainstorm, the preset response time window is set very short, perhaps only ten seconds, to ensure that safety measures can be implemented quickly. For non-urgent, advisory operations, the preset response time window can be set longer, such as one minute, giving the user ample time to consider. Simultaneously, if historical data shows that a user typically responds within five seconds of receiving a notification, the intelligent door and window management system can appropriately shorten the preset response time window for that user to improve overall operational efficiency. This adaptive preset response time window setting makes the interaction rhythm of the intelligent door and window management system more flexible and user-friendly.

[0143] Specifically, the preset response time window can be calculated using the urgency level from the urgency information and the average response time from the user's historical response pattern information as input parameters, with the output being the calculated preset response time window duration. The specific input-output model can be a linear regression model, which can be pre-trained based on the user's historical response pattern information.

[0144] Furthermore, user feedback is monitored within a preset response time window. If user feedback is received, the door and window control decisions are adjusted in real time based on the feedback, including:

[0145] S61: Detects the force, direction, and duration of manual pushing and pulling by the user;

[0146] S62: Determine the type and degree of correction of user feedback based on force, direction and duration;

[0147] S63: Convert the determined correction type and correction degree into incremental or absolute position adjustment instructions for doors and windows, and execute the adjustment instructions.

[0148] When a user chooses to interact directly with the door or window by hand within the response time window, the torque sensor and position encoder built into the motor controller or door / window frame will start working. The intelligent door and window management system will detect the magnitude of the pushing or pulling force applied by the user to the door or window in real time, determine whether the operation is to open or close, and record the duration of this operation.

[0149] After acquiring these physical parameters, the intelligent door and window management system makes intelligent judgments. For example, a brief but forceful push or pull might be interpreted as the user wanting the door or window to move quickly to its limit, i.e., fully open or fully closed. A gentle but continuous force might be interpreted as the user wanting to make fine adjustments to the position; the speed and distance of the door or window's movement will be proportional to the duration and magnitude of the force applied by the user, and the door or window will stop when the user stops applying force. The intelligent door and window management system can even learn different users' operating habits; for example, it can identify that a user's habitual light tap operation represents increasing the opening degree of the curtains by 10%. In this way, the intelligent door and window management system can determine the type of feedback from the user's physical actions—whether it's a full open or closed position, a proportional adjustment, an incremental adjustment, and the specific degree of correction.

[0150] Finally, the intelligent door and window management system converts the determined correction type and degree into precise control commands for the door and window drive motors in real time. For example, if the system determines that the user intends to make a fine adjustment, it generates an incremental position adjustment command, such as increasing the current position by 15 centimeters. If the system determines that the user intends to close quickly, it generates an absolute position adjustment command, such as moving to the zero position. These commands are immediately sent to the motor controller and executed, thus achieving a highly intuitive and immediate response to the user's physical feedback.

[0151] Furthermore, in step S6, if there is no user feedback, the door and window actuator is driven to complete the corresponding action according to the target door and window state determined by the door and window control decision or user preference correction instruction, or the current door and window state is kept unchanged, including:

[0152] S64: Obtain the target door / window status determined by the door / window control decision or user preference correction instruction;

[0153] S65: Determine the execution priority of door and window control decisions or user preference correction instructions based on the type of such instructions;

[0154] S66: Adjust the operating speed of the door and window actuator according to the execution priority, and drive the door and window to reach the target door and window state or keep the current door and window state unchanged.

[0155] When the preset response time window expires and the smart door and window management system has not received any user feedback, it initiates the default execution process. The first step is to define the action goal. The smart door and window management system extracts the specific target door and window status from the door and window control decisions to be executed or user preference correction instructions, such as opening the living room curtains to 75% or completely closing the bedroom window.

[0156] Before activating the doors and windows, the intelligent door and window management system performs a priority evaluation of the instruction to be executed. This evaluation is based on the type of instruction and its associated scenario context. For example, a safety-related instruction, such as automatically opening windows to ventilate smoke upon detecting a high concentration of smoke, will have the highest priority. An instruction related to user health, such as closing windows when air quality is poor, will have a lower priority. And an instruction purely for improving comfort, such as fine-tuning curtains based on changes in light, will have a relatively lower priority.

[0157] This execution priority directly affects the way doors and windows operate. The intelligent door and window management system dynamically adjusts the operating speed of the door and window actuators, i.e., the motors, based on the determined execution priority. For the highest priority safety commands, the doors and windows will be driven at the fastest speed to ensure the target state is reached in the shortest possible time. For medium priority commands, a standard speed may be used. For low priority comfort adjustment commands, a very slow and smooth speed will be used to drive the doors and windows. This slow-speed design avoids disturbing users with sudden, rapid movement of doors and windows, especially in scenarios where users may be resting or require a quiet environment. Furthermore, the slow movement makes the entire automation process more elegant and quiet, enhancing the user's subjective experience. If the pending door and window control decisions or user preference correction commands explicitly state that the current door and window state should remain unchanged, no adjustments will be made. Through this mechanism of adjusting operating speed based on priority, the intelligent door and window management system can complete its automation tasks in a more intelligent and human-centered way, even without human intervention.

[0158] In step S66, the operating speed is adjusted by regulating the PWM (Pulse Width Modulation) duty cycle of the motor driver. For low-priority comfort adjustment commands (such as fine-tuning according to lighting conditions), the controller outputs a PWM signal with a 20% duty cycle, keeping the motor operating in a quiet, slow-speed range. For high-priority emergency commands (such as rain protection), the controller outputs a PWM signal with a 100% duty cycle. Simultaneously, the real-time speed is calculated using position pulses fed back from the Hall sensor, and the stability of the target speed is maintained through a PID (Proportional-Integral-Derivative) closed-loop control algorithm.

[0159] In the above method, in order to overcome speed fluctuations caused by uneven track friction, wind pressure changes, or aging resistance during the operation of doors and windows, the system introduces a PID (Proportional-Integral-Derivative) control algorithm to dynamically adjust the motor drive signal. Specifically, the system first calculates the speed deviation. , i.e., target speed With real-time speed Difference: Subsequently, the PID controller calculates the control output based on the deviation. Its discretization formula is as follows:

[0160] Among them, in the proportional term middle, This is a proportionality coefficient used to quickly respond to current speed deviations. When doors or windows encounter obstructions, causing speed... When the motor descends, the proportional term rapidly increases the output, thereby increasing the motor torque.

[0161] In the integral term middle, Integral coefficients are used to eliminate steady-state errors. Indicates the time from system startup ( ) to the current time ( ), velocity deviation of all sampling points The sum of . The sampling period is defined as follows. As long as there is a speed deviation, the integral term will continue to accumulate until the real-time speed fully reaches the target speed, ensuring that the doors and windows can ultimately operate at a precise speed.

[0162] In differential terms middle, These are the differential coefficients, used to predict the trend of deviation changes. This represents the difference between the error value at the current time and the error value at the previous sampling time. This is an approximation (slope) of the derivative of the error with respect to time. When the speed changes rapidly, the differential term produces a counter-regulating effect, thereby suppressing overshoot and oscillations, making the operation of doors and windows smoother.

[0163] Calculated control quantity The duty cycle is mapped to a PWM (Pulse Width Modulation) signal. The microcontroller determines this based on... The value dynamically adjusts the PWM waveform output to the motor drive bridge (such as an H-bridge circuit). When When the PID output increases, the PWM duty cycle increases, the average motor voltage increases, and the speed increases; when When the PID output decreases, the PWM duty cycle decreases, and the motor decelerates. When the two are equal, the motor speed remains constant.

[0164] Please refer to Figure 2 A control device for an intelligent door and window management system, used to implement any of the above methods, the device comprising:

[0165] Acquisition module 201: Acquires ambient light information, user schedule information, indoor sound characteristics information, and other smart device status information;

[0166] Recognition Module 202: Comprehensively analyzes ambient light information, user schedule information, indoor sound feature information, and other smart device status information to identify the user's scene context;

[0167] Acquisition module 203: Based on the scene context, acquire door and window control decisions from the stored scene-related preset behavior rules;

[0168] Generation module 204: Monitors user manual operations on doors and windows. If manual operations are found, it generates user preference correction instructions associated with the scene context and executes the user preference correction instructions first.

[0169] Inquiry module 205: Before executing door and window control decisions or user preference correction instructions, it sends suggestions or inquiries to the user and sets a preset response time window;

[0170] Execution module 206: Listens for user feedback within a preset response time window. If user feedback is present, the door and window control decision is corrected in real time based on the feedback. If no user feedback is present, the door and window actuator is driven to complete the corresponding action according to the target door and window state determined by the door and window control decision or the user preference correction instruction, or the current door and window state is kept unchanged.

[0171] The various modules of this device work together to form a complete intelligent control hub. The data acquisition module 201 provides continuous, multi-dimensional data input to the intelligent door and window management system by connecting to various sensors and data interfaces.

[0172] Specifically, the acquisition module 201 includes a microcontroller (e.g., STM32 series) and peripheral interfaces. The method for acquiring indoor sound characteristic information is as follows: ambient audio is acquired through a microphone array at a sampling rate of 44.1kHz. The system divides the 20Hz-20kHz audio into three frequency bands: low frequency (<300Hz), mid frequency (300Hz-3400Hz), and high frequency (>3400Hz). If the energy proportion of the mid frequency band exceeds 60% and there are intermittent pauses, it is identified as human conversation; if the energy distribution across the entire frequency band is uniform and the duration exceeds 5 minutes, it is identified as background noise or entertainment sound effects.

[0173] The identification module 202 processes and understands the raw data sent by the acquisition module, extracting profound insights into user intent and environmental conditions—the scene context—through sophisticated analysis algorithms. The acquisition module 203, based on the identification module's judgment, quickly finds the most suitable action plan from a vast rule knowledge base. The generation module 204 keenly captures every manual intervention by the user and transforms it into valuable personalized experience, continuously enriching and optimizing the knowledge base. The inquiry module 205 intelligently communicates with the user before taking action, ensuring that every action of the intelligent door and window management system meets the user's expectations. Finally, the execution module 206 receives the final instructions and precisely and smoothly controls the physical movements of the doors and windows, while also sensitively sensing and responding to the user's immediate physical feedback. The organic combination of these modules enables the device to effectively execute the aforementioned intelligent control methods, thereby providing users with a truly automated, personalized, and superior intelligent door and window management solution.

[0174] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A control method for an intelligent door and window management system, characterized in that, The method includes the following steps: S1: Collects ambient light information, user schedule information, indoor sound characteristics information, and other smart device status information; S2: By comprehensively analyzing the ambient light information, user schedule information, indoor sound feature information, and other smart device status information, the user's scene context is identified; S3: Based on the scene context, obtain door and window control decisions from the stored scene-related preset behavior rules; S4: Monitor user manual operations on doors and windows. If manual operations are detected, generate user preference correction instructions associated with the scene context and execute the user preference correction instructions first. S5: Before executing the door and window control decision or the user preference correction instruction, issue a suggestion or inquiry to the user and set a preset response time window; S6: Listen for user feedback within the preset response time window. If user feedback is present, the door and window control decision is corrected immediately based on the feedback. If no user feedback is present, the door and window actuator is driven to complete the corresponding action according to the target door and window state determined by the door and window control decision or the user preference correction instruction, or the current door and window state is kept unchanged.

2. The control method for an intelligent door and window management system according to claim 1, characterized in that, Step S2 includes: S21: Set the user schedule information and the ambient light information as the first priority information, and set the indoor sound feature information and other smart device status information as the second priority information; other smart devices include at least a smart thermostat, an air purifier, a smart TV, and a smart speaker; S22: Based on the first priority information, perform a first-level scene classification, which includes at least sleep-related scenes, work-related scenes, and daily activity scenes. S23: Combine the second priority information and the results of the first-level scene classification to perform a second-level scene judgment and obtain the scene context.

3. The control method for an intelligent door and window management system according to claim 2, characterized in that, Step S23 includes: S231: Under the first-level scene classification result of the sleep-related scene, if the indoor sound feature information is identified as quiet and the status information of other smart devices shows that the device is in the off state, then the scene context is finally identified as a quiet sleep scene. S232: Under the first-level scene classification result of the work-related scene, if the indoor sound feature information is identified as a slight human voice but no obvious entertainment noise, and the status information of other smart devices shows that the device is in the off state, then the scene context is finally identified as a focused work scene. S233: Under the first-level scene classification result of the daily activity scenario, if the indoor sound feature information is identified as a sound pattern with high loudness and wide frequency distribution, and the status information of other smart devices shows that the smart TV or smart speaker is on, then the scene context is finally identified as a home entertainment scenario.

4. The control method for an intelligent door and window management system according to claim 1, characterized in that, Step S3 includes: S31: Identify all preset behavior rules associated with the scene context and doors and windows; S32: Evaluate the conditions of the preset behavior rules and determine the set of preset behavior rules that meet the conditions; S33: If there are multiple preset behavior rules in the set of preset behavior rules that meet the conditions, and their corresponding door and window control decisions conflict with each other, then conflict resolution is performed. The conflict resolution includes: Based on the preset rule priority order, select the door and window control decision corresponding to the rule with the higher priority from the conflicting door and window control decisions; If the conditions of multiple conflicting rules with the same priority are all met, the preset default door and window control decision is selected. The door and window control decision determined after conflict resolution or in the absence of conflict is taken as the door and window control decision.

5. The control method for an intelligent door and window management system according to claim 1, characterized in that, Step S4 includes: S41: When the doors and windows move automatically, detect the user's manual operation on the doors and windows; S42: If a manual operation is detected, obtain the current position of the door / window when the manual operation begins, the direction of the manual operation, and the final position of the door / window when the manual operation ends. S43: Obtain the target position for the automatic movement of the doors and windows, and compare the final position with the target position; S44: If the direction of the manual operation is opposite to the direction of the automatic movement, or if the deviation between the final position and the target position exceeds a preset threshold, a first type of user preference correction instruction is generated. S45: If the deviation between the final position and the target position does not exceed a preset threshold, then generate a second type of user preference correction instruction; S46: Associate the first type of user preference correction instruction or the second type of user preference correction instruction with the scenario context, and execute the first type of user preference correction instruction or the second type of user preference correction instruction first.

6. The control method for an intelligent door and window management system according to claim 5, characterized in that, Step S46 includes: S461: Store the first type of user preference correction instruction or the second type of user preference correction instruction and the scene context as key-value pairs in the rule database; S462: Set the priority flag for the correction instructions in the stored key-value pairs; S464: When making decisions on door and window control, retrieve correction instructions with priority identifiers from the rule database based on the current scene context; S464: Apply the retrieved correction instruction with priority identifier to execute either the first type of user preference correction instruction or the second type of user preference correction instruction.

7. The control method for an intelligent door and window management system according to claim 1, characterized in that, Step S5 includes: S51: Before executing the door and window control decision or the user preference correction instruction, obtain the urgency information of the current scenario and the user's historical response pattern information; S52: Based on the current scene context, determine a set of interaction channels for issuing suggestions or inquiries to the user, wherein the set of interaction channels includes at least one of smart display screen, voice broadcast and mobile application push; S53: Calculate the duration of the preset response time window based on the urgency information of the current scenario and the user's historical response pattern information.

8. The control method for an intelligent door and window management system according to claim 1, characterized in that, In step S6, user feedback is monitored within the preset response time window. If user feedback is received, the door and window control decision is immediately corrected based on the feedback, including: S61: Detects the force, direction, and duration of manual pushing and pulling by the user; S62: Determine the type and degree of correction of the user feedback based on the force, direction and duration; S63: Convert the determined correction type and correction degree into incremental or absolute position adjustment instructions for doors and windows, and execute the adjustment instructions.

9. The control method for an intelligent door and window management system according to claim 1, characterized in that, In step S6, if there is no user feedback, the door and window actuator is driven to complete the corresponding action according to the target door and window state determined by the door and window control decision or the user preference correction instruction, or the current door and window state is kept unchanged, including: S64: Obtain the target door / window status determined by the door / window control decision or the user preference correction instruction; S65: Determine the execution priority of the door and window control decision or the user preference correction instruction based on the type of the door and window control decision or the user preference correction instruction; S66: Based on the execution priority, adjust the operating speed of the door and window actuator, and drive the door and window to reach the target door and window state or keep the current door and window state unchanged.

10. A control device for an intelligent door and window management system, characterized in that, The device includes: Data acquisition module: Collects ambient light information, user schedule information, indoor sound characteristics information, and other smart device status information; Recognition module: By comprehensively analyzing the ambient light information, user schedule information, indoor sound feature information, and other smart device status information, the recognition module identifies the user's scene context. Acquisition module: Based on the scene context, acquire door and window control decisions from the stored scene-related preset behavior rules; Generation module: Monitors user manual operations on doors and windows. If manual operations are detected, it generates user preference correction instructions associated with the scene context and executes the user preference correction instructions first. Inquiry module: Before executing the door and window control decision or the user preference correction instruction, it issues suggestions or inquiries to the user and sets a preset response time window; Execution module: Listens for user feedback within the preset response time window. If user feedback is present, the door and window control decision is corrected in real time based on the feedback. If no user feedback is present, the door and window actuator is driven to complete the corresponding action according to the target door and window state determined by the door and window control decision or the user preference correction instruction, or the current door and window state is kept unchanged.

Images