Intelligent environment purification and humidification integrated linkage regulation and control method and system

Abstract

The invention relates to an intelligent environment purification and humidification integrated linkage regulation and control method and system, and solves the problem that a regulation and control strategy lacks pertinence and accuracy. The method comprises the steps that a preset similarity calculation algorithm is utilized to calculate the overall matching degree with rules in an information matching rule base one by one, and the overall matching degree is calculated; selecting a rule associated with the highest overall matching degree as a successfully matched rule, and integrating the matched personnel information, scene information and corresponding rule content to form a personalized information set; and inputting the personalized information set into a pre-constructed relevant equipment regulation and control strategy analysis model, outputting a relevant equipment regulation and control strategy, and executing the relevant equipment regulation and control strategy. The method has the following effects that purification and humidification equipment is accurately adjusted according to different scenes and personnel requirements, and a comfortable and healthy indoor environment is created for a user.

Description

Technical Field

[0001] The present invention relates to the field of smart homes, and in particular to an intelligent environmental purification and humidification integrated linkage control method and system. Background Art

[0002] In modern life, people have increasing requirements for the comfort and health of indoor environment. The quality of indoor environment not only affects people's quality of life, but is also closely related to people's physical health. As an important means to improve the indoor environment, environmental purification and humidification, and its integrated linkage control technology have gradually become a hot spot for research and application.

[0003] At present, some related technologies have been applied in the field of environmental control. For example, some smart devices can collect basic environmental parameters such as indoor temperature, humidity, and air quality through sensors, and control the operation of purification and humidification equipment according to preset thresholds. There are also some technologies that can automatically adjust the working status of the equipment in different time periods according to time settings.

[0004] However, these existing technologies have many drawbacks. They are relatively simple in acquiring and analyzing personnel and scene information, and can only be controlled based on simple environmental parameters or time settings. They cannot fully perceive the specific status, behavioral habits, and complex scene changes of personnel, resulting in a lack of pertinence and precision in the control strategy. Summary of the invention

[0005] In order to accurately adjust the purification and humidification equipment according to different scenarios and personnel needs, and create a comfortable and healthy indoor environment for users, the present application provides an intelligent environment purification and humidification integrated linkage control method and system.

[0006] In the first aspect, the present application provides an intelligent environmental purification and humidification integrated linkage control method, which adopts the following technical solutions: A method for integrated linkage control of intelligent environmental purification and humidification, comprising: Obtain information about people in the room and the scene they are in; Filter out key data from personnel information and scene information; Compare the filtered personnel information and scene information with the rules in the information matching rule library one by one; If the comparison is completely consistent, the matching personnel information, scene information and corresponding rule content are integrated to form a personalized information set; If there is no complete match, the preset similarity calculation algorithm is used to calculate the overall matching degree of the rules in the information matching rule library one by one, and the rule associated with the highest overall matching degree is selected as the successfully matched rule, and the matched personnel information, scene information and corresponding rule content are integrated to form a personalized information set; The personalized information set is input into a pre-built relevant device control strategy analysis model, the relevant device control strategy is output, and the relevant device control strategy is executed.

[0007] By adopting the above technical solution, the method can accurately collect indoor personnel and scene information, form a personalized information set through screening and comparison, and then output the control strategy through the equipment control strategy analysis model. This process realizes the intelligent and personalized environmental control, and can accurately adjust the purification and humidification equipment according to different scenes and personnel needs, creating a comfortable and healthy indoor environment for users, and can also improve equipment operation efficiency and reduce energy consumption.

[0008] Optionally, steps are also included in parallel with filtering out key data from personnel information and scene information, as follows: Analyze whether key data related to scene recognition and conversion is obtained, including the activity posture of people, the type and intensity of sound, and precise location; If no, keep the original setting. If yes, then determine whether the person moves between rooms based on the key data obtained related to scene recognition and conversion; If yes, it is determined that there is a scene transition; If not, keep the original setting.

[0009] By adopting the above technical solutions, this parallel step can accurately judge the scene transition by capturing key data such as the posture, sound, and position of personnel activities. It can enable the intelligent environmental control system to keenly perceive environmental changes, and quickly respond and adjust the purification and humidification strategies in time when personnel move and switch scenes, ensuring that the environment is always comfortable and pleasant in each scene, optimizing the user experience, and making environmental control more intelligent and more in line with actual scene needs.

[0010] Optionally, forming a personalized information set includes: Analyze whether there is a scene transition; If not, maintain the original setting; If yes, obtain the multimodal information involved in this scene conversion, and match the rules corresponding to the multimodal information of this scene conversion from the information matching rule library, integrate the multimodal information of this scene conversion and the corresponding rules and matching personnel information, scene information and corresponding rule content to form a personalized information set.

[0011] By adopting the above technical solutions, this process analyzes scene transitions, collects multimodal information and matches corresponding rules when a transition is confirmed, and then integrates them into a personalized information set. This enables the system to keenly capture scene changes, integrate rich information with rules, provide accurate basis for subsequent equipment control strategies, realize dynamic optimization of environmental control, meet users' diverse needs in different scenarios, and improve indoor environmental comfort.

[0012] Optionally, the personalized information set is input into a pre-built relevant device control strategy analysis model, and the output relevant device control strategy includes: Analyze whether the personalized information set contains multimodal information and the rules corresponding to the multimodal information; If not, maintain the original setting; If yes, then the rules corresponding to the multimodal information, the rules corresponding to the personnel information, and the rules corresponding to the scene information are integrated; According to the importance of the rules, the scope of applicable scenarios and the relevance to the current scenario, a preset weighted average algorithm is used to calculate the comprehensive score, and the integrated rules are prioritized according to the results of the comprehensive score ranking; The rule inference engine is used to generate a preliminary equipment control strategy based on the priority rules.

[0013] By adopting the above technical solution, this step is based on personalized information collection, analyzes multimodal information and rules, integrates various rules, calculates and sorts comprehensive scores according to importance, applicable scenarios, etc., and generates strategies with the help of rule reasoning engine. This ensures that the control strategy fits the actual scenario, takes into account personnel needs and rule characteristics, realizes precise and intelligent control, optimizes equipment operation, and creates a comfortable, healthy and energy-saving indoor environment for users.

[0014] Optionally, a rule inference engine is used to generate a preliminary device control strategy based on the prioritized rules, including: Continuously collect real-time data of the current scene through various sensors and perform data preprocessing; Compare the preprocessed data with the prioritized rule base, and quickly locate all rules that meet the data characteristics of the current scenario in order of priority; According to the matched rules, the inference engine uses the forward reasoning mechanism, starting from the current scene data, and gradually derives the preliminary equipment control strategy based on the rules. For each triggered rule, it determines the corresponding equipment operation instructions.

[0015] By adopting the above technical solutions, sensors are used to collect and preprocess data in real time, and the positioning and adaptation rules are compared with the rule library, and then the forward reasoning mechanism is used to derive strategies and determine device instructions. This makes the control strategy closely fit the real-time scene, ensuring that the equipment operates according to accurate rules, achieving precise and efficient environmental control, and being able to quickly respond to environmental changes, providing users with a stable and comfortable indoor environment experience.

[0016] Optionally, a rule inference engine is used to generate a preliminary device control strategy based on the prioritized rules, including: Set environmental control goals based on current scene requirements and user preferences; Starting from the set goals, the inference engine reversely searches the rule base for rules that can achieve these goals; Based on the rule chain obtained by reverse lookup, the inference engine backtracks to generate specific equipment control strategies to achieve the goal.

[0017] By adopting the above technical solutions, this method sets goals based on scene requirements and user preferences, reversely searches for rules and generates strategies retroactively. It can closely control the environment around user needs, make the equipment operation fit the actual scene, and avoid invalid operations. At the same time, through reverse deduction of rules, the strategy formulation process is optimized, the control accuracy is improved, and a comfortable and personalized indoor environment is created for users, enhancing the user experience.

[0018] Optionally, the method further includes the following steps after using the rule inference engine to generate a preliminary device control strategy according to the rules sorted by priority: Obtain user feedback data and use natural language processing technology to convert user text feedback into quantitative rule adjustment requirements; Classify the quantified rule adjustment requirements into environment parameter adjustment requirements, device operation requirements, and scene mode adjustment requirements; Prioritize each requirement based on its urgency and impact on user experience using a pre-built evaluation matrix; According to the analysis results of the rule adjustment requirements, the rules in the rule base are modified accordingly.

[0019] By adopting the above technical solutions, the process is guided by user feedback, uses natural language processing to quantify needs, and after classification, prioritizes them according to urgency and user experience impact, and then modifies the rule base accordingly. This enables the system to continuously learn and dynamically optimize, accurately respond to various user demands, continuously improve the fit and comfort of environmental control, make equipment operation more in line with user wishes, and significantly enhance user satisfaction with the intelligent environmental control system.

[0020] In the second aspect, the present application provides an intelligent environmental purification and humidification integrated linkage control system, which adopts the following technical solutions: A smart environment purification and humidification integrated linkage control system includes a memory, a processor, and a program stored in the memory and executable on the processor. When the program is loaded and executed by the processor, the smart environment purification and humidification integrated linkage control method as described in the first aspect is implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a flow chart of an integrated linkage control method for intelligent environmental purification and humidification in an embodiment of the present application.

[0022] Figure 2 It is a flowchart of another embodiment of the present application and the steps of screening key data from personnel information and scene information in parallel.

[0023] Figure 3 It is a flowchart of forming a personalized information set according to another embodiment of the present application.

[0024] Figure 4 Another embodiment of the present application is to input a personalized information set into a pre-built related equipment control strategy analysis model, and output a flow chart of the related equipment control strategy. DETAILED DESCRIPTION

[0025] The present application is further described in detail below in conjunction with the accompanying drawings.

[0026] Reference Figure 1 , is a method for integrated linkage control of intelligent environmental purification and humidification disclosed in this application, comprising: Step S100, obtaining information about people in the room and information about the scene in which they are located.

[0027] Personnel information: including identity, number, health status, and behavioral habits, which is used to determine personnel's needs for the environment. For example, the requirements of the elderly and children for temperature and humidity are different from those of young people. Personnel information is obtained in the following ways: using smart door locks and face recognition devices to identify identities, cameras to count the number of people; using smart bracelets to collect health data; and analyzing smart device operation records to understand behavioral habits.

[0028] Scene information: refers to the indoor state, such as time, spatial location (living room, bedroom, etc.), activity type (party, sleep, etc.). Different scenes have different environmental requirements. For example, the sleeping scene requires a quiet and dim environment. The scene information is obtained as follows: directly obtain the system time; use Bluetooth and Wi-Fi positioning to determine the spatial location, and the camera combines algorithms to determine the functional area; the microphone collects sound and combines voice recognition to determine the activity type, and the camera can also analyze the posture to determine.

[0029] Operation process: The system starts, triggering the information collection equipment to work. The identity recognition equipment monitors the entry and exit of personnel and transmits information, the health monitoring equipment regularly uploads health data, the scene information collection equipment collects and analyzes environmental data in real time, and finally the system summarizes and integrates this information.

[0030] For example: At 7 o'clock in the evening, the user enters the bedroom through the smart door lock, the smart watch synchronizes the heart rate data to the system, the camera recognizes that the user is lying on the bed, and the microphone does not detect obvious conversation sounds. The system determines it as a sleeping scene, integrates this information, and completes step S100.

[0031] Step S200, filtering out key data from personnel information and scene information.

[0032] Specifically, from the obtained information on personnel and scenes, select the data that plays a key role in environmental purification and humidification control.

[0033] In terms of personnel information, the focus is on the identities of special groups (such as pregnant women and people with allergies), abnormal health conditions (such as those related to respiratory diseases), and behavioral habits related to environmental regulation (such as adjusting the temperature before bedtime).

[0034] Scene information focuses on special time periods (late night, lunch break), specific rooms (bedroom, study) and specific activities (party, sleep, exercise). These scenes have different requirements for environmental indicators such as indoor temperature and humidity, and air quality.

[0035] The system quickly analyzes all information based on pre-set rules, accurately screens out key data and integrates them to provide a basis for the subsequent formulation of purification and humidification control strategies. For example, when it detects that a user with allergies is preparing to rest in the bedroom at 10 pm, the system extracts key information such as allergies, time, location and sleeping scene.

[0036] Step S300, the screened personnel information and scene information are compared one by one with the rules in the information matching rule library.

[0037] The personnel and scene key data screened out in step S200 are compared one by one with the rules in the information matching rule library. The information matching rule library is pre-built and contains a set of rules for environmental purification and humidification control corresponding to various scenes and personnel.

[0038] Execution method: The system will carefully compare the characteristics of key data with the conditions set by each rule in the rule library according to certain algorithm logic. For example, key information such as the identity, health status, behavior habits of personnel, as well as the time, location, activity type, etc. of the scene are all used as comparison factors.

[0039] Operation process: The system starts from the first rule in the rule base and substitutes the key data into the rule conditions for judgment. It judges whether the key data fully meets the scene and personnel conditions set by the rule. If all the conditions of a rule match the key data, the comparison is considered successful; if no completely matching rule is found after traversing the entire rule base, the subsequent similarity calculation process is entered.

[0040] Step S400: If the comparison is completely consistent, the matching personnel information, scene information and corresponding rule content are integrated to form a personalized information set.

[0041] First, the system confirms that the comparison result is a complete match. Then, the specific information related to the matching rule is extracted from the database that stores personnel information and scene information. At the same time, the corresponding rule content is obtained from the rule library, including the control targets of environmental parameters (such as temperature, humidity, air quality indicators, etc.) and the operating mode of the equipment (such as purifier wind speed, humidifier gear, etc.). Finally, this information is summarized into a structured personalized information set, waiting to be input into the equipment control strategy analysis model.

[0042] The example is shown as follows: Assume that after the comparison in step S300, it is determined that the filtered key data (the person is an elderly person, the time is afternoon, the location is in the living room, and the activity type is leisure watching TV) completely matches a rule in the rule library. The content of the rule is "When the person is an elderly person, leisure watching TV in the living room in the afternoon, keep the indoor temperature at around 25°C, the humidity at 45%-55%, the air purifier turns on the low wind speed continuous purification mode, and the humidifier is set to medium gear". The system will integrate the identity information of the elderly, the afternoon time, the living room location, the leisure watching TV scene information, and the above rule content to form a personalized information set.

[0043] Step S500: If there is no complete match, the preset similarity calculation algorithm is used to calculate the overall matching degree of the rules in the information matching rule library one by one, and the rule associated with the highest overall matching degree is selected as the successfully matched rule, and the matching personnel information, scene information and corresponding rule content are integrated to form a personalized information set.

[0044] Similarity calculation algorithm: an algorithm that quantitatively compares the similarity between the selected personnel and scene key data and the rules in the information matching rule library. The value calculated by this algorithm, namely the overall matching degree, can intuitively reflect the degree of fit between the data and the rules.

[0045] Overall matching degree: It is a quantitative indicator to measure the matching degree between key data and each rule in the rule base. The higher the value, the higher the similarity between key data and corresponding rules, and the more suitable it is as a basis for regulation.

[0046] The operation process is as follows: Rule traversal: After the system determines that there is no rule that is completely consistent with the key data, it will read each rule in the information matching rule base in turn.

[0047] Data preprocessing: Convert key data and various information in the read rule content, such as personnel identity, time, location, etc., into a unified data format to facilitate subsequent calculations. For example, the personnel identity is coded with numbers, 0 represents children, 1 represents teenagers, etc.

[0048] Feature extraction and weight assignment: Extract key features from key data and rules, such as the health status of personnel, the activity type of the scene, etc. And assign different weights to these features based on their importance to environmental regulation. For example, in a sleeping scene, the weight of quietness will be higher than that of a normal scene.

[0049] Similarity calculation: Use the preset similarity calculation algorithm, such as the cosine similarity algorithm, to calculate the overall matching degree between the key data and each rule. The cosine similarity algorithm determines the similarity between the key data and the rule by calculating the cosine value of the angle between the key data and the rule in the high-dimensional space vector. The closer the value is to 1, the higher the similarity.

[0050] Rule selection and information integration: Select the rule with the highest overall matching degree, integrate the matching personnel information, scene information and corresponding rule content according to a specific data structure to form a personalized information set.

[0051] Assume that the key data selected is: the person is an adult with mild respiratory disease, the time is 9 pm, the location is in the bedroom, and the activity type is sleeping. There is no completely matching rule in the rule base. The system starts to calculate the overall matching degree, encodes the person's identity as 1, the health status as a specific value representing mild respiratory disease, the time is converted into a value for the evening time period, the location is encoded as the value corresponding to the bedroom, and the activity type is encoded as the value corresponding to sleep. The rules in the rule base are also processed in the same way and calculated using the cosine similarity algorithm. Finally, it is found that the rule "the person is an adult, sleeping in the bedroom at night, and those with respiratory diseases need to pay attention to air quality" has the highest matching degree. The content of the rule is "the indoor temperature is maintained at 26℃, the humidity is 40%, and the air purifier is turned on in high-efficiency purification mode". The system integrates the information of the person, the scene and the content of the rule to form a personalized information set.

[0052] Step S600: input the personalized information set into a pre-built relevant device control strategy analysis model, output the relevant device control strategy, and execute the relevant device control strategy.

[0053] Related equipment control strategy analysis model: This is an intelligent model built on a large amount of data training and algorithms. It can analyze the personnel information, scene information and corresponding rules contained in the input personalized information set, thereby generating targeted environmental purification and humidification equipment control strategies. The model integrates machine learning, data analysis and other technologies to achieve accurate decision-making on environmental control.

[0054] Equipment control strategy: refers to the specific operation plan formulated for indoor purification and humidification equipment, including the setting of parameters such as the equipment's on / off status, operation mode (such as the wind speed level of the air purifier, the humidification intensity of the humidifier, etc.), and working hours, to meet the needs of people for the indoor environment in the current scenario.

[0055] The operation process is as follows: 1. Model input: The personalized information set generated in step S500 is accurately input into the pre-built relevant equipment control strategy analysis model according to the data format and interface specifications required by the model. This step ensures that the model can obtain complete and valid data, providing a basis for subsequent analysis and strategy generation. 2. Strategy analysis and generation: After receiving the input data, the model first deeply analyzes the personalized information set and extracts the key features and rules therein. Then, based on the preset algorithms and logic inside the model, combined with historical data and experience knowledge of environmental control, this information is comprehensively analyzed. During the analysis process, the model will consider a variety of factors, such as different requirements of personnel for temperature, humidity, and air quality in different scenarios, performance parameters and energy consumption of equipment, etc. Finally, through a series of complex calculations and reasoning, a detailed set of equipment control strategies is generated, which clearly stipulates the specific operating parameters and working methods of various types of purification and humidification equipment in the current scenario. 3. Strategy execution: The generated equipment control strategy is transmitted to the corresponding equipment control system through the communication interface. After receiving the control strategy, the equipment control system controls the indoor purification and humidification equipment in real time according to the instructions in the strategy. For example, it sends instructions to the air purifier to adjust its wind speed to achieve the appropriate air purification effect; it sends signals to the humidifier to control its humidification intensity and working time to keep the indoor humidity within an appropriate range. During the operation of the equipment, the equipment control system will also monitor the operating status of the equipment and changes in environmental parameters in real time to ensure that the equipment operates stably according to the control strategy.

[0056] Reference Figure 2 The intelligent environment purification and humidification integrated linkage control method also includes steps in parallel with selecting key data from personnel information and scene information, as follows: Step SA00, analyzing whether key data related to scene recognition and conversion is obtained. If not, executing step SB00; if yes, executing step SC00.

[0057] Key data related to scene recognition and conversion: including information such as people's activity posture, sound type and intensity, precise location, etc. These data are crucial for determining whether the indoor scene has changed. For example, if a person moves from the living room to the bedroom, or changes from a quiet state to a party, these scene changes can be reflected through these key data.

[0058] The operation process of step SA00 is as follows: Data monitoring: The system continuously collects information about people's movements, sounds, and locations through various sensors deployed indoors, such as cameras, microphones, and positioning sensors. The camera captures people's movements and analyzes whether they are standing, walking, or lying down; the microphone records the frequency and volume of the sound to determine whether the sound is a conversation, music, or other noise; the positioning sensor uses Bluetooth, Wi-Fi, or other indoor positioning technologies to track the precise location of people in real time.

[0059] Data analysis: The collected data will be transmitted in real time to the system's data processing module, which uses image recognition algorithms to analyze people's activity posture data, uses sound recognition technology to analyze sound data, and uses positioning algorithms to process location data. For example, image recognition algorithms can identify whether people are doing daily activities or special sports; sound recognition technology can distinguish whether the sound in the current environment comes from normal communication or a lively party; positioning algorithms can accurately determine the room where people are located and whether they have moved.

[0060] Judgment and decision: After data analysis, the system determines whether key data related to scene recognition and conversion has been obtained based on preset judgment rules. If these key data have valid collection and analysis results, the system determines "yes" and proceeds to the next step SC00; if the key data is missing or the analysis results are invalid, the system determines "no" and executes step SB00.

[0061] Step SB00, maintain the original settings.

[0062] The original settings here refer to the operating parameters and working modes of the indoor environment purification and humidification equipment set by the system based on the previous scene and personnel information before the current judgment. These parameters include the wind speed and running time of the air purifier, the humidification amount and working frequency of the humidifier, and the set values ​​of indoor temperature and humidity.

[0063] Step SC00, based on the key data related to scene recognition and conversion obtained, determine whether the person moves between different rooms. If yes, execute step SD00; if no, execute step SE00.

[0064] The process of determining whether a person moves between different rooms is as follows: Location data tracking: The system uses indoor positioning technology (such as Bluetooth positioning, Wi-Fi positioning, etc.) to continuously track the precise location of people and record the coordinate information of people at different time points. For example, the location data of a person is recorded every 5 seconds and stored in a temporary database.

[0065] Data comparison and analysis: Compare the current person's location data with the previous and historical location data. If the person's location coordinates are found to have changed significantly, and the change is consistent with the characteristics of moving from one room to another (such as crossing the boundaries of the positioning areas of different rooms), then proceed to the next step of analysis. At the same time, combined with the person's activity posture data, if the posture shows a walking state and the duration is consistent with the time of moving from one room to another, as well as the change in sound type and intensity (such as from a quiet study environment to a noisy living room environment), the rationality of the location change is further confirmed.

[0066] Movement determination decision: Based on the comprehensive analysis of location changes, activity postures, sounds and other data, the system determines whether a person moves between rooms based on preset determination rules. If all data support the effective movement of the person's position, the system determines "yes" and executes step SD00; if the data is contradictory or insufficient, and it is impossible to clearly determine the movement of the person, the system determines "no" and executes step SE00.

[0067] Step SD00, determining whether there is a scene transition.

[0068] When the system executes step SC00, it determines that there is a scene transition after comprehensively judging that the person moves between different rooms based on the person's precise location, activity posture, sound and other multimodal data. This means that the use requirements of the indoor environment may have changed. For example, when a person enters the living room from the bedroom, the requirements for environmental factors such as temperature, humidity, air quality and lighting may be different, and the system needs to adjust the operation strategy of related equipment accordingly.

[0069] Step SE00, maintain the original settings.

[0070] The operation process is as follows: Equipment status query: The system will first retrieve the operating parameters and working mode records of various current environmental purification and humidification equipment. These records include the wind speed gear, operating time setting, purification mode of the air purifier; the humidification amount, humidification frequency, working time of the humidifier; and the temperature and humidity setting values ​​of air conditioners and other equipment.

[0071] Instructions are issued to maintain: Based on the queried device status information, the system sends instructions to the control modules of each device to maintain the current status. After receiving the instructions, the control module keeps the operating parameters of the device unchanged and continues to operate in the original working mode. For example, if the air purifier is currently in smart mode, it automatically adjusts the wind speed between 2-4 gears according to the indoor air quality. After the system instructions are issued, the air purifier still maintains the smart mode and operates according to the original adjustment logic.

[0072] Reference Figure 3 , forming a personalized information set including: Step S510, analyzing whether there is a scene transition. If not, executing step S520; if yes, executing step S530.

[0073] In the process of generating a personalized information set, step S510 plays a key role in connecting the previous and the next. It focuses on judging whether there is a scene conversion based on the analysis results of the personnel and scene-related data in the previous steps. The scene conversion here not only includes the movement of personnel between different rooms, but also covers the changes in the nature of the scene caused by factors such as activity type and time stage, such as the conversion from an office scene during the day to a leisure scene at night. This judgment result will directly determine whether the subsequent system will maintain the existing settings or start a new scene adaptation process.

[0074] Step S520, maintaining the original settings.

[0075] This means that when the scene remains stable, the existing operating parameters and working mode of the equipment will continue to be maintained to ensure the stability and continuity of the indoor environment, while avoiding energy waste and equipment loss due to unnecessary equipment adjustments.

[0076] Step S530, obtain the multimodal information involved in this scene conversion, and match the rules corresponding to the multimodal information of this scene conversion from the information matching rule library, integrate the multimodal information of this scene conversion and the corresponding rules and matching personnel information, scene information and corresponding rule content to form a personalized information set.

[0077] After determining that there is a scene conversion, the system executes step S530. This step aims to comprehensively collect multimodal information related to this scene conversion, which includes the behavior and physiological data of personnel in the new scene, the environmental parameters of the new scene, and related time information. This information is integrated with the corresponding rules obtained from the information matching rule library, as well as the original personnel information and scene information, to form a personalized information set that adapts to the new scene, providing a data basis for the subsequent generation of accurate equipment control strategies.

[0078] The operation process is as follows: Multimodal information collection: Collect multimodal information in new scenarios through various sensors. For example, use cameras to capture people's behaviors in new scenarios, such as sitting, standing and walking while watching TV in the living room; use smart bracelets and other devices to obtain people's physiological data, such as heart rate, blood oxygen saturation, etc.; deploy temperature and humidity sensors and air quality sensors in the environment to collect environmental parameters such as temperature, humidity, and air quality in new scenarios; and record the current time information at the same time.

[0079] Rule matching: Match the collected multimodal information with the rules in the information matching rule library. For example, if a new scene is identified as watching TV in the living room at night, the system will find the corresponding rules in the rule library. The rules may include adjusting the indoor temperature to 25°C, keeping the humidity at 45%-55%, and turning on the low wind speed purification mode of the air purifier.

[0080] Information integration: Integrate the multimodal information of this scene conversion, the matched rules, and the previously existing personnel information (such as identity, health status, etc.) and scene information (such as the original scene location, etc.). Arrange and combine them according to a specific data structure to form a complete and organized personalized information set, which is convenient for subsequent system analysis and processing.

[0081] Reference Figure 4 , input the personalized information set into the pre-built relevant equipment control strategy analysis model, and output the relevant equipment control strategy including: Step S610, analyzing whether the personalized information set contains multimodal information and rules corresponding to the multimodal information. If not, executing step S620; if yes, executing step S630.

[0082] After the personalized information set is input into the pre-built relevant equipment control strategy analysis model, step S610 focuses on analyzing the set to determine whether it contains multimodal information and the rules corresponding to the multimodal information. Multimodal information covers data on personnel behavior, environmental parameters, and other aspects. This information and its corresponding rules are the key basis for generating accurate equipment control strategies. If such key content is included, the system will make a more in-depth strategy based on it; if not, the original settings will be maintained to ensure that the system maintains stable operation when data is insufficient.

[0083] Step S620, maintaining the original settings.

[0084] When the system determines in step S610 that the personalized information set does not contain multimodal information and corresponding rules, step S620 will be executed. Maintaining the original settings means that when the current data is insufficient to support the generation of a new and more accurate device control strategy, the existing environmental purification and humidification equipment operating parameters and working modes will continue to be maintained. This can ensure the stability of the indoor environment, avoid discomfort that may be caused by blindly adjusting the equipment, and also prevent energy waste and loss caused by frequent start and stop of the equipment.

[0085] Step S630, integrating the rules corresponding to the multimodal information, the rules corresponding to the personnel information, and the rules corresponding to the scene information.

[0086] The operation process is as follows: Rule extraction and classification: The system accurately extracts the corresponding rules for multimodal information, personnel information, and scene information from the personalized information set. For example, if the multimodal information detects that the indoor carbon dioxide concentration is too high and the personnel activities are frequent, the corresponding rule may be to increase the ventilation volume of the fresh air system; if the personnel information shows that the person has allergies, the corresponding rule may be to increase the purification level of the air purifier; if the scene information is a sleeping scene, the corresponding rule may be to reduce the operating noise of the equipment and adjust the indoor temperature and humidity to a range suitable for sleeping. The extracted rules are classified into the categories of multimodal, personnel, and scene information rules.

[0087] Rule integration and conflict detection: Integrate the above three types of rules to form a preliminary rule set. During the integration process, the system will start the conflict detection mechanism to check whether there are contradictions between different types of rules. For example, the multimodal information rule requires increasing the indoor temperature to accelerate air circulation, while the sleep scene rule in the scene information requires lowering the indoor temperature to facilitate sleep, which leads to a conflict.

[0088] Conflict resolution and rule optimization: Once a rule conflict is detected, the system will handle it according to the preset conflict resolution strategy. Common strategies include judging based on the priority of the rules and retaining the rules with higher priority first; or making a compromise based on the actual situation. For example, in the above temperature conflict example, the temperature is set to an intermediate value that takes into account air circulation and sleeping comfort. After the conflict is resolved, the integrated rules are optimized to ensure the rationality and effectiveness of the rule set.

[0089] Step S640, according to the importance of the rule, the scope of applicable scenarios and the relevance to the current scenario, a preset weighted average algorithm is used to calculate a comprehensive score, and the integrated rules are prioritized according to the result of the comprehensive score sorting.

[0090] The operation process is as follows: Indicator quantification: First, quantify the three key indicators of the importance of the rule, the scope of applicable scenarios, and the relevance to the current scenario. For importance, assign a score of 1-5 based on the impact of the rule on user health, comfort, etc., with 5 representing extremely important and 1 representing relatively unimportant. For example, the importance of air purification rules related to the respiratory health of people with severe allergies can be assigned a value of 5; while some lighting adjustment rules that only affect slight environmental comfort can be assigned an importance of 1. The scope of applicable scenarios is quantified based on the proportion of the number of scenarios to which the rule applies to the total number of scenarios. For example, if a rule applies to 10 scenarios and the total number of scenarios is 50, the quantified value of the scope of applicable scenarios is 0.2. The relevance to the current scene is quantified by comparing the degree of match between the scene conditions set by the rule and the current actual scene, with a complete match of 1 and a complete mismatch of 0.

[0091] Weight allocation: Assign weights to each quantitative indicator. Assume that the importance weight is set to 0.5, the applicable scenario weight is set to 0.2, and the relevance weight to the current scenario is set to 0.3. These weights are set based on a large amount of experimental data and actual application experience to ensure that the comprehensive score can accurately reflect the priority of the rule.

[0092] Comprehensive score calculation: The weighted average algorithm is used to calculate the comprehensive score of each rule. The formula is: Comprehensive score = importance quantification value × importance weight + applicable scenario range quantification value × applicable scenario range weight + relevance quantification value with the current scenario × relevance weight with the current scenario. For example, the importance quantification value of a rule is 4, the applicable scenario range quantification value is 0.3, and the relevance quantification value with the current scenario is 0.8. According to the above weight distribution, its comprehensive score = 4×0.5+0.3×0.2+0.8×0.3=2+0.06+0.24=2.3.

[0093] Priority sorting: After calculating the comprehensive scores of all rules, sort the rules from high to low according to the scores. The rule with the highest score has the highest priority and will be executed first when generating the device control strategy; the rule with a lower score has a lower priority and will be considered for execution only after the high-priority rule is met.

[0094] Step S650, using the rule inference engine to generate a preliminary equipment control strategy according to the priority-sorted rules.

[0095] After completing the rule priority sorting in step S640, proceed to step S650. This step uses the rule reasoning engine to generate a preliminary device control strategy based on the prioritized rules through a series of logical reasoning and conditional judgments. The rule reasoning engine is an intelligent module based on artificial intelligence technology. It can understand and process rule sets, derive reasonable device control solutions based on the logical relationship between different rules and the actual situation of the current scene, and provide a basis for achieving precise environmental control.

[0096] The operation process is as follows: Rule reading and parsing: The rule inference engine first reads the prioritized rule set from the rule repository. Each rule is parsed in detail to identify the condition part (such as people sleeping, poor indoor air quality, etc.) and action part (such as reducing equipment noise, increasing air purifier wind speed, etc.) in the rule, and clarify the scope of application and specific requirements of the rule.

[0097] Logical reasoning and policy deduction: Logical reasoning is performed starting from the high-priority rules in the order of priority of the rules. For high-priority rules, if their conditions are met in the current scenario, the reasoning engine will execute the action part of the rule and incorporate it into the preliminary equipment control strategy. For example, the high-priority rule is "When the person has severe allergies and the indoor pollen concentration exceeds the standard, the air purifier turns on the highest wind speed and starts the high-efficiency filtration mode." If the person is detected to have severe allergies and the pollen concentration exceeds the standard in the current scenario, the reasoning engine will add the control action of "the air purifier turns on the highest wind speed and starts the high-efficiency filtration mode" to the preliminary strategy. When the high-priority rules are processed, the low-priority rules are processed in turn, and the low-priority rule actions that meet the conditions are also included in the strategy without conflicting with the high-priority rules.

[0098] Strategy integration and preliminary generation: After reasoning and processing all rules, the control actions corresponding to each rule are integrated to form a preliminary equipment control strategy. This strategy includes specific control instructions for various types of environmental purification and humidification equipment, such as the operation mode and wind speed setting of air purifiers, the humidification amount and working time setting of humidifiers, and the temperature and air volume adjustment of air conditioners.

[0099] The rule inference engine is used to generate preliminary equipment control strategies based on the priority-sorted rules, including: Step S651, continuously collect real-time data of the current scene through various sensors and perform data preprocessing.

[0100] Real-time data collection: In the process of generating preliminary equipment control strategies using the rule inference engine, step S651 plays a basic role in data support. Through various sensors deployed in the environment, various data of the current scene are continuously collected, including but not limited to temperature, humidity, air quality (such as PM2.5, carbon dioxide concentration, etc.), personnel activity status (such as whether they are exercising, whether they are sleeping, etc.), sound intensity and other information. These data are the basis for subsequent decision-making and strategy generation, and can reflect the actual status of the current scene.

[0101] Data preprocessing: Since the data collected by sensors may have different formats, precisions, and ranges, the collected data needs to be preprocessed to facilitate subsequent rule comparison and inference engine processing. Preprocessing mainly involves standardizing and normalizing the data to make the data have uniform specifications and comparability.

[0102] Sensor data collection: The system activates and calls various sensors, including but not limited to temperature and humidity sensors, air quality sensors, cameras (for monitoring personnel activities), microphones (for monitoring sound), etc., to continuously collect data. For example, the temperature and humidity sensor will measure the temperature and humidity of the environment at a certain frequency (such as every 5 seconds) and transmit the data to the data processing unit in real time; the air quality sensor will monitor the concentration of indoor pollutants such as PM2.5 and carbon dioxide in real time; the camera will analyze the activity posture and behavior of personnel through image recognition technology; the microphone will monitor the intensity and frequency of sound in the environment.

[0103] The data preprocessing operations are as follows: Standardization: For different types of data, convert them into a unified unit of measurement or representation. For example, convert temperature data from Fahrenheit to Celsius (if the sensor collects Fahrenheit); adjust the sound intensity from the original decibel value to a specific range (such as 0-100) to facilitate subsequent comparison and calculation.

[0104] Normalization: For numerical data with different ranges, normalize them to the interval [0,1]. Taking carbon dioxide concentration as an example, if its original measurement range is 0-1000ppm, it is normalized to the interval [0,1] through the formula (current concentration - minimum concentration) / (maximum concentration - minimum concentration), so that the data of different sensors have equal importance and comparability in subsequent comparisons.

[0105] Categorical data encoding: For some categorical data, such as the activity status of a person (sleep, exercise, stillness, etc.), encoding can be used for conversion. For example, "sleep" is encoded as 0, "exercise" is encoded as 1, and "stillness" is encoded as 2, so that the categorical data can be processed by the inference engine in numerical form.

[0106] Step S652, compare the preprocessed data with the priority-sorted rule base, and quickly locate all rules that meet the characteristics of the current scene data in order of priority.

[0107] Comparison of data and rule base: This step is to match the data preprocessed in step S651 with the priority-sorted rule base, aiming to find the rules that match the data characteristics of the current scene. By comparing the processed data with the triggering conditions of each rule in the rule base, the rules that meet the current environment can be screened out, providing targeted rule basis for subsequent reasoning decisions.

[0108] Rule positioning: Position the rules according to their priority order, and first check whether the high-priority rules are suitable for the current scenario, so that the subsequent inference engine can execute the rules in a reasonable order, ensuring that the most critical rules are applied first, and improving the effectiveness and rationality of equipment regulation.

[0109] The operation process is as follows: Data reading: First, the system will read the data pre-processed in step S651. These data contain various information of the current scene, such as the temperature, humidity, activity status of personnel, air quality and other standardized and normalized data of the environment.

[0110] Rule base traversal and condition comparison: Traverse the rules in the rule base in order from high to low priority. For each rule, check its trigger condition part. For example, the trigger condition of a rule may be "when the indoor temperature is higher than 25℃ and the person is in motion", the system will compare the current temperature data and the person's activity status data with it. When comparing, the corresponding comparison operation will be performed according to the specific data type. For numerical data, use precise numerical comparison; for categorical data, use the encoded category value for equality comparison.

[0111] Rule matching and screening: If the data of the current scene meets the triggering conditions of a rule, the rule will be marked as matching the current scene. The system will continue to perform such traversal and comparison operations until all rules have been checked and all rules that meet the characteristics of the current scene data have been found. In this process, since the check is carried out in order of priority, high-priority rules will be considered first. Only when the high-priority rules are not met will the low-priority rules be considered to ensure that the screening and matching of rules meet the optimized priority order.

[0112] Step S653, based on the matched rules, the inference engine uses the forward reasoning mechanism, starting from the current scene data, and gradually derives the preliminary equipment control strategy based on the rules, and determines the corresponding equipment operation instructions for each triggered rule.

[0113] Forward reasoning mechanism: In this step, forward reasoning is the process of gradually deriving the device control strategy based on the matched rules starting from the data of the current scene. It is a reasoning method based on known facts (current scene data) and rules. By satisfying the condition part of the rule, the action part of the rule is triggered, and then the corresponding device operation instructions are determined. This mechanism can systematically generate a series of reasonable device control operations based on the logical relationship between scene data and rules, providing specific operation guidelines for achieving precise environmental control.

[0114] Equipment operation instructions: These instructions are determined based on the matched rules and specify the operation details of environmental purification and humidification equipment (such as air purifiers, humidifiers, air conditioners, etc.), including the opening or closing of the equipment, the operating mode, and the working parameters (such as wind speed, humidification amount, temperature setting, etc.), aiming to make the operating status of the equipment meet the needs of the current scenario.

[0115] The operation process is as follows: Rule trigger check: The system will check the rules matched in step S652 one by one. For each matched rule, check whether its triggering condition is fully satisfied by the data of the current scene. If satisfied, the rule will be considered activated and ready for the next reasoning operation. For example, if there is a rule that "when the indoor temperature is higher than 28°C and people are in motion, the air conditioner should turn on the cooling mode, set the temperature to 25°C, and the wind speed to medium", and the current scene data shows that the indoor temperature is 30°C and people are exercising, then the rule will be triggered.

[0116] Equipment operation deduction: When a rule is triggered, the equipment operation instructions are deduced based on the action part of the rule. In the above example, based on the action part of the rule, the inference engine will deduce the operation instructions for the air conditioner, including turning on the air conditioner in cooling mode, setting the temperature to 25°C, and adjusting the wind speed to medium. This deduction operation will be performed for each triggered rule.

[0117] Strategy generation and integration: The device operation instructions derived from different rules are integrated to form a complete preliminary device control strategy. This means that multiple operations on multiple devices may be involved. The system will classify and organize these operations according to the device type and operation type to form a unified set of device control strategies. For example, in addition to the above-mentioned air conditioner operation instructions, there may be instructions such as "air purifier turns on high-efficiency purification mode" and "humidifier temporarily turns off" derived from another rule. These instructions will be organized into a complete policy set, including operation requirements for air conditioners, air purifiers, humidifiers and other equipment.

[0118] The rule inference engine is used to generate preliminary equipment control strategies based on the priority-sorted rules, including: Step S65a, setting the goal of environmental control according to the current scene requirements and user preferences.

[0119] Environmental control target: This is the desired state of the indoor environment determined based on the current scene requirements and user preferences, covering multiple aspects such as temperature, humidity, air quality, noise level, etc. These targets provide clear direction for the subsequent operation of the rule reasoning engine and determine the final desired effect of equipment control, aiming to create a comfortable, healthy and efficient environment for users.

[0120] Scenario requirements: refers to the various environmental requirements in the current scenario. For example, in an office scenario, suitable temperature and good air quality may be required to ensure work efficiency; in a sleeping scenario, a quiet environment and suitable temperature and humidity may be required to ensure sleep quality; in a sports scenario, sufficient fresh air and suitable temperature may be required to ensure the sports experience.

[0121] User preferences: reflect the special needs of individual users. For example, some users may prefer slightly cooler temperatures, higher humidity, or lower noise levels. These preferences will affect the setting of environmental control goals.

[0122] The operation process is as follows: Scenario demand analysis: The system will first analyze the current scene and identify the type and main features of the scene. For example, through camera and sensor data, it can be determined whether the current scene is an office scene (based on the activity posture of the personnel, the use of office equipment, etc.), a sleeping scene (based on the posture of the personnel in bed, the time, the light and sound of the environment, etc.), or a sports scene (based on the movement status of the personnel, the use of sports equipment, etc.). For different scenes, the system will have corresponding preset environmental demand templates. For example, the preset environmental requirements for office scenes may be a temperature between 22°C and 25°C, a humidity between 40% and 60%, and good air quality.

[0123] User preference acquisition: The system will obtain user preference information from the user's setting files or historical operation records. This information can be preferences explicitly set by the user, such as the preferred temperature range, humidity range, noise sensitivity, etc., which are clearly stated in the user settings. It can also be preferences mined from the user's historical usage data. For example, if the user often sets the air conditioner temperature to 23°C, it is inferred that the user prefers this temperature.

[0124] Goal setting: Consider the needs of the current scene and user preferences, and set the final environmental control goal. For example, if the current scene is an office scene, and the user prefers a cooler environment and has a mild history of allergies, then the environmental control goal may be set to a temperature of around 22°C, a humidity of around 50%, and high-quality air quality (such as PM2.5 concentration below 35μg / m³). At the same time, the cleanliness and circulation of indoor air must be guaranteed, and the noise level must be minimized to provide users with a comfortable, healthy, and quiet office environment.

[0125] Step S65b, starting from the set goals, the inference engine reversely searches the rule base for rules that can achieve these goals.

[0126] Reverse search: Unlike forward reasoning, which derives results from known conditions, reverse search starts from the set environmental control goals and searches for rules in the rule base that can achieve these goals. This search method focuses on the realization of the goal and determines the rules that need to be executed to meet the environmental control needs through reverse thinking.

[0127] Rule base: stores a large number of pre-set rule sets, which are formulated based on the logical relationship between various environmental conditions, equipment operations, and the desired environmental state. Each rule contains a condition part and an action part. When the condition is met, the corresponding action will be executed.

[0128] Goal-oriented reasoning: Guided by the environmental control goal, a targeted search is conducted in the rule base to find a series of rules whose execution can guide the current environmental state to gradually transform to the target state.

[0129] The operation process is as follows: Goal decomposition: The inference engine first decomposes the environmental control goal set in step S65a into multiple specific and actionable sub-goals. For example, if the environmental control goal is "maintaining the indoor temperature at 25°C, humidity at 50%, noise below 30 decibels, and good air quality in the sleeping scene", it will be decomposed into sub-goals such as "temperature reaches 25°C", "humidity reaches 50%, noise is reduced to below 30 decibels", and "air quality is optimized to good".

[0130] Rule base traversal: For each sub-goal, the inference engine starts to traverse the rule base. The rules in the rule base are stored according to a certain data structure for quick retrieval. During the traversal process, the inference engine matches each sub-goal with the action part of the rule. The action part of the rule describes the expected change in environmental state after executing the rule. For example, a rule may be "When the air conditioner is set to cooling mode, the temperature is set to 25°C, and the wind speed is low, the indoor temperature can be reduced to about 25°C." The inference engine will recognize that the action part of the rule matches the sub-goal of "the temperature reaches 25°C."

[0131] Rule screening: When a rule matching a sub-goal is found, the inference engine will further check the conditional part of the rule. The rule will only be retained if the conditional part of the rule is likely to be met in the current scenario. For example, for the above temperature-related rules, the conditional part may be "the current indoor temperature is higher than 25°C", and the inference engine will determine whether the condition is likely to be true based on the real-time temperature data of the current scenario. If the current temperature is indeed higher than 25°C, the rule is screened out as a rule that may achieve the goal; conversely, if the current temperature is lower than 25°C, the rule does not meet the requirements and is excluded.

[0132] Record rule chains: For each sub-goal, the inference engine will record all the filtered rules to form a rule chain. These rule chains represent possible paths from the current scene state to the realization of each sub-goal. For example, for the sub-goal of "humidity reaches 50%", it may be filtered out "When the humidifier is turned on, the humidification volume is set to 200 ml per hour, and it continues to run for 2 hours, the indoor humidity can reach about 50% (condition: the current humidity is lower than 50%, and there is sufficient water in the water tank)" and "When the fresh air system is turned on, the air with suitable humidity outside is introduced to adjust the indoor humidity (condition: the humidity of the outside air is suitable, and the indoor and outdoor air circulation conditions allow)" and other rules, and record them to form a rule chain for the humidity sub-goal.

[0133] Step S65c, based on the rule chain obtained by reverse search, the inference engine backtracks to generate a specific device control strategy to achieve the goal.

[0134] Backtracking generation: After the rule chain is obtained by reverse search in step S65b, this step will use the inference engine to backtrack and generate specific device control strategies based on these rule chains. The backtracking process is a refinement process from the goal to the executable operation, which aims to convert the logical relationship in the rule chain into specific operation instructions for the device, so that the environment can achieve the desired control goal from the current state.

[0135] Equipment control strategy: It specifies the operating details of various environmental purification and humidification equipment (such as air purifiers, humidifiers, air conditioners, ventilation systems, etc.), including turning on / off the equipment, adjusting the operating mode, setting working parameters, etc., to achieve precise control of the environment and meet the set environmental goals.

[0136] The operation process is as follows: Rule chain analysis: The inference engine first performs a detailed analysis of the rule chain obtained in step S65b. For each rule chain corresponding to each sub-goal, the engine will check the logical relationship and execution conditions of each rule in the chain. For example, for the sub-goal of "adjusting the indoor temperature to 25°C", its rule chain may include "when the indoor temperature is higher than 25°C, the air conditioner turns on the cooling mode with a low wind speed, and continues to run for a period of time to reduce the temperature" and "when the temperature difference between indoor and outdoor is large, open the windows and use natural ventilation to reduce the indoor temperature (assuming this rule exists)".

[0137] Backtracking reasoning operation: Starting from the final environmental control goal, the reasoning engine gradually determines the specific device operation according to the rules in the rule chain. For the above temperature adjustment example, if the current indoor temperature is higher than 25°C and the current scene meets the conditions of the first rule, the engine will give priority to this rule and convert it into a device operation instruction, such as sending a command to the air conditioner to turn on the cooling mode and set a low wind speed.

[0138] During the backtracking process, the order and dependencies between rules will be considered. If there are multiple rules, they may need to be executed in a certain order. For example, open the window for ventilation first, and then decide whether to turn on the air conditioner based on the temperature changes after a period of time. At the same time, for operations involving multiple devices, the operation sequence and time intervals of these devices will be coordinated. For example, the air purifier may be turned on to purify the indoor air for a period of time, and then the working state of the humidifier may be adjusted according to the air quality to ensure the consistency and effectiveness of the equipment operation.

[0139] For some complex rule chains, conditional judgments and loop operations may be involved. For example, in some rules, it is stipulated that "if the environmental parameters do not meet expectations after the device is turned on for a period of time, adjust the device working parameters to continue running", the inference engine will judge and process this situation, continuously monitor environmental data, and make decisions on subsequent operations based on actual conditions.

[0140] Strategy integration: Integrate the device operation instructions generated by the rule chains of different sub-goals to form a complete device control strategy. This includes detailed information such as the opening sequence, operating time, working mode and parameter settings of different devices. For example, the final device control strategy may include: open the window for ventilation for 5 minutes first, if the indoor temperature is still above 25℃, close the window, turn on the air conditioner, set it to cooling mode, and the wind speed is low; turn on the air purifier at the same time, set it to medium purification mode; according to the indoor humidity, turn on or adjust the humidification amount of the humidifier, etc.

[0141] The intelligent environment purification and humidification integrated linkage control method also includes steps after using the rule reasoning engine to generate a preliminary equipment control strategy according to the priority-sorted rules, which are as follows: Step S660, obtaining user feedback data, and using natural language processing technology to convert the user's text feedback into quantitative rule adjustment requirements.

[0142] User feedback data: Feedback information about the integrated control effect of environmental purification and humidification provided to the system through various channels by users based on their own feelings and actual experience of the environment. These feedbacks can be evaluations of the current indoor temperature, humidity, and air quality, opinions on the operating status of the equipment (such as noise, wind speed), and feelings about the overall environmental comfort, etc. Feedback forms include but are not limited to text descriptions, voice evaluations, etc.

[0143] Natural language processing technology: a series of technologies designed to enable computers to understand, process and generate human language. In this step, it is mainly used to convert the user's text feedback expressed in natural language into quantitative information that can be understood and processed by computers, so as to make targeted adjustments to the rules later.

[0144] Quantified rule adjustment requirements: Convert the user's vague natural language feedback into specific, measurable requirements for rule adjustment. For example, convert "feeling that the room is too dry" into "need to increase the indoor humidity by X%", where X is the quantified value, so as to clarify the direction and degree of rule adjustment.

[0145] The operation process is as follows: Feedback data collection operates as follows: Multi-channel collection: The system obtains user feedback data in a variety of ways. On the one hand, a special feedback portal is set up in the application interface of the smart device, and users can directly enter text feedback in this portal. For example, on the mobile phone APP of the environmental purification and humidification equipment, there is a "user feedback" button, which can be clicked to enter the text input box. On the other hand, voice feedback is supported. Users can speak their feedback content through the built-in microphone of the device, and the system automatically converts the voice into text. In addition, feedback data can also be collected from users' discussions about the environmental control system on relevant forums and social media.

[0146] Data collation and storage: The collected feedback data will be collated and stored in a dedicated database. The data will be classified and stored according to time, user ID, feedback channel and other information to facilitate subsequent query and processing. For example, different folders are created according to user ID, and user feedback records are stored in chronological order in each folder.

[0147] Natural language processing is as follows: Text preprocessing: Preprocess the collected text feedback to improve processing efficiency and accuracy. This includes removing noise information in the text, such as special characters, HTML tags (if the feedback is collected from web pages), etc.; performing lexical analysis to split the text into individual words or phrases; performing part-of-speech tagging to determine the part of speech of each word (such as noun, verb, adjective, etc.). For example, for the feedback text "The indoor humidity seems to be a bit low, and I feel uncomfortable", after preprocessing, the word sequence ["indoor", "humidity", "seems", "a bit", "low", "feeling", "uncomfortable"] can be obtained, and the part of speech of each word is marked.

[0148] Semantic understanding: Use semantic analysis technology in natural language processing to understand the actual meaning of user text feedback. This may involve named entity recognition, identifying key entities mentioned in the text, such as environmental parameters (temperature, humidity, air quality, etc.), device names (air purifiers, humidifiers, etc.), scene modes (sleep mode, office mode, etc.). For example, "humidity" is identified as an environmental parameter entity from the above text. At the same time, semantic role labeling is used to determine the semantic role of each word in the sentence, such as "low" is an attribute that describes the "humidity" state.

[0149] Quantitative conversion: Based on the results of semantic understanding, user feedback is converted into quantitative rule adjustment requirements. This requires combining the preset conversion rules and knowledge base. For example, if the knowledge base stipulates that "a bit low" corresponds to a 5%-10% increase in humidity, then "the indoor humidity seems a bit low" can be quantified as "increase the indoor humidity by 7.5%." For some complex feedback, reasoning and calculation may be required. For example, "It feels much drier than yesterday, and the humidity was 40% yesterday." By comparing with historical data and preset conversion rules, it can be quantified as "increase the indoor humidity to 45%-50%."

[0150] Step S670, classifying the quantified rule adjustment requirements, the rule adjustment requirements are divided into environmental parameter adjustment requirements, device operation requirements, and scene mode adjustment requirements.

[0151] Quantifying rule adjustment requirements: This refers to converting the user's feedback expressed in natural language into specific, measurable requirements for rule adjustment through step S660. These requirements are the basis for subsequent classification and processing, and they clarify the direction and degree of change that the user expects to make to the environmental control rules.

[0152] Environmental parameter adjustment requirements: mainly involve the need to change physical parameters related to the indoor environment, such as temperature, humidity, air quality indicators (such as PM2.5, formaldehyde concentration), light intensity, etc. This type of demand aims to make the environmental parameters reach a state that is more satisfactory to the user by adjusting the rules.

[0153] Equipment operation requirements: Requirements related to specific operating instructions for environmental purification and humidification equipment, including turning the equipment on or off, switching the operating mode (such as switching between the standard mode, sleep mode, and strong purification mode of the air purifier), changing the equipment's operating parameters (such as adjusting the humidification volume of the humidifier and the wind speed of the air conditioner), etc. These requirements directly affect the operation mode of the equipment to meet the user's expectations for the equipment's functions.

[0154] Scene mode adjustment requirements: requirements related to environmental control modes in different scenarios. Different scenarios (such as sleeping, working, leisure and entertainment, etc.) usually have their own suitable environmental parameters and equipment operation configurations. The scene mode adjustment requirements are that users want to change the control rules corresponding to these scenarios to better match their personal needs in specific scenarios.

[0155] The operation process is as follows: Demand analysis: The system first analyzes the quantitative rule adjustment demand generated in step S660 in detail. By analyzing the demand content, key information is extracted, such as the object involved (environmental parameters, equipment or scene mode), the expected change direction (increase, decrease, switch, etc.) and the specific quantitative value (if any). For example, for the demand "increase indoor humidity to 50%", the system recognizes that the object involved is the environmental parameter "humidity", the change direction is "increase", and the quantitative value is "50%".

[0156] Classification of environmental parameter adjustment requirements: Parameter identification: Check whether specific environmental parameters are mentioned in the requirements. If environmental parameter-related words such as "temperature", "humidity", "PM2.5 concentration" are identified, and the requirements revolve around the value change or range adjustment of the parameter, they are classified as environmental parameter adjustment requirements. For example, requirements such as "lower the indoor temperature by 2°C" and "control the PM2.5 concentration below 30μg / m³" are clearly environmental parameter adjustment requirements.

[0157] Parameter range definition: For requirements involving environmental parameters, further determine the reasonable range of the parameters. Based on the preset knowledge base, the system determines whether the parameter adjustment proposed by the user is within a safe and reasonable range. For example, for indoor temperature, the general reasonable range is between 16℃-30℃. If the user's requirements exceed this range, the system may prompt the user to confirm or make corresponding corrections.

[0158] Equipment operation requirements classification: Equipment name identification: Check whether the demand contains the name of the environmental purification and humidification equipment, such as "air purifier", "humidifier", "air conditioner", etc. If the equipment name is mentioned and the demand revolves around the operation of the equipment (such as turning on, off, adjusting the mode, changing parameters, etc.), it will be classified as an equipment operation demand. For example, demands such as "adjust the air purifier wind speed to high" and "turn off the humidifier" are equipment operation demands.

[0159] Operation confirmation: For the needs involving equipment operation, the specific operation and its rationality are clarified. The system determines whether the operation required by the user is feasible based on the functions and operating specifications of the equipment. For example, for an air purifier, if the user requires the wind speed to be set to a level beyond the level supported by the device itself, the system will handle it accordingly, such as prompting the user to select the correct range of levels.

[0160] Classification of scene mode adjustment requirements: Scene mode identification: Identify whether a specific scene mode is mentioned in the demand, such as "sleep mode", "office mode", "sports mode", etc. If the demand revolves around changes in the environmental parameter settings of these scene modes, the device operating status, etc., it is classified as a scene mode adjustment demand. For example, "In sleep mode, set the air conditioning temperature to 26°C and the air purifier to silent mode" is a scene mode adjustment demand.

[0161] Scene mode association: Associating scene mode adjustment requirements with the scene modes preset in the system, confirming the specific scene mode and its related rules that the user wants to change. The system checks whether the user's requirements match the functions and features of the existing scene mode. If there are conflicts or unreasonable aspects, corresponding prompts or processing will be given.

[0162] Step S680, based on the urgency of the demand and the impact on the user experience, a pre-built evaluation matrix is ​​used to determine the priority of each demand.

[0163] Urgency of demand: refers to the urgency with which the rule adjustment demand needs to be processed. If some demands are not handled in a timely manner, they may have a significant impact on the user's life, health or the normal operation of the equipment. Such demands have a high degree of urgency. For example, if the indoor air quality is seriously exceeded, the user feedback needs to immediately improve the purification capacity of the air purifier. This demand has a high degree of urgency.

[0164] Impact on user experience: Measures the degree to which the rule adjustment requirements change the user experience in terms of environmental comfort, ease of use, etc. Although some requirements are not urgent, they may significantly improve or reduce the user experience and also need to be focused on. For example, users hope that the device will run with less noise in sleep mode, which is crucial to improving the sleep experience. Even if it does not involve urgent environmental safety issues, the impact on the user experience is still high.

[0165] Evaluation Matrix: A pre-built data structure used to comprehensively evaluate the priority of each rule adjustment requirement. The rows and columns of the matrix represent different evaluation factors (such as urgency, impact on user experience, etc.) and their different levels of severity. The priority of requirements under different combinations of factors is determined by the values ​​of the matrix elements. For example, the values ​​of the elements in the matrix corresponding to "high urgency" and "high impact on user experience" may be set to the highest priority.

[0166] The operation process is as follows: Urgency Assessment: Set the urgency level: The system divides the urgency of the demand into different levels, such as "high", "medium" and "low". For demands involving key issues such as user health and safety, they are judged to be "high" urgency. For example, when a user reports that the indoor formaldehyde concentration is too high, which may pose a serious threat to health, the urgency of such demands is "high". If the demand has a certain impact on the environment or equipment operation, but will not immediately have serious consequences, it is judged to be "medium" urgency. For example, if a user reports that the air purifier filter is clogged, it will not affect health immediately, but it will reduce the purification effect in the long run. The urgency of such demands is "medium". For some general demands that do not affect normal use and health and safety, they are judged to be "low" urgency. For example, if a user wants to fine-tune the humidification amount of the humidifier, it will have little impact on the environment in the short term, and the urgency is "low".

[0167] Judgment based on standards: The system evaluates each quantified rule adjustment requirement based on the preset urgency judgment standards. These standards are formulated based on multiple factors such as industry specifications, equipment operating characteristics, and user health and safety. For example, for environmental parameter requirements, refer to relevant environmental quality standards. When indoor environmental parameters deviate seriously from the safe range, it is judged to be of high urgency. For equipment operation requirements, if the equipment fails or there are operational requirements that may affect the normal operation of the equipment, it is judged to be of medium to high urgency.

[0168] User experience impact assessment: Define the level of impact: The impact on the user experience is also divided into three levels: "high", "medium" and "low". Requirements that can significantly improve or seriously reduce user comfort and convenience are judged to have a "high" impact level. For example, users reported that the device noise was too loud in sleep mode, which seriously affected the quality of sleep. The impact of this requirement on the user experience is "high". Requirements that have a certain improvement or impact on the user experience, but the degree is relatively small, are judged to have a "medium" impact level. For example, users hope to fine-tune the ambient temperature in office mode to make the office experience more comfortable. The impact of such requirements on the user experience is "medium". Requirements that have a slight impact on the user experience are judged to have a "low" impact level. For example, users hope to adjust the brightness of the device display panel, which has little impact on the overall experience and the impact level is "low".

[0169] User feedback and analysis: The system analyzes user feedback content and past user behavior data to determine the impact of the demand on the user experience. The impact is determined by analyzing factors such as the tone and wording of user feedback and the frequency of feedback on similar demands from past users. For example, if a large number of users have reported a problem about the ease of device operation, the impact of this demand on the user experience may be judged to be high.

[0170] Use the evaluation matrix to prioritize: Matrix matching: Use the urgency of each requirement and the degree of impact on user experience as the index of the evaluation matrix to find the corresponding matrix element. For example, if the urgency of a requirement is "high" and the degree of impact on user experience is "medium", find the element corresponding to the combination of these two factors in the evaluation matrix.

[0171] Determine the priority: Determine the priority of the requirement based on the value of the corresponding element in the evaluation matrix. The value of the matrix element can be a digital code, such as 1-5, and the larger the number, the higher the priority. For example, the matrix element value corresponding to "high urgency" and "medium impact on user experience" is 4, which means that the priority of this requirement is level 4. In this way, the system determines the priority of all quantified rule adjustment requirements so that the rules can be adjusted in order of priority later.

[0172] Step S690: modify the rules in the rule base accordingly according to the analysis result of the rule adjustment requirement.

[0173] Operation process: Demand priority sorting and retrieval: The system adjusts the demand priority according to the rules determined in step S680, and retrieves the demands from high to low. High-priority demands are processed first to ensure that issues that have a greater impact on user experience and environmental regulation are resolved in a timely manner. For example, if the high-priority demand is "reduce device noise to less than 30 decibels in the sleep scene", the system will first search and modify the rule library for this demand.

[0174] Rule location and matching: For each requirement, the system locates the relevant rules in the rule base. By parsing the requirements, key information is extracted, such as the environmental parameters, device types, scene modes, etc. involved, and then matched with the conditions and actions of the rules in the rule base. For example, for the above requirement to reduce the noise of equipment in the sleeping scene, the system will search for all rules related to the sleeping scene and equipment noise in the rule base.

[0175] Rule modification: According to the demand analysis results, the located rules are modified accordingly. The modification methods include but are not limited to adjusting the triggering conditions of the rules, the execution instructions of the action part, and the priority of the rules.

[0176] Trigger condition adjustment: If the demand is for stricter environmental parameters in a specific scenario, the trigger condition of the rule may need to be modified. For example, the original rule is "when the indoor temperature is higher than 28°C, turn on the air conditioning for cooling", if the user feedback wants to turn on the cooling at 26°C, then the temperature threshold in the trigger condition is changed to 26°C.

[0177] Modify execution instructions: For device operation requirements, directly modify the device execution instructions in the rule action part. For example, if the requirement is "increase the wind speed of the air purifier to the highest level", find the corresponding rule and modify the wind speed setting instruction of the air purifier to the highest level.

[0178] Priority adjustment: If the priority setting of some rules is found to be unreasonable during the analysis process, the priority of the rules will be readjusted according to the needs. For example, the original priority of the rule for general environment purification is higher than the low-noise rule for the sleeping scene. According to user feedback, the low-noise rule for the sleeping scene should be given higher priority, so the priorities of the two are swapped.

[0179] Rule verification and update: After modifying the rules, the system verifies the new rules. By simulating different environmental scenarios and input data, check whether the rules can be triggered correctly and whether the execution results meet expectations. If the verification passes, the modified rules are updated to the rule base; if the verification fails, analyze the reasons and modify the rules again until the rules meet the requirements.

[0180] Based on the same inventive concept, an embodiment of the present invention provides an intelligent environment purification and humidification integrated linkage control system, including a memory and a processor, wherein the memory stores a program that can be executed on the processor to implement the following Figures 1 to 4 Procedure for either method.

[0181] The embodiments of this specific implementation method are all preferred embodiments of the present application, and are not intended to limit the protection scope of the present application. Therefore, all equivalent changes made based on the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for integrated linkage control of intelligent environmental purification and humidification, characterized in that: include: Obtain information about people in the room and the scene they are in; Filter out key data from personnel information and scene information; Compare the filtered personnel information and scene information with the rules in the information matching rule library one by one; If the comparison is completely consistent, the matching personnel information, scene information and corresponding rule content are integrated to form a personalized information set; If there is no complete match, the preset similarity calculation algorithm is used to calculate the overall matching degree of the rules in the information matching rule library one by one, and the rule associated with the highest overall matching degree is selected as the successfully matched rule, and the matched personnel information, scene information and corresponding rule content are integrated to form a personalized information set; The personalized information set is input into a pre-built relevant device control strategy analysis model, the relevant device control strategy is output, and the relevant device control strategy is executed.

2. The intelligent environment purification and humidification integrated linkage control method according to claim 1 is characterized in that: It also includes steps parallel to filtering out key data from personnel information and scene information, as follows: Analyze whether key data related to scene recognition and conversion is obtained, including the activity posture of people, the type and intensity of sound, and precise location; If no, keep the original setting. If yes, then determine whether the person moves between rooms based on the key data obtained related to scene recognition and conversion; If yes, it is determined that there is a scene transition; If not, keep the original setting.

3. The intelligent environment purification and humidification integrated linkage control method according to claim 2 is characterized in that: Forming a personalized information set includes: Analyze whether there is a scene transition; If not, maintain the original setting; If yes, obtain the multimodal information involved in this scene conversion, and match the rules corresponding to the multimodal information of this scene conversion from the information matching rule library, integrate the multimodal information of this scene conversion and the corresponding rules and matching personnel information, scene information and corresponding rule content to form a personalized information set.

4. The intelligent environment purification and humidification integrated linkage control method according to claim 3 is characterized in that: The personalized information set is input into the pre-built relevant equipment control strategy analysis model, and the output relevant equipment control strategies include: Analyze whether the personalized information set contains multimodal information and rules corresponding to the multimodal information; If not, maintain the original setting; If yes, then the rules corresponding to the multimodal information, the rules corresponding to the personnel information, and the rules corresponding to the scene information are integrated; According to the importance of the rules, the scope of applicable scenarios and the relevance to the current scenario, a preset weighted average algorithm is used to calculate the comprehensive score, and the integrated rules are prioritized according to the results of the comprehensive score ranking; The rule inference engine is used to generate a preliminary equipment control strategy based on the priority rules.

5. The intelligent environment purification and humidification integrated linkage control method according to claim 4 is characterized in that: The rule inference engine generates preliminary equipment control strategies based on the priority-sorted rules, including: Continuously collect real-time data of the current scene through various sensors and perform data preprocessing; Compare the preprocessed data with the prioritized rule base, and quickly locate all rules that meet the data characteristics of the current scenario in order of priority; According to the matched rules, the inference engine uses the forward reasoning mechanism, starting from the current scene data, and gradually derives the preliminary equipment control strategy based on the rules. For each triggered rule, it determines the corresponding equipment operation instructions.

6. The intelligent environment purification and humidification integrated linkage control method according to claim 4 is characterized in that: The rule inference engine generates preliminary equipment control strategies based on the priority-sorted rules, including: Set environmental control goals based on current scene requirements and user preferences; Starting from the set goals, the inference engine reversely searches the rule base for rules that can achieve these goals; Based on the rule chain obtained by reverse lookup, the inference engine backtracks to generate specific equipment control strategies to achieve the goal.

7. The intelligent environment purification and humidification integrated linkage control method according to claim 4 is characterized in that: The steps are as follows: Obtain user feedback data and use natural language processing technology to convert user text feedback into quantitative rule adjustment requirements; Classify the quantified rule adjustment requirements into environment parameter adjustment requirements, device operation requirements, and scene mode adjustment requirements; Prioritize each requirement based on its urgency and impact on user experience using a pre-built evaluation matrix; According to the analysis results of the rule adjustment requirements, the rules in the rule base are modified accordingly.

8. An intelligent environmental purification and humidification integrated linkage control system, characterized in that: It includes a memory, a processor, and a program stored in the memory and executable on the processor, wherein the program can be loaded and executed by the processor to implement an intelligent environment purification and humidification integrated linkage control method as described in any one of claims 1 to 7.

Images