Illuminating lamp system with air quality detection function

Abstract

The invention provides a lighting lamp system with an air quality detection function, and relates to the field of lamps. The system is composed of eight core modules. The air quality detection device is arranged in the lamp and is used for continuously monitoring the air quality; the user data acquisition module collects user health information; the indicator light control rule generation module creates a personalized control rule based on the user health data and the knowledge graph; the control module processes the data and generates an indicator light control signal; the alarm indicating lamp visually displays the air quality condition through color change; the internet-of-things control module automatically controls the environment adjusting equipment when the air quality is reduced; the machine learning prediction module analyzes real-time data to predict future air quality changes; and the mobile terminal interaction module realizes two-way communication with a user mobile phone. The system combines environment monitoring, data analysis, machine learning and Internet of Things technologies, and provides personalized air quality monitoring and automatic regulation and control services for users.

Description

Technical Field

[0001] This application relates to the field of lighting fixtures, and particularly to a lighting fixture system with an air quality detection function. Background Art

[0002] With the acceleration of urbanization and the increasing prominence of environmental pollution problems, indoor air quality has become an important factor affecting people's health. Research shows that people spend an average of about 80%-90% of their time indoors, and indoor air often contains various pollutants such as formaldehyde, volatile organic compounds, PM2.5, etc. These pollutants may cause respiratory discomfort, allergic reactions, and even more serious health problems. Therefore, the real-time monitoring and management of indoor air quality become particularly important.

[0003] As an essential facility indoors, lighting fixtures are almost everywhere in homes and workplaces. At the same time, air quality detection devices, as independent devices, are also gradually becoming popular. Currently, some lighting fixtures integrating air quality detection functions have emerged on the market. Such products usually integrate sensors inside the fixtures, can display basic air quality parameters such as temperature, humidity, PM2.5, etc., and reflect the air quality status by changing the light color or brightness.

[0004] However, the existing lighting fixtures integrating air quality detection functions have a significant defect: they cannot provide personalized air quality monitoring and warning services according to the individual health conditions of users. These products generally use a unified standard threshold to judge the quality of air, ignoring the significant differences in the sensitivity and needs of different users to air quality. Summary of the Invention

[0005] This application provides a lighting fixture system with an air quality detection function to achieve personalized air quality monitoring and warning based on the individual health conditions of users.

[0006] In a first aspect, this application provides a lighting fixture system with an air quality detection function, and the system includes: An air quality detection device, a user data acquisition module, a control module, an indicator light control rule generation module, an alarm indicator light, an Internet of Things control module, a machine learning prediction module, and a mobile terminal interaction module; The air quality detection device is arranged inside the lighting fixture and is used for continuously collecting real-time air quality data reflecting the air quality; The user data acquisition module is used for acquiring the user's health data; The indicator light control rule generation module is used for determining the target indicator light control rule based on the user health data, the Internet of Things control module, and a preset health environment knowledge graph; The control module, connected to the air quality detection device and the user data acquisition module, is configured to generate an alarm indicator control signal according to the target indicator control rule, user data, and real-time air quality data; The alarm indicator, connected to the control module, is configured to display different colors according to the alarm indicator control signal of the control module; The Internet of Things control module, connected to the control module, is configured to communicate with multiple environmental control devices through a wireless communication network, receive the real-time air quality data sent by the control module, and send a device control instruction to the environmental control devices when the real-time air quality data is lower than a preset air quality threshold, where the environmental control devices include air purification devices and ventilation facilities; The machine learning prediction module, connected to the control module, is configured to obtain real-time air quality data and real-time weather data, input the real-time air quality data and the real-time weather data into a preset air quality prediction model to generate air quality prediction data, and generate a warning message when the air quality prediction data is lower than the preset air quality threshold; The mobile terminal interaction module, connected to the control module, is configured to establish a data connection with a mobile terminal through a wireless communication network, send the real-time air quality data, the air quality prediction data, and the warning message to the mobile terminal, and receive a user parameter setting instruction sent by the mobile terminal.

[0007] In the above technical solution, the lighting fixture system with air quality detection function in this application realizes real-time air quality monitoring by setting the air quality detection device inside the fixture. At the same time, combined with the user health data obtained by the user data acquisition module, the air quality monitoring can be associated with the individual health status of the user, laying a foundation for personalized monitoring. The indicator light control rule generation module determines the target indicator light control rule based on the user health data, the Internet of Things control module, and the preset healthy environment knowledge graph, realizing the personalized customization of the air quality evaluation standard, enabling the system to dynamically adjust the air quality judgment standard according to the health status of different users. The control module integrates the target indicator light control rule, user data, and real-time air quality data to generate an alarm indicator light control signal, and the alarm indicator light displays different colors according to this control signal, intuitively reflecting the air quality status after personalized evaluation. The Internet of Things control module establishes connections with multiple environmental control devices such as air purification equipment and ventilation facilities through a wireless communication network, and automatically triggers environmental control when the real-time air quality data is lower than the preset air quality threshold, forming an intelligent closed-loop system from monitoring to control. The machine learning prediction module analyzes the real-time air quality data and real-time weather data to predict the future air quality change trend, and generates a warning message when the air quality prediction data is lower than the preset air quality threshold, realizing the function upgrade from passive response to active prevention. The mobile terminal interaction module realizes the seamless connection between the system and the user's mobile terminal, transmits the real-time air quality data, air quality prediction data, and warning information to the user, and receives the user parameter setting instruction, enhancing the interactivity and user experience of the system. Through the collaborative work of this series of modules, this system not only solves the problem that traditional air quality monitoring devices cannot perform personalized monitoring according to the individual health status of users, but also realizes a complete functional chain from monitoring, prediction, warning to automatic control, significantly improving the intelligent level and accuracy of indoor air quality management, and providing a safer and healthier indoor environment guarantee for users, especially special groups sensitive to air quality.

[0008] In summary, one or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: By using a multi-source data acquisition model to collect vibration data, acoustic data, exhaust gas composition data, and temperature distribution data during the operation of the engine, comprehensive monitoring of the engine state is achieved, effectively avoiding the problem that a single data source is easily affected by the environment. Through the heterogeneous data deep fusion module, data conversion is performed on the multi-source data to obtain a unified multi-source feature representation, which not only solves the problem of inconsistent time series of heterogeneous data but also extracts the deep correlation features between the data. Combining with the multi-level diagnosis module, multi-level diagnosis is performed on the unified multi-source feature representation. Through the progressive process of feature extraction, time series analysis, knowledge reasoning, and diagnostic decision-making, the accuracy and interpretability of fault diagnosis are improved. Finally, based on the multi-level diagnosis results, the predictive maintenance module conducts maintenance analysis. Through the comprehensive evaluation of the predicted life value, fault evolution path, and optimal maintenance time, the transformation from passive maintenance to active predictive maintenance is realized, which not only reduces the maintenance cost but also improves the maintenance efficiency. This complete technical solution from data acquisition, feature fusion to diagnostic prediction effectively solves the technical problem of low accuracy of traditional single-data-source diagnosis methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 FIG. is a schematic structural diagram of a lighting fixture with an air quality detection function provided by an embodiment of the present application; Figure 2 FIG. is a schematic structural diagram of a control rule generation module provided by an embodiment of the present application; Figure 3 FIG. is a schematic structural diagram of a multi-level diagnosis module provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0010] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of this specification. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments.

[0011] In the description of the embodiments of the present application, words such as "for example" or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design solution described as "for example" or "for instance" in the embodiments of the present application should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Exactly speaking, the use of words such as "for example" or "for instance" is intended to present related concepts in a specific manner.

[0012] In the description of the embodiments of the present application, the term "plural" means two or more. For example, plural systems refer to two or more systems, and plural screen terminals refer to two or more screen terminals. In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of such features. The terms "include", "comprise", "have" and their variants all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0013] Please refer to Figure 1 , Figure 1 which is a schematic structural diagram of a lighting fixture with an air quality detection function provided by an embodiment of the present application. This system can be implemented depending on a computer program or run as an independent tool - type application. Specifically, in the embodiments of the present application, this method can be applied to a server, but can also be applied to electronic devices such as a server. A lighting fixture system with an air quality detection function includes the following modules: an air quality detection device 1, a user data acquisition module 2, a control module 3, an indicator light control rule generation module 4, an alarm indicator light 5, an Internet of Things control module 6, a machine learning prediction module 7, and a mobile terminal interaction module 8; The air quality detection device is arranged inside the lighting fixture and is used to continuously collect real - time air quality data reflecting the air quality; Specifically, the air quality detection device 1 is arranged inside the lighting fixture. This arrangement makes full use of the characteristics that the lighting fixture is usually located at the center of the room and works continuously, ensuring the representativeness of the indoor air sampling. The air quality detection device 1 continuously collects real - time air quality data reflecting the air quality through a variety of built - in sensors, including multiple parameters such as PM2.5, PM10, carbon dioxide, formaldehyde, temperature, and humidity. These sensors are connected to the control module 3 through circuits and transmit the collected data in real - time for processing.

[0014] The user data acquisition module is used to acquire the user's health data; Specifically, the user data acquisition module 2 acquires the user's health data through the connection with the user's mobile terminal or a dedicated input interface. The user health data includes information such as the user's age, gender, whether suffering from asthma or allergic diseases, respiratory disease history, and daily health detection data. The purpose of obtaining these data is to establish a digital file of the user's health status as the basic basis for personalized air quality monitoring.

[0015] The indicator light control rule generation module is used to determine the target indicator light control rule based on the user health data, the Internet of Things control module, and the preset healthy environment knowledge graph; Specifically, the indicator light control rule generation module 4 determines the target indicator light control rule based on the user health data, the environment control information provided by the Internet of Things control module 6, and the preset healthy environment knowledge graph. The healthy environment knowledge graph is a structured knowledge base that contains the sensitivity relationship and safety threshold of various air quality parameters under different health conditions. The indicator light control rule generation module 4 maps the user health data to the concepts in the knowledge graph through semantic analysis technology to form a personalized air quality evaluation standard, and converts these standards into specific indicator light control rules, such as setting a lower PM2.5 threshold for asthma patients.

[0016] Based on the above embodiments, as an alternative embodiment, the indicator light control rule generation module further includes: a knowledge data acquisition module, a knowledge graph construction module, and a control rule generation module; The knowledge data acquisition module is used to acquire healthy environment knowledge data; Specifically, the knowledge data acquisition module 41 is used to acquire healthy environment knowledge data. The healthy environment knowledge data refers to a collection of professional knowledge related to the impact of air quality on human health, including medical literature, environmental science research results, health guidelines, and expert suggestions. The knowledge data acquisition module 41 connects to an external knowledge base or professional data platform through a data interface to regularly update the healthy environment knowledge data. This module also has a built-in data crawler function that can automatically obtain the latest research results and guidance suggestions from the public websites of authoritative medical institutions and environmental monitoring organizations. The acquired data is preprocessed and classified for storage for subsequent use. This continuous update mechanism ensures that the system always works based on the latest healthy environment research results.

[0017] The knowledge graph construction module is used to construct a healthy environment knowledge graph based on the healthy environment knowledge data; Specifically, the knowledge graph construction module 42 is used to construct a healthy environment knowledge graph based on the healthy environment knowledge data. The healthy environment knowledge graph is a structured semantic network used to represent the complex association relationships between air quality parameters and health conditions. The knowledge graph construction module 42 first performs natural language processing on the healthy environment knowledge data provided by the knowledge data acquisition module 41, extracts key entities such as "PM2.5", "asthma", "allergy", etc., and the relationships between entities such as "causes", "aggravates", "threshold", etc. Subsequently, the knowledge graph construction module 42 applies entity recognition and relationship extraction algorithms to organize these entities and relationships into a network structure, where each node represents an entity and the edge represents the relationship between entities, and assigns semantic weights. For example, construct the association between "PM2.5 concentration exceeds 50 μg / m³" and "the risk of symptom aggravation in asthma patients increases by 30%", and quantify the strength of this association. This structured representation makes complex medical and environmental knowledge machine - understandable and provides a scientific basis for setting personalized air quality thresholds.

[0018] The control rule generation module is used to determine the target indicator light control rule based on the user's health data and the healthy environment knowledge graph.

[0019] Specifically, the control rule generation module 43 is used to determine the target indicator light control rule based on the user's health data and the healthy environment knowledge graph. The control rule generation module 43 first receives the user's health data transmitted by the user data acquisition module 2, and extracts the user's health characteristics, such as tags like "asthma patient", "elderly people over 65 years old", or "sensitive to formaldehyde". Subsequently, the control rule generation module 43 retrieves the nodes related to these health characteristics in the healthy environment knowledge graph, and calculates the degree of influence of different air quality parameters on this user through graph algorithms. For example, for asthma patients, the system will lower the thresholds of PM2.5 and PM10; for pregnant women, the system will increase the sensitivity to formaldehyde and benzene. Based on these personalized impact assessments, the control rule generation module 43 generates specific indicator light control rules, such as "when the PM2.5 concentration exceeds 30 μg / m³, the alarm indicator light shows yellow; when it exceeds 50 μg / m³, it shows red". These rules are transmitted to the control module 3 to guide the color display of the alarm indicator light 5.

[0020] Based on the above - mentioned embodiment, as an optional embodiment, the control rule generation module includes: a health feature extraction module, a semantic association module, a control rule element graph generation module, and a target indicator light control rule generation module; The health feature extraction module is used to perform semantic extraction on the user's health data by using natural language processing technology to obtain key health features; Specifically, the health feature extraction module 431 is used to perform semantic extraction on user health data using natural language processing technology to obtain key health features. The health feature extraction module 431 receives user health data from the user data acquisition module 2, which may include structured information (such as numerical indicators like age, gender, body mass index, etc.) and unstructured text (such as medical history descriptions, symptom records, doctor diagnoses, etc.). For unstructured text, the health feature extraction module 431 uses natural language processing technology for analysis. First, it splits the text into basic semantic units through a word segmentation algorithm, and then uses named entity recognition technology to identify medical entities such as disease names, symptom descriptions, and drug names in the text. Subsequently, the health feature extraction module 431 applies dependency syntax analysis to understand the relationships between these entities, for example, to distinguish the semantic differences between "having a history of asthma" and "having no history of asthma". To improve the extraction accuracy, the health feature extraction module 431 has a built-in medical vocabulary library and a health status classification system, which can identify common health problems and their synonymous expressions. Finally, the health feature extraction module 431 standardizes the extracted information into a set of key health features, such as medically significant labels like "asthma patient", "allergic constitution", "cardiopulmonary insufficiency", "pregnant woman", "infant", etc. This method of health feature extraction based on natural language processing greatly improves the automation and accuracy of data processing, overcomes the inconvenience of the traditional method that requires users to manually fill in complex forms, and enables the system to capture key health information from users' daily descriptions.

[0021] The semantic association module maps the key health features to the health environment knowledge graph to obtain corresponding semantic associations; Specifically, the semantic association module 432 maps the key health features to the health environment knowledge graph to obtain the corresponding semantic associations. The semantic association module 432 receives the key health features from the health feature extraction module 431 and searches for relevant nodes in the health environment knowledge graph. The health environment knowledge graph is a complex network structure that contains the association relationships between various health conditions, environmental factors, pollutants, and health risks. The semantic association module 432 adopts a graph traversal algorithm, starting from the key health features, to explore all the environmental factor nodes directly related to them. For example, for the feature of "asthma patients", the system will identify the highly relevant environmental factors including PM2.5, PM10, pollen, humidity, etc. At the same time, the semantic association module 432 calculates the strength weights of these associations, quantifying the impact degree of different environmental factors on specific health conditions based on medical research literature and clinical statistical data. In addition, the semantic association module 432 also considers multi-feature interactions. When the user has multiple health features simultaneously (such as being both an asthma patient and an elderly person), the system will comprehensively evaluate the superimposed effects of these features. Finally, the semantic association module 432 generates a semantic network that includes the user's health features, relevant environmental factors, and their association strengths. This semantic association method based on the knowledge graph enables the system to transform abstract health features into specific environmental sensitivity indicators, providing a scientific basis for the subsequent generation of control rules.

[0022] The control rule element graph generation module is used to generate a control rule element graph based on the semantic association; Specifically, the control rule element graph generation module 433 is used to form a control rule element graph based on semantic associations. The control rule element graph is a special directed graph structure that describes the mapping relationship from environmental parameters to indicator light states. The control rule element graph generation module 433 receives the semantic association network from the semantic association module 432 and reconstructs it into a control-oriented structure through a graph transformation algorithm. In this process, environmental factor nodes are converted into input variable nodes, association strengths are converted into threshold correction factors, and health risks are mapped to trigger conditions for indicator light states. For example, based on the semantic association of "asthma patients are highly sensitive to PM2.5", the system generates a control rule element of "PM2.5 concentration × 1.5 times sensitivity factor > 30 μg / m³ → yellow warning". The control rule element graph generation module 433 also considers the multi-factor synergistic effect, connects multiple conditions through logical operators such as AND, OR, and NOT to form complex decision paths. At the same time, to improve the practicality of the system, the control rule element graph generation module 433 has a built-in rule simplification algorithm to merge similar conditions and eliminate redundant paths, ensuring that the generated control rule element graph is both complete and concise. The finally formed control rule element graph includes core elements such as environmental parameters, correction factors, threshold conditions, and indicator light states, as well as the logical relationships between them. This structured control rule element graph provides a clear framework for subsequent rule reasoning, enabling personalized air quality evaluation criteria to be accurately expressed and calculated.

[0023] The target indicator light control rule generation module is used to perform inference and judgment on the control rule element graph based on the healthy environment knowledge graph to form the target indicator light control rule.

[0024] Specifically, the target indicator light control rule generation module 434 is used to perform inference and judgment on the control rule element diagram based on the healthy environment knowledge graph to form the target indicator light control rule. The target indicator light control rule is the specific execution instruction that the system finally uses to control the color display of the alarm indicator light. The target indicator light control rule generation module 434 first verifies the completeness and consistency of the control rule element diagram to check whether there are logical conflicts or incomplete coverage. Subsequently, the target indicator light control rule generation module 434 applies the rule inference engine to optimize and expand the rules based on the domain knowledge in the healthy environment knowledge graph. This process utilizes the causal relationships and expert experience contained in the knowledge graph to make the rules more in line with the actual application scenarios. For example, the system may infer based on medical knowledge that "in an environment with a humidity higher than 70%, even if the PM2.5 concentration is relatively low, the risk for asthma patients will increase", thereby adjusting the corresponding warning threshold. The target indicator light control rule generation module 434 also performs formal expression of the rules, converting the control rule element diagram into conditional-action statements with clear syntax, such as "IF PM2.5>35μg / m³ AND user_type=asthma THEN light_color=yellow". Finally, the target indicator light control rule generation module 434 generates a complete set of target indicator light control rules, including detailed regulations on the color, brightness, and flashing mode that the indicator light should display under various environmental conditions. These rules are stored in a structured format for easy retrieval and execution by the control module 3. Through this knowledge-based reasoning process, the system can generate indicator light control rules that are both scientific and personalized, ensuring the accuracy and pertinence of air quality warnings.

[0025] The control module is connected to the air quality detection device and the user data acquisition module, and is used to generate an alarm indicator light control signal according to the target indicator light control rule, user data, and real-time air quality data; Specifically, the control module 3 is connected to the air quality detection device 1 and the user data acquisition module 2, receives real-time air quality data and user health data, and at the same time receives the target indicator light control rule provided by the indicator light control rule generation module 4. The control module 3 performs standardization processing and outlier filtering on the air quality data through internal data processing algorithms, and combines the target indicator light control rule to generate an alarm indicator light control signal. This process realizes the conversion from raw data to visual warning instructions.

[0026] Based on the above embodiments, as an alternative embodiment, please refer to Figure 2 The control module includes: a preprocessing module, a semantic extraction module, and a signal generation module; The preprocessing module is used to preprocess the real-time air quality data to obtain the preprocessed real-time air quality data; Specifically, the preprocessing module 31 is used to preprocess the real-time air quality data to obtain the preprocessed real-time air quality data. The preprocessing module 31 receives the original data stream from the air quality detection device 1, and these data include parameters collected by multiple sensors, such as PM2.5 concentration, PM10 concentration, carbon dioxide content, temperature, humidity, and volatile organic compounds. Due to factors such as sensor accuracy limitations, environmental interference, and data transmission, the original data usually has problems such as noise, outliers, and inconsistencies, and direct use will affect the accuracy of subsequent processing. The preprocessing module 31 first filters the noise of the data and uses digital signal processing techniques such as moving average and median filtering to eliminate high-frequency noise and sudden interference. Subsequently, outlier detection is performed, and the Z-score method based on statistics or the density-based local outlier factor algorithm is used to identify and process data points that deviate from the normal range. Then, data standardization processing is performed to convert parameters with different dimensions (such as PM2.5 in micrograms per cubic meter and humidity in percentage) to the same scale range for subsequent comprehensive analysis. Finally, time alignment and frequency unification are implemented to ensure that the sampling times of all parameters are consistent, and the interpolation method is used to process parameters with different sampling frequencies. After this series of processing, the preprocessing module 31 outputs high-quality preprocessed real-time air quality data, which have the characteristics of low noise, no outliers, consistent dimensions, and uniform time distribution, providing a reliable basis for subsequent semantic analysis. This strict data preprocessing mechanism is the key prerequisite for the system to accurately evaluate the air quality status and avoids misjudgment and incorrect control caused by data quality problems.

[0027] Based on the above embodiments, as an optional embodiment, the preprocessing module includes: a data cleaning sub-module, a data standardization sub-module, and a data fusion sub-module; The data cleaning sub-module is used to perform outlier detection processing on the real-time air quality data to obtain the first real-time air quality data; Specifically, the data cleaning sub-module 311 is used to detect and process outliers in the real-time air quality data to obtain the first real-time air quality data. The original real-time air quality data collected by the air quality detection device 1 is often interfered by various factors, including measurement errors of the sensor itself, instantaneous environmental interference (such as short-term pollution sources like cooking fumes, cigarettes, perfumes, etc.), and noise during data transmission. These factors can cause outliers that do not conform to the actual environmental state in the data. The data cleaning sub-module 311 adopts a multi-level outlier detection strategy. First, it uses statistical methods to conduct a preliminary screening of the data, including parameter boundary detection based on the 3-sigma rule to identify data points that deviate significantly from the historical mean; then it applies the density-based local outlier factor (LOF) detection algorithm to identify outliers that are significantly different from the surrounding data points in the multi-dimensional parameter space; finally, it combines time series analysis to predict the reasonable change range of the data through the ARIMA model and identify suspicious data points with drastic changes in a short period. For the detected outliers, the data cleaning sub-module 311 does not simply delete them, but adopts an intelligent repair strategy, using different processing methods according to the degree of abnormality: for mild abnormalities, local smoothing is used to replace the abnormal points with the weighted average of the surrounding normal data points; for moderate abnormalities, historical pattern replacement is used to fill in the data with data from similar historical scenarios; for severe abnormalities, they are marked as missing values and left for subsequent processing. After this series of processing, the first real-time air quality data output by the data cleaning sub-module 311 is basically free from obvious outlier interference and is closer to the actual air quality state, providing a reliable data basis for subsequent processing.

[0028] The data standardization sub-module is used to standardize the first real-time air quality data to obtain the second real-time air quality data; Specifically, the data standardization sub-module 312 is used to standardize the first real-time air quality data to obtain the second real-time air quality data. Air quality involves various different types of parameters, such as PM2.5 (μg / m³), carbon dioxide (ppm), temperature (°C), humidity (percentage), etc. These parameters use different measurement units and have very different value ranges. For example, PM2.5 is usually 0 - 500 μg / m³, while carbon dioxide can reach 400 - 5000 ppm. This inconsistency in dimension and value range makes it difficult to directly compare or comprehensively process these parameters. The data standardization sub-module 312 performs a unified transformation on the received first real-time air quality data, mainly using two standardization methods: MinMax standardization linearly maps each parameter to the [0,1] interval, retaining the relative distribution characteristics of the parameter; Z-score standardization converts the parameter into a standard normal distribution with a mean of 0 and a standard deviation of 1, highlighting the degree of dispersion of the parameter. For different application scenarios, the data standardization sub-module 312 automatically selects the appropriate standardization method. For example, MinMax standardization is used when calculating the comprehensive air quality index, and Z-score standardization is used in anomaly detection. At the same time, to handle the importance differences of different parameters, the data standardization sub-module 312 also introduces a parameter weight mechanism, assigning different weights according to the degree of impact of different parameters on health. For example, PM2.5 gets a greater weight due to its higher health risk. In addition, the data standardization sub-module 312 also saves the parameters used in the standardization process (such as minimum value, maximum value, mean, standard deviation, etc.) for subsequent inverse transformation to ensure data interpretability. After standardization, different types of air quality parameters are converted to a unified scale space, forming the second real-time air quality data with a consistent structure, which is convenient for subsequent comprehensive analysis and processing.

[0029] The data fusion sub-module is used to perform spatio-temporal fusion processing on the second real-time air quality data to obtain the preprocessed real-time air quality data.

[0030] Specifically, the data fusion sub-module 313 is used to perform spatio-temporal fusion processing on the second real-time air quality data to obtain preprocessed real-time air quality data. Air quality data has obvious temporal and spatial correlations. Relying solely on single-point measurements at a certain moment or a certain location is difficult to comprehensively reflect the environmental state. The data fusion sub-module 313 uses multi-source data fusion technology to comprehensively consider information in the time dimension and the space dimension, and generates a more comprehensive and stable air quality representation. In the time dimension, the data fusion sub-module 313 adopts sliding analysis of time windows, not only considering the data at the current moment, but also combining historical data over a period of time (such as 5 minutes, 15 minutes, 1 hour) to calculate trend features and fluctuation features, such as dynamic indicators like the rising rate, peak frequency, and duration. In the space dimension, if the system deploys multiple sensor nodes, the data fusion sub-module 313 uses spatial interpolation algorithms (such as the inverse distance weighting method or the Kriging method) to construct an indoor air quality distribution map, identify pollution hotspots and diffusion paths. For cases where multi-point measurements are lacking, the data fusion sub-module 313 infers the possible air quality states at different locations based on the indoor airflow dynamics model. In addition, the data fusion sub-module 313 also integrates outdoor air quality data from external APIs to analyze the correlation and interactive effects between indoor and outdoor air quality. Through these spatio-temporal fusion processes, the data fusion sub-module 313 generates preprocessed real-time air quality data containing rich spatio-temporal information, which not only reflects the current state, but also includes multi-dimensional information such as historical trends, spatial distributions, and expected changes, providing a more comprehensive basis for system decision-making.

[0031] The semantic extraction module is used to perform semantic extraction on the preprocessed real-time air quality data to obtain air quality semantic data; Specifically, the semantic extraction module 32 is used to perform semantic extraction on the preprocessed real-time air quality data to obtain air quality semantic data. Semantic extraction is the process of converting numerical sensor data into high-level information with environmental significance. The semantic extraction module 32 first analyzes the independent semantics of each air quality parameter. For example, it interprets the PM2.5 concentration of 35 μg / m³ as "mild pollution" and the carbon dioxide concentration of 1200 ppm as "insufficient ventilation". This conversion is based on air quality standards and a professional knowledge base, transforming abstract numbers into specific descriptions of environmental states. Subsequently, the semantic extraction module 32 conducts correlation analysis among parameters to identify the comprehensive environmental states reflected by parameter combinations. For example, "PM2.5 mild pollution + high humidity + high temperature" may indicate "dust accumulation in a humid and stuffy environment". In addition, the semantic extraction module 32 also performs time pattern analysis to identify the trends and periodic characteristics of data changes over time, such as dynamic characteristics like "continuous increase in PM2.5 concentration" or "daily peak in carbon dioxide concentration". Finally, the semantic extraction module 32 generates structured air quality semantic data, including multi-dimensional semantic information such as classification labels, comprehensive scores, key pollutants, change trends, and health impacts of the current air state. This conversion from numerical to semantic enables the system to "understand" the actual meaning of air quality data rather than mechanically comparing numerical magnitudes, providing a richer decision-making basis for subsequent intelligent control. At the same time, the semantic data form is also more convenient for matching with knowledge-based control rules, improving the accuracy of system response.

[0032] The signal generation module is used to generate the indicator light control signal based on the air quality semantic data and the target indicator light control rule.

[0033] Specifically, the signal generation module 33 is used to generate an indicator light control signal based on the air quality semantic data and the target indicator light control rule. The signal generation module 33 receives the air quality semantic data from the semantic extraction module 32 and the target indicator light control rule from the indicator light control rule generation module 4, and associates the two through a rule matching engine. The target indicator light control rule defines the indicator light states to be displayed under different air quality semantic conditions, such as color, brightness, and blinking mode, etc. The signal generation module 33 first performs rule retrieval, searching for control rules that match the current air quality semantic data in the rule library. For multiple matching rules, the system applies a priority mechanism to resolve conflicts, usually giving priority to stricter warning rules to ensure safety. After determining the finally applied rule, the signal generation module 33 performs signal encoding, converting the abstract control instructions into specific electrical signal parameters, such as RGB color values, PWM duty cycles (to control brightness), and timing parameters (to control blinking), etc. Finally, the signal generation module 33 outputs the encoded indicator light control signal to the alarm indicator light 5 through a standard interface. To improve the user experience, the signal generation module 33 also implements smooth state transition processing. When the indicator light needs to switch from one state to another (such as from green to yellow), a gradual change effect is adopted instead of a sudden change, reducing the visual impact. This rule-based signal generation mechanism ensures the consistency and predictability of the system response, and at the same time, through the combination with personalized control rules, realizes a customized display effect for different user health conditions.

[0034] The alarm indicator light is connected to the control module and is used to display different colors according to the alarm indicator light control signal of the control module; Specifically, the alarm indicator light 5 is connected to the control module 3 and is the most intuitive output interface of the system. The alarm indicator light 5 displays different colors according to the alarm indicator light control signal sent by the control module 3. For example, green indicates excellent air quality, yellow indicates mild pollution, red indicates severe pollution, etc. This color coding method enables users to intuitively understand the current air quality status without the need for professional knowledge.

[0035] The Internet of Things control module is connected to the control module and is used to communicate and connect with multiple environmental control devices through a wireless communication network, and receive the real-time air quality data sent by the control module. When the real-time air quality data is lower than the preset air quality threshold, it sends a device control instruction to the environmental control devices, where the environmental control devices include air purification devices and ventilation facilities; Specifically, the Internet of Things control module 6 is connected to the control module 3 and communicates with multiple environmental control devices through a wireless communication network such as Wi-Fi, ZigBee, or Bluetooth. The Internet of Things control module 6 receives the real-time air quality data sent by the control module 3 and sends a device control instruction to the environmental control device when the data is lower than the preset air quality threshold. The environmental control devices include air purification devices and ventilation facilities, such as air purifiers, fresh air systems, or smart windows. This automatic linkage control mechanism eliminates the need for user manual intervention and improves the efficiency of air quality management.

[0036] Based on the above embodiments, as an alternative embodiment, please refer to Figure 3 , the Internet of Things control module further includes: a device communication control unit, an environmental data analysis unit, and a device control unit; The device communication control unit is used to establish a communication connection with multiple environmental control devices through the wireless communication network and receive the real-time air quality data; Specifically, the device communication control unit 61 is used to establish a communication connection with multiple environmental control devices through the wireless communication network and receive the real-time air quality data. The device communication control unit 61 is built-in with multiple wireless communication protocol stacks, including Wi-Fi, ZigBee, Bluetooth, and NB-IoT, etc., enabling it to be compatible and docked with environmental control devices with different communication standards on the market. When the system starts, the device communication control unit 61 first performs the automatic discovery and identification process of environmental control devices, and confirms the types of controllable devices, communication protocols, and control interfaces around through broadcast detection signals. After the identification is completed, the device communication control unit 61 establishes an encrypted communication channel to ensure the security of control instruction transmission. At the same time, the device communication control unit 61 receives the processed real-time air quality data from the control module 3, including key indicators such as PM2.5, PM10, formaldehyde, and carbon dioxide concentration, and transfers these data to the environmental data analysis unit 62 for further processing. This multi-protocol support and automatic discovery mechanism greatly improves the compatibility and usability of the system, and users can include various brands of air purification devices and ventilation facilities within the control range without complex configuration.

[0037] The environmental data analysis unit is used to analyze and process the real-time air quality data, generate an air quality score, and judge whether the air quality is lower than the preset air quality threshold according to the air quality score. When the air quality is lower than the preset air quality threshold, an air quality warning signal is sent to the device control unit; Specifically, the environmental data analysis unit 62 is used to analyze and process real-time air quality data, generate an air quality score, and determine whether the air quality is lower than a preset air quality threshold based on the air quality score. When the air quality is lower than the preset air quality threshold, an air quality warning signal is sent to the device control unit 63. The environmental data analysis unit 62 adopts a comprehensive scoring algorithm, which weights and calculates various air quality indicators according to their health impact weights to generate an air quality score ranging from 0 to 100, where 100 points represent the best air quality and 0 points represent the worst. For example, the PM2.5 indicator has a greater weight due to its higher health risk. The calculated air quality score is then compared with the preset air quality threshold, which is usually set at 60 points but can be adjusted according to the special needs of users. When the air quality score is lower than the preset air quality threshold, the environmental data analysis unit 62 generates an air quality warning signal, which contains information such as the score value, the type of key pollutants, and the degree of pollution, and sends it to the device control unit 63. The environmental data analysis unit 62 also has a short-term trend analysis function. By analyzing the data change trend in the recent 30 minutes, it issues a warning signal in advance to achieve early intervention in pollution prevention. This comprehensive scoring and trend analysis method avoids the one-sidedness of single-index monitoring and makes environmental regulation more comprehensive and forward-looking.

[0038] The device control unit is used to generate corresponding device control instructions according to the air quality warning signal and a preset control strategy, and transmit the device control instructions to the device communication control unit; Specifically, the device control unit 63 is used to generate corresponding device control instructions according to the air quality warning signal and a preset control strategy, and transmit the device control instructions to the device communication control unit 61. The device control unit 63 has a variety of intelligent control strategies built-in, and automatically selects the optimal regulation plan for different pollutants and pollution levels. The control strategy takes into account various factors such as device type, energy efficiency ratio, noise level, and usage cost, and while achieving the air quality improvement goal, tries to reduce energy consumption and noise interference as much as possible. For example, for mild PM2.5 pollution, the system preferentially starts the air purifier in the low-power mode; for excessive formaldehyde, the ventilation facilities are preferentially turned on to increase the ventilation rate. The device control unit 63 also has a device operation status monitoring function, which receives the working status feedback of each environmental regulation device in real time. When it finds that the device is abnormal or the efficiency decreases, it automatically adjusts the control strategy or issues a maintenance reminder. The generated device control instructions follow a standardized format, including fields such as device ID, operation type, operation parameters, and execution priority, to ensure the accurate execution of the instructions. This policy-based intelligent control method greatly improves the accuracy and efficiency of environmental regulation and avoids the energy waste and equipment wear caused by blind regulation.

[0039] The device communication control unit is further configured to receive the device control instruction and send the device control instruction to the corresponding environmental control device, where the environmental control device includes an air purification device and a ventilation facility.

[0040] Specifically, after receiving the device control instruction transmitted from the device control unit 63, the device communication control unit 61 sends these instructions to the corresponding environmental control devices, including an air purification device and a ventilation facility, through the previously established wireless communication channel. To ensure the reliable transmission of the control instruction, the device communication control unit 61 adopts an instruction confirmation mechanism, requiring the receiving device to return an execution confirmation message. If the confirmation is not received within a predetermined time, the system automatically resends the instruction or attempts an alternative communication method. At the same time, the device communication control unit 61 maintains a device communication status database, recording the connection quality and response time of each device, providing a basis for communication fault diagnosis and optimization. This reliable instruction transmission mechanism ensures the timely and effective implementation of environmental control, greatly improving the stability of the system and the user experience.

[0041] The machine learning prediction module, connected to the control module, is configured to obtain real-time air quality data and real-time weather data, and input the real-time air quality data and the real-time weather data into a preset air quality prediction model to generate air quality prediction data. When the air quality prediction data is lower than the preset air quality threshold, a warning message is generated; Specifically, the machine learning prediction module 7 is connected to the control module 3 and is responsible for the prediction and analysis of air quality. The machine learning prediction module 7 obtains real-time air quality data and real-time weather data obtained through a network interface, including outdoor air quality index, wind speed, air pressure, and precipitation probability, etc. The machine learning prediction module 7 inputs these data into a preset air quality prediction model, which is trained by a deep learning algorithm and can predict the air quality change trend in the next few hours. When the air quality prediction data is lower than the preset air quality threshold, the machine learning prediction module 7 generates a warning message, including possible pollution time, pollution degree, and duration, etc.

[0042] Based on the above embodiments, as an optional embodiment, the machine learning prediction module further includes: a data acquisition unit, a prediction analysis unit, and a warning generation unit; The data acquisition unit is configured to receive in real time the real-time air quality data collected by the air quality monitoring device and obtain the real-time weather data through a network interface; Specifically, the data acquisition unit 71 is used to receive the real-time air quality data collected by the air quality monitoring device in real time and obtain the real-time weather data through the network interface. The data acquisition unit 71 is set with a dual-channel data acquisition mechanism. The internal channel is directly connected to the air quality detection device 1 through the system bus, and the real-time air quality data is collected once a minute, including multiple indicators such as PM2.5, PM10, carbon dioxide, volatile organic compounds, etc. At the same time, the external channel establishes a secure connection with the meteorological service API through the HTTP / HTTPS protocol, and the real-time weather data is updated once an hour, including meteorological factors such as outdoor temperature, humidity, air pressure, wind speed, wind direction, and precipitation probability. To ensure data quality, the data acquisition unit 71 preprocesses the received data, including operations such as outlier detection, missing value filling, and time series alignment. Outlier detection uses the statistical 3-sigma rule to identify and mark data points that deviate significantly from the normal range; missing value filling adopts different strategies according to the data type, such as linear interpolation or nearest neighbor filling; time series alignment unifies the data collected at different frequencies to the same time scale for subsequent model processing. This comprehensive and accurate data acquisition and preprocessing mechanism provides high-quality input for predictive analysis and is the basis for accurate prediction.

[0043] The prediction analysis unit is used to input the real-time air quality data and the real-time weather data into the preset air quality prediction model to generate the air quality prediction data; Specifically, the prediction analysis unit 72 is used to input the real-time air quality data and the real-time weather data into the preset air quality prediction model to generate the air quality prediction data. The prediction analysis unit 72 incorporates a variety of machine learning algorithms, and the core is a deep learning model based on the long short-term memory network (LSTM). This model is specifically optimized for processing time series data and can capture the long-term dependencies and short-term fluctuation characteristics of air quality changes. The preset air quality prediction model consists of three layers: the input layer receives the preprocessed data from the data acquisition unit 71; the hidden layer contains 128 LSTM units, which learn the temporal patterns of the data through the memory gate and forget gate mechanisms; the output layer generates the air quality prediction data for the next 6 hours, with a 30-minute interval, for a total of 12 time points of prediction values. Model training uses a historical data set containing air quality and weather data for the past year, and optimizes the network parameters through the backpropagation algorithm. To improve prediction accuracy, the model also introduces an attention mechanism to dynamically adjust the weights of different input features. For example, under specific weather conditions, the influence weights of wind speed and wind direction are increased. The prediction analysis unit 72 updates the prediction results every 15 minutes and records the prediction errors for continuous optimization of the model performance. This deep learning-based prediction method has higher accuracy and robustness compared to traditional statistical models, can handle complex and variable environmental factors, and provides reliable air quality prediction information for users.

[0044] The warning generation unit is configured to compare the air quality prediction data with the preset air quality threshold, and generate a warning message when the air quality prediction data is lower than the preset threshold.

[0045] Specifically, the warning generation unit 73 is configured to compare the air quality prediction data with the preset air quality threshold, and generate a warning message when the air quality prediction data is lower than the preset threshold. The warning generation unit 73 has multiple levels of warning thresholds built in. Based on the national air quality standards and referring to the recommendations of the World Health Organization, different levels of warning thresholds are set according to different pollutants. For example, for PM2.5, 35 μg / m³ is set as the mild warning threshold, 75 μg / m³ as the moderate warning threshold, and 150 μg / m³ as the severe warning threshold. The warning generation unit 73 compares the air quality prediction data generated by the prediction and analysis unit 72 with these thresholds in real time. When the prediction data shows that the air quality index at a certain future time point will be lower than the corresponding threshold, the warning mechanism is triggered. The warning message includes content such as the warning time, warning level, key pollutants, predicted value, duration, and recommended measures. For example, "Warning: PM2.5 will reach 90 μg / m³ (moderate pollution) at 15:30, is expected to last for 3 hours, and it is recommended to turn on the air purification equipment and reduce outdoor activities". To avoid user fatigue caused by frequent warnings, the warning generation unit 73 adopts an intelligent warning strategy to suppress short-term slight over-standard situations and only issues warnings when the pollution level is relatively high or the duration is relatively long. The generated warning message is transmitted to the control module 3 and the mobile terminal interaction module 8, which are used to trigger automatic environmental regulation and send notifications to users respectively. This predictive warning mechanism enables the system to change from passive response to active prevention, greatly improving the forward-looking and effectiveness of air quality management.

[0046] On the basis of the above implementation, as an optional embodiment, the warning generation unit further includes: a threshold comparator and a warning message generator; The threshold comparator is configured to compare the air quality prediction data with the preset air quality threshold in real time, identify the time point and specific parameters exceeding the threshold, and obtain a comparison result; Specifically, the threshold comparator 731 is used to compare the air quality prediction data with the preset air quality threshold in real time to identify the time points and specific parameters exceeding the threshold, and obtain the comparison result. The threshold comparator 731 has a built-in multi-dimensional threshold matrix, and sets different threshold standards for different air quality parameters and different sensitive populations. These thresholds are determined based on the air quality index (AQI) calculation method and the results of clinical medical research. Taking PM2.5 as an example, the mild pollution threshold for the general population is 35 μg / m³, while the same-level threshold for asthma patients is reduced to 25 μg / m³. The threshold comparator 731 receives the air quality prediction data from the prediction analysis unit 72, and the prediction data includes the predicted values of various air quality parameters at multiple time points within the next 6 hours. The threshold comparator 731 uses a sliding window algorithm to compare these predicted values with the corresponding thresholds point by point. When the predicted value of a certain parameter at a certain time point exceeds the corresponding threshold, the time point, parameter type, predicted value, and degree of exceeding are recorded. To improve the robustness of the comparison, the threshold comparator 731 also performs a persistence analysis. Only when multiple consecutive time points exceed the threshold will it be confirmed as a valid warning point to avoid false alarms caused by short-term fluctuations. After the comparison is completed, the threshold comparator 731 generates a structured comparison result, including detailed information such as all time points exceeding the threshold, duration, specific parameters, and their predicted values. This refined threshold comparison mechanism significantly improves the accuracy and pertinence of the warning, enabling the system to identify truly noteworthy air quality anomalies.

[0047] The warning information generator is used to generate the warning information including the warning time, influence range, and warning level based on the comparison result.

[0048] Specifically, the early warning information generator 732 is used to generate early warning information including the early warning time, the affected range, and the early warning level based on the comparison result. The early warning information generator 732 first sorts the comparison results provided by the threshold comparator 731 according to their priorities, and determines the severity of the event based on factors such as the type of pollutant, the degree of exceeding the threshold, and the duration. Subsequently, the early warning information generator 732 generates corresponding early warning levels according to the severity levels, usually divided into three levels: mild early warning (yellow), moderate early warning (orange), and severe early warning (red). The early warning information generator 732 uses natural language generation technology to convert the structured comparison results into early warning information that is easy for users to understand, including five key elements: the early warning time (the expected time point when the pollution starts), the affected range (the indoor area affected), the early warning level (the classification of the pollution severity), the duration (the expected time period for the pollution to continue), and the recommended measures (the coping suggestions for this type of pollution). For example, a complete early warning information may be: "Severe early warning: It is expected that starting from 2:30 this afternoon, the PM2.5 in the living room area will exceed 150 μg / m³, at the red warning level, and is expected to last for 4 hours. It is recommended to immediately turn on the high-efficiency air purification equipment, close the doors and windows, reduce indoor activities, and the elderly and children should wear masks." To enhance the early warning effect, the early warning information generator 732 also sets different notification methods according to the early warning level. The mild early warning is only displayed on the application interface, the moderate early warning will send a push notification, and the severe early warning will send an alarm through multiple channels (push notification, text message, email, etc.). This hierarchical early warning and personalized notification mechanism ensures that users can receive early warning information that matches the risk level, avoiding information overload or alarm fatigue.

[0049] The mobile terminal interaction module is connected to the control module and is used to establish a data connection with the mobile terminal through a wireless communication network, send the real-time air quality data, the air quality prediction data, and the early warning information to the mobile terminal, and receive the user parameter setting instructions sent by the mobile terminal.

[0050] Specifically, the mobile terminal interaction module 8 is connected to the control module 3 and establishes a data connection with mobile terminals such as the user's smart phone and tablet through a wireless communication network. The mobile terminal interaction module 8 sends the real-time air quality data, the air quality prediction data, and the early warning information to the mobile terminal, enabling the user to understand the indoor air condition at any time. At the same time, the mobile terminal interaction module 8 receives the user parameter setting instructions sent by the mobile terminal, such as adjusting the alarm threshold, updating the health information, or modifying the environmental control strategy, to realize the remote control and personalized configuration of the system.

[0051] Based on the above embodiments, as an alternative embodiment, the mobile terminal interaction module further includes: a communication connection unit, a data sending unit, and an instruction receiving unit; The communication connection unit is used to establish a connection with the control module, maintain data interaction, and transmit the established connection status information to the data sending unit and the instruction receiving unit; Specifically, the communication connection unit 81 is used to establish a connection with the control module 3, maintain data interaction, and transmit the established connection status information to the data sending unit 82 and the instruction receiving unit 83. The communication connection unit 81 adopts a variety of wireless communication technologies, including Wi-Fi direct connection, Bluetooth low energy, and mobile data network, etc., and automatically selects the optimal connection method according to the communication environment. During system initialization, the communication connection unit 81 first starts the device discovery service and broadcasts a connection request signal to the surroundings. When the user mobile terminal runs the supporting application program and responds to the connection request, the two parties perform identity authentication and encryption key exchange. After successful authentication, the communication connection unit 81 establishes an encrypted communication channel with the user mobile terminal and regularly sends heartbeat packets to ensure the connection remains valid. The communication connection unit 81 monitors the connection status in real time, records key indicators such as signal strength, latency, and packet loss rate, and generates connection status information. When it detects a decrease or interruption in the connection quality, the communication connection unit 81 automatically starts the reconnection mechanism and attempts to restore or reconstruct the connection. This multi-protocol support and automatic reconnection mechanism ensure the stable and reliable communication between the system and the user mobile terminal, and can maintain good connection performance even in a complex network environment.

[0052] The data sending unit is used to confirm whether the connection status is normal according to the connection status information. When the connection status is normal, it sends the real-time air quality data, the air quality prediction data, and the warning information to the user mobile terminal; Specifically, the data sending unit 82 is used to confirm whether the connection status is normal according to the connection status information. When the connection status is normal, it sends the real-time air quality data, air quality prediction data, and warning information to the user mobile terminal. The data sending unit 82 first checks the connection status information provided by the communication connection unit 81 to determine whether the current connection status meets the requirements of data transmission. The judgment criteria include: the signal strength is higher than -70 dBm, the delay time is less than 200 milliseconds, the packet loss rate is lower than 5%, etc. After confirming that the connection status is normal, the data sending unit 82 obtains the real-time air quality data from the control module 3, the air quality prediction data and warning information from the machine learning prediction module 7, formats these data, and forms a standardized data packet. To optimize data transmission, the data sending unit 82 adopts a differential transmission strategy, only sending the data that has changed compared with the previous transmission, which greatly reduces the amount of data transmission. The data packet is in JSON format, containing fields such as data type, timestamp, measurement value, and unit, which is convenient for the mobile terminal application to parse and display. For the warning information, the data sending unit 82 sets a higher transmission priority to ensure that the warning information can be delivered to the user mobile terminal in a timely manner. This intelligent data processing and transmission strategy not only improves the communication efficiency, reduces the power consumption of the system and mobile devices, but also ensures that users can obtain accurate air quality information and warning prompts in real time.

[0053] The instruction receiving unit is configured to receive the user parameter setting instruction sent by the user mobile terminal, parse and verify the received instruction to obtain a verified valid instruction, and transfer the verified valid instruction to the control module.

[0054] Specifically, the instruction receiving unit 83 is configured to receive the user parameter setting instruction sent by the user mobile terminal, parse and verify the received instruction to obtain the verified valid instruction, and transmit the verified valid instruction to the control module 3. The instruction receiving unit 83 continuously monitors the instruction requests from the user mobile terminal. When a user parameter setting instruction is received, it immediately performs instruction parsing. The user parameter setting instruction usually includes operations such as threshold adjustment, device control strategy modification, notification method setting, etc. The instruction receiving unit 83 first checks whether the instruction format conforms to the predefined protocol specification, and then verifies whether the parameters in the instruction are within a reasonable range. For example, it checks whether the PM2.5 threshold setting is within the effective range of 0 - 200 μg / m³. After passing the format and parameter verification, the instruction receiving unit 83 further verifies the user's authority to ensure that the user has the right to execute this operation. After all verifications are passed, the instruction receiving unit 83 generates the verified valid instruction and transmits it to the control module 3 for execution through the internal communication interface. At the same time, the instruction receiving unit 83 returns an instruction reception confirmation message to the user mobile terminal to inform the user that the instruction has been accepted by the system. If problems are found during the verification process, the instruction receiving unit 83 generates detailed error messages to explain the reasons for rejection and help the user correct the instruction. This strict instruction verification mechanism effectively prevents incorrect operations and malicious attacks, ensuring the safe and stable operation of the system.

[0055] Based on the above embodiments, as an alternative embodiment, the system further includes: A user feedback module for obtaining the subjective perception feedback of the user on the current air quality status; Specifically, the user feedback module 9 is used to obtain the subjective perception feedback of users on the current air quality status. The evaluation of air quality not only depends on objective measurement data but is also closely related to users' subjective feelings. Different users may have different sensitivities and comfort feelings towards the same air quality parameters. The user feedback module 9 provides a simple and intuitive feedback channel for users through the interface of the mobile terminal interaction module 8, including various forms such as satisfaction ratings, comfort descriptions, and specific symptom reports. When the system displays a certain air quality status, users can provide subjective perception feedback through simple operations (such as sliding the rating bar, selecting an emoticon icon, or filling in a short description). For example, users can express specific feelings such as "the current air feels fresh and comfortable", "slightly stuffy and uncomfortable", or "feeling throat discomfort". To improve the convenience of feedback, the user feedback module 9 also supports voice input and gesture operations, and users can quickly provide feedback through voice commands such as "the air quality is bad" or simple gestures. In addition, the user feedback module 9 also implements a feedback collection strategy that combines active inquiry and passive reception: on the one hand, the system will actively ask users about their feelings after obvious changes in air quality or after maintaining stability for a long time; on the other hand, users can also actively provide feedback at any time. Each feedback message is associated and stored with the real-time air quality data, environmental conditions, and user information at that time to form complete marked data, providing a basis for subsequent personalized learning. This user feedback mechanism enables the system to capture subjective perception information that is difficult to obtain solely by sensors, making up for the limitations of objective measurement.

[0056] The personalized parameter learning module is used to perform personalized adjustment on the threshold parameters in the indicator light control rule by using an enhanced learning algorithm based on the user feedback, the real-time air quality data, and the user health data to obtain a personalized indicator light control rule; Specifically, the personalized parameter learning module 10 is used to perform personalized adjustment on the threshold parameters in the indicator light control rule based on user feedback, real-time air quality data, and user health data, and obtain a personalized indicator light control rule. The personalized parameter learning module 10 realizes the adaptive learning ability of the system. By analyzing the corresponding relationship between user feedback and environmental data, it continuously optimizes the control rule to better match the individual needs of users. The personalized parameter learning module 10 adopts a model-based reinforcement learning algorithm, takes the air quality state as the environmental state, the indicator light color and environmental regulation behavior as actions, and the user satisfaction feedback as the reward signal to construct a complete learning framework. The learning process is divided into three stages: First, user mode recognition is performed. Based on user health data and historical feedback, a user sensitivity model is established to identify the user's sensitivity and preference for different pollutants. Then, parameter fine-tuning is implemented. The threshold parameters in the indicator light control rule are gradually adjusted through the gradient ascent method. For example, the yellow warning threshold of PM2.5 is adjusted from 35 μg / m³ to 30 μg / m³ to make the system response more in line with user expectations. Finally, rule evaluation is performed. The effectiveness of the adjusted rule is verified and evaluated through simulation tests and historical data to ensure that the new rule can maintain rationality in various situations. To prevent overfitting to individual feedback, the personalized parameter learning module 10 also introduces regularization constraints and expert knowledge boundaries to ensure that the adjusted parameters are still within the medical safety range. As the learning progresses, the system gradually forms stable personalized indicator light control rules, which fully consider the user's specific health status and subjective feelings, achieving true personalized intelligence.

[0057] The control module is also used to generate an alarm indicator light control signal according to the personalized indicator light control rule.

[0058] Specifically, in addition to generating an alarm indicator control signal according to the target indicator control rule, the control module 3 also generates a more precise alarm indicator control signal according to the personalized indicator control rule. When the personalized parameter learning module 10 generates the personalized indicator control rule, these rules are transmitted to the control module 3 and integrated with the original rules. The control module 3 adopts a priority strategy, preferentially applying the personalized rules and only falling back to the general rules in specific situations (such as when the personalized rules have not been formed or health and safety requirements). In specific operations, the control module 3 receives real-time air quality data, passes it to the preprocessing module 31 for processing, and then the semantic extraction module 32 extracts the air quality semantic data. Based on these semantic data, the signal generation module 33 applies the personalized indicator control rule to generate the final alarm indicator control signal. For example, for the same PM2.5 concentration of 40 μg / m³, the system may display green (indicating good) for general users and yellow (indicating warning) for users detected with a history of respiratory allergies and relevant discomfort feedback. The generation of such personalized control signals enables the alarm indicator 5 to provide visual feedback that better meets the actual needs of users, enhancing the pertinence and practicality of the early warning.

[0059] On the other hand, in a specific embodiment, this embodiment further includes a security monitoring function, an intelligent access control linkage function, a voice interaction system, an emergency assistance function, an environmental perception expansion function, a positioning and navigation function, and a health monitoring interconnection function; The security monitoring function realizes comprehensive home security monitoring by integrating a micro high-definition camera and an infrared sensor on the top and side of the lamp. The micro high-definition camera uses a CMOS sensor with a resolution of 1080P and a super wide-angle lens to achieve a 120° field of view coverage; the infrared sensor uses passive infrared technology (PIR) and can detect the movement of heat sources in low-light environments. The built-in motion detection algorithm of the system uses the image sequence collected by the camera and the heat source data of the infrared sensor to identify abnormal movements by combining the background difference method and the optical flow method. At the same time, the system integrates face recognition technology based on deep learning, compares the captured facial features with the pre-stored facial features of family members, and the accuracy rate reaches 99.3%. When the system detects an unrecognized face or an abnormal movement pattern, it automatically starts the recording function and pushes alarm information to the user through the mobile terminal interaction module. This integrated security function avoids the trouble of installing separate monitoring devices, and the widespread distribution of the lamps ensures the comprehensiveness of security coverage.

[0060] The intelligent access control linkage function connects the lights with the home access control system to create a seamless space transition experience. The lights are equipped with low-power Bluetooth (BLE 5.0) and Near Field Communication (NFC) modules, with operating frequencies of 2.4 GHz and 13.56 MHz respectively, enabling the system to detect the smartphones or access cards carried by users. When family members approach the home area, the Received Signal Strength Indicator (RSSI) algorithm calculates the approximate distance and direction of the users, activating the lighting system in advance. Meanwhile, based on the user identity and ambient light conditions, the system intelligently adjusts the light brightness (adjustable from 200 to 800 lumens) and color temperature (adjustable from 2700K to 6500K), providing a comfortable lighting environment the moment the user enters the room. Especially when returning home at night, this intelligent linkage avoids the inconvenience of fumbling for the light switch in the dark, improving the quality of life and safety.

[0061] The voice interaction system realizes natural voice control and interaction functions by integrating a ring microphone array and a full-frequency speaker inside the lights. The ring microphone array consists of 6 omnidirectional microphones, adopting sound source localization and beamforming technologies, enabling the system to accurately capture voice commands within a range of 5 meters. Even in an environment with background noise, the recognition rate can reach over 95%. The full-frequency speaker has a power output of 3 watts and a frequency response range of 80 Hz - 20 kHz, ensuring that the voice responses are clear and distinguishable. The system is built-in with voice recognition and natural language processing engines, supporting a hybrid processing mode of local and cloud, and can understand complex semantic commands such as "Adjust the living room lights to a brightness and color temperature suitable for reading". Meanwhile, the voice interaction system also supports the home broadcast function, allowing users to send voice messages to specified rooms or all lights in the house through a mobile application, constructing a convenient home intercom network. This voice interaction method is especially suitable for the elderly and children with limited mobility, significantly enhancing the usability of smart homes.

[0062] The emergency assistance function provides a quick help channel for users in critical situations. The lights are integrated with a physical SOS button and a voice-activated trigger mechanism. Users can activate the emergency mode by pressing the hidden red emergency button on the edge of the light or shouting out preset emergency words (such as "Emergency help", "Call for rescue"). After the system starts the emergency procedure, the control module immediately controls the alarm indicator to flash red (frequency: 2 Hz), and at the same time sends an emergency notification containing the user's location information and on-site images to the preset emergency contacts through the mobile terminal interaction module. For severe emergencies, the system can also directly call the emergency or police number and achieve remote communication through the voice interaction system. This function is especially applicable to the elderly living alone and those with limited mobility, providing them with safety protection and enabling them to obtain help quickly in case of emergencies.

[0063] The environmental perception extension function realizes all-round monitoring of the indoor environment by adding multiple types of sensors. The lamp integrates an acoustic sensor, a carbon monoxide sensor, and a methane sensor, which respectively monitor the indoor noise level (measurement range 30 - 120 dB), carbon monoxide concentration (detection range 0 - 1000 ppm, accuracy ±5 ppm), and methane concentration (detection range 0 - 1000 ppm, accuracy ±10 ppm). These sensors adopt a miniaturized design and share the data acquisition circuit with the air quality detection device, optimizing the system structure. When abnormal values are detected, such as when the carbon monoxide concentration exceeds 50 ppm or the methane concentration reaches 10% of the lower explosion limit, the system immediately triggers a multi-level warning mechanism: the alarm indicator shows a red warning, and at the same time, warning messages are broadcast through the voice interaction system, and the ventilation facilities are started through the Internet of Things control module. This multi-dimensional environmental monitoring goes beyond the traditional air quality monitoring scope and provides more comprehensive protection for home safety, especially for colorless, odorless but extremely dangerous threats such as gas leakage and carbon monoxide poisoning.

[0064] The positioning and navigation function provides intelligent indoor navigation services for users with the help of Bluetooth Beacon technology. The lamp is built-in with a low-power Bluetooth transmitter that broadcasts signal packets containing location codes at intervals of 100 ms. When the user requests navigation services through the mobile application, the system plans the optimal path according to the user's destination and provides dim light guidance through the lighting fixtures along the way. The dim light guidance adopts a gradient lighting mode, and the lighting fixtures along the travel route are lit in turn with soft light (brightness is 15% of normal lighting, warm light with a color temperature of 3000K), forming an intuitive light path guidance. This positioning and navigation function is especially suitable for night scenes, providing safe indoor movement guidance for the elderly and children, avoiding the risks of falling and collision caused by the dark environment, and being more energy-efficient and comfortable than full-bright lighting at the same time.

[0065] The health monitoring and interconnection function connects the lamp with the user's wearable device through the Internet of Things technology, establishing an intelligent lighting system based on biological rhythms. The lamp establishes a connection with the smart watch or health bracelet worn by the user through low-power Bluetooth and receives biological data such as heart rate, activity level, and sleep quality. The biological rhythm analysis algorithm of the system combines these data and the user's work and rest rules to generate a personalized lighting plan. For example, when it is detected that the user is in a highly fatigued state, the system automatically adjusts the light to a high-color-temperature cold light above 6000K to improve concentration and alertness; when the user is about to rest, the light is gradually adjusted to a warm light of 2700K to promote melatonin secretion and assist in falling asleep. In this way, the health monitoring and interconnection function elevates lighting from a simple visual function to a physiological regulation tool, helping users maintain a healthy biological rhythm and improve sleep quality and daytime energy levels.

[0066] The above are only exemplary embodiments of the present disclosure, and the scope of the present disclosure cannot be limited thereby. That is, any equivalent changes and modifications made in accordance with the teachings of the present disclosure still fall within the scope covered by the present disclosure. Those skilled in the art will readily conceive of other embodiments of the present disclosure after considering the specification and the disclosure of the practical truth.

[0067] This application is intended to cover any variations, uses, or adaptations of the present disclosure, which follow the general principles of the present disclosure and include well-known common general knowledge or conventional technical means in the technical field not recorded in the present disclosure. The specification and the embodiments are only regarded as exemplary, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. A lighting fixture system with air quality detection function, characterized in that: The system comprises: Air quality detection device, user data acquisition module, control module, indicator light control rule generation module, alarm indicator light, Internet of Things control module, machine learning prediction module and mobile terminal interaction module; The air quality detection device is arranged inside the lamp and is used to continuously collect real-time air quality data reflecting the air quality; The user data acquisition module is used to acquire the user health data of the user; The indicator light control rule generation module is used to determine the target indicator light control rule based on the user health data, the Internet of Things control module and the preset health environment knowledge graph; The control module is connected to the air quality detection device and the user data acquisition module, and is used to generate an alarm indicator light control signal according to the target indicator light control rule, user data and real-time air quality data; The alarm indicator light is connected to the control module and is used to display different colors according to the alarm indicator light control signal of the control module; The Internet of Things control module is connected to the control module, and is used to communicate with multiple environmental control devices through a wireless communication network, and receive real-time air quality data sent by the control module. When the real-time air quality data is lower than a preset air quality threshold, a device control instruction is sent to the environmental control device, wherein the environmental control device includes air purification equipment and ventilation facilities; The machine learning prediction module is connected to the control module and is used to obtain real-time air quality data and real-time weather data, and input the real-time air quality data and the real-time weather data into a preset air quality prediction model to generate air quality prediction data, and generate warning information when the air quality prediction data is lower than the preset air quality threshold; The mobile terminal interaction module is connected to the control module and is used to establish a data connection with the mobile terminal through a wireless communication network, send the real-time air quality data, the air quality prediction data and the warning information to the mobile terminal, and receive user parameter setting instructions sent by the mobile terminal.

2. The system according to claim 1, characterized in that The indicator light control rule generation module also includes: a knowledge data acquisition module, a knowledge graph construction module, and a control rule generation module; The knowledge data acquisition module is used to acquire health environment knowledge data; The knowledge graph construction module is used to construct a health environment knowledge graph based on the health environment knowledge data; The control rule generation module is used to determine the target indicator light control rule based on the user health data and the health environment knowledge graph.

3. The system according to claim 1, characterized in that The Internet of Things control module also includes: a device communication control unit, an environmental data analysis unit and a device control unit; The device communication control unit is used to establish a communication connection with the plurality of environmental control devices through the wireless communication network and receive the real-time air quality data; The environmental data analysis unit is used to analyze and process the real-time air quality data to generate an air quality score, and determine whether the air quality is lower than the preset air quality threshold according to the air quality score, and send an air quality warning signal to the device control unit when the air quality is lower than the preset air quality threshold; The device control unit is used to generate corresponding device control instructions according to the air quality warning signal and the preset control strategy, and transmit the device control instructions to the device communication control unit; The device communication control unit is also used to receive the device control instruction and send the device control instruction to the corresponding environmental control device, wherein the environmental control device includes air purification equipment and ventilation facilities.

4. The system according to claim 1, characterized in that The mobile terminal interaction module further includes: a communication connection unit, a data sending unit and an instruction receiving unit; The communication connection unit is used to establish a connection with the control module and maintain data interaction, and transmit the established connection status information to the data sending unit and the instruction receiving unit; The data sending unit is used to confirm whether the connection status is normal according to the connection status information, and when the connection status is normal, send the real-time air quality data, the air quality prediction data and the warning information to the user mobile terminal; The instruction receiving unit is used to receive the user parameter setting instruction sent by the user mobile terminal, parse and verify the received instruction, obtain the verified valid instruction, and pass the verified valid instruction to the control module.

5. The system according to claim 1, characterized in that The machine learning prediction module further includes: a data acquisition unit, a prediction analysis unit and an early warning generation unit; The data acquisition unit is used to receive the real-time air quality data collected by the air quality monitoring device in real time, and obtain the real-time weather data through the network interface; The prediction and analysis unit is used to input the real-time air quality data and the real-time weather data into the preset air quality prediction model to generate the air quality prediction data; The warning generation unit is used to compare the air quality prediction data with the preset air quality threshold, and generate warning information when the air quality prediction data is lower than the preset threshold.

6. The system according to claim 5, characterized in that The warning generation unit further includes: a threshold value comparator and a warning information generator; The threshold comparator is used to compare the air quality prediction data with the preset air quality threshold in real time to identify the time point and specific parameters exceeding the threshold, and obtain a comparison result; The warning information generator is used to generate the warning information including the warning time, impact scope and warning level based on the comparison result.

7. The system according to claim 2, characterized in that The control rule generation module includes: a health feature extraction module, a semantic association module, a control rule element map generation module and a target indicator light control rule generation module; The health feature extraction module is used to perform semantic extraction on the user health data using natural language processing technology to obtain key health features; The semantic association module maps the key health features to the health environment knowledge graph to obtain corresponding semantic associations; The control rule element graph generation module is used to form a control rule element graph based on the semantic association; The target indicator light control rule generation module is used to infer and judge the control rule element graph based on the health environment knowledge graph to form the target indicator light control rule.

8. The system according to claim 1, characterized in that The control module includes: a preprocessing module, a semantic extraction module and a signal generation module; The preprocessing module is used to preprocess the real-time air quality data to obtain preprocessed real-time air quality data; The semantic extraction module is used to perform semantic extraction on the preprocessed real-time air quality data to obtain air quality semantic data; The signal generating module is used to generate the indicator light control signal based on the air quality semantic data and the target indicator light control rule.

9. The system according to claim 8, characterized in that The preprocessing module includes: a data cleaning submodule, a data standardization submodule and a data fusion submodule; The data cleaning submodule is used to perform abnormal value detection processing on the real-time air quality data to obtain first real-time air quality data; The data standardization submodule is used to perform standardization processing on the first real-time air quality data to obtain second real-time air quality data; The data fusion submodule is used to perform spatiotemporal fusion processing on the second real-time air quality data to obtain the pre-processed real-time air quality data.

10. The system according to claim 1, characterized in that The system further comprises: User feedback module, used to obtain users' subjective perception feedback on the current air quality status; A personalized parameter learning module, used to perform personalized adjustment on the threshold parameters in the indicator light control rule by using a reinforcement learning algorithm based on the user feedback, the real-time air quality data and the user health data, so as to obtain a personalized indicator light control rule; The control module is also used to generate an alarm indicator light control signal according to the personalized indicator light control rule.

Images