Classroom air quality management and epidemic disease early warning robot system and dynamic regulation and control method

Abstract

The invention relates to the technical field of robots, in particular to a classroom air quality management and epidemic disease early warning robot system and a dynamic regulation and control method. Comprising a mobile monitoring module, an intelligent analysis module, an early warning decision module and a dynamic regulation and control module. The mobile monitoring module is used for collecting air quality data in a classroom and feeding back the air quality data to the intelligent analysis module; the mobile monitoring module carries a micro gas chromatography module, a laser particle counter, a non-dispersive infrared CO2 sensor, an infrared thermal imager, a temperature and humidity sensor and a gas sensitive sensor array; the intelligent analysis module carries out fusion processing on the air quality data collected by the mobile monitoring module and constructs an aerosol diffusion model, and the intelligent analysis module and the early warning decision module realize data interaction through a shared memory; the disease transmission risk can be reduced through intelligent regulation and control.

Description

Technical Field

[0001] This invention relates to the field of robotics, and in particular to a classroom air quality management and epidemic early warning robot system and its dynamic control method. Background Technology

[0002] In the construction of smart campuses, classroom air quality and epidemic prevention are crucial aspects of ensuring a safe teaching environment. Current technologies for classroom environmental monitoring mostly employ fixed-point, single-parameter monitoring models, capable of detecting only basic indicators such as CO2 concentration, temperature, and humidity. They cannot simultaneously capture key parameters closely related to air quality, such as VOCs composition and aerosol particle distribution within the classroom. Particularly noteworthy is the lack of coupled analysis capabilities for air pollutants and pathogen transmission. These systems cannot assess the risk of pathogen aggregation due to insufficient ventilation through abnormal formaldehyde concentration changes, nor can they evaluate the probability of epidemic transmission by combining spatial distribution characteristics of body temperature data. Consequently, early warning mechanisms rely solely on single threshold judgments, resulting in low accuracy and significant delays. Furthermore, control measures are mostly independent start-stop operations for fresh air systems or disinfection equipment, making it difficult to achieve multi-device coordinated operation and dynamically optimize control strategies based on real-time risks.

[0003] Based on the above problems, there is an urgent need for a technical solution that can integrate multi-dimensional environmental parameters and personnel status information to achieve accurate correlation analysis between air quality and the risk of epidemic transmission, and reduce the risk of disease transmission through intelligent regulation. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and proposes a classroom air quality management and epidemic early warning robot system, comprising: The system includes a mobile monitoring module, an intelligent analysis module, an early warning and decision-making module, and a dynamic control module. The mobile monitoring module is used to collect air quality data in the classroom and feed it back to the intelligent analysis module; The mobile monitoring module is equipped with a miniature gas chromatograph module, a laser particle counter, a non-dispersive infrared CO2 sensor, an infrared thermal imager, a temperature and humidity sensor, and a gas sensor array. The intelligent analysis module fuses and processes the air quality data collected by the mobile monitoring module and constructs an aerosol diffusion model. The intelligent analysis module and the early warning decision module achieve data interaction through shared memory. The warning signal of the warning decision module is transmitted to the dynamic control module through an optocoupler isolation circuit. The warning decision module predicts the probability of an influenza outbreak by analyzing the geospatial correlation between abnormal formaldehyde concentration and infrared temperature measurement data and combining it with a causal reasoning algorithm. The dynamic control module uses a model predictive control algorithm and works in conjunction with the fresh air system and ultraviolet disinfection lamp for dynamic control.

[0005] Preferably, the mobile monitoring module's micro gas chromatography module uses a micro-column with a length of 1.2 meters and an inner diameter of 0.25 millimeters. The stationary phase of the micro-column is 5% phenyl-methyl polysiloxane, and the carrier gas is high-purity nitrogen. The mobile monitoring module is used to simultaneously detect the particle number concentration in six particle size channels: 0.3 μm, 0.5 μm, 1 μm, 2.5 μm, 5 μm, and 10 μm.

[0006] More preferably, the intelligent analysis module adopts a multi-threaded parallel processing architecture, which includes a main thread and slave threads. The main thread receives real-time data from each sensor. The slave threads include a first slave thread, a second slave thread, and a third slave thread. The first slave thread performs baseline correction and peak identification on the VOCs detection data. The second slave thread performs statistical analysis on the aerosol particle data. The third slave thread performs filtering processing on the CO2 concentration, temperature, and humidity environmental parameters.

[0007] More preferably, the gas sensor array comprises 12 sensing elements, each of which is made of metal oxide nanomaterials. Six of the sensing elements are based on SnO2 and doped with Pt, Pd, and Au metal elements, respectively. The other six sensing elements are based on SnO2 and ZnO and doped with In, Ga, and Sb metal elements, respectively. The operating temperature of the sensing elements is controlled within the range of 200-400℃ by a temperature control circuit, and the specific identification of VOCs components is achieved by the response differences of different elements.

[0008] More preferably, in the causal reasoning algorithm of the early warning decision module, the formula for calculating the probability of an influenza outbreak is: ; in, Indicates the probability of an influenza outbreak; Indicates formaldehyde concentration; Indicates CO2 concentration; Indicates aerosol particle number concentration; This indicates the abnormal body temperature rate, which is the ratio of the number of people with a body temperature exceeding 37.3℃ to the total number of people in the classroom. This indicates the density of people, which is the ratio of the total number of people in the classroom to the classroom area. The ventilation efficiency coefficient is calculated by the ratio of the actual ventilation volume to the theoretically required ventilation volume. , , , , , These are the weight coefficients of each influencing factor, determined through multiple linear regression analysis of historical data, with values ​​ranging from 0 to 1. This is a threshold parameter, with a value range of 5-10.

[0009] More preferably, the objective function calculation formula for the linkage strategy between the fresh air system and the ultraviolet disinfection lamp in the model predictive control algorithm of the dynamic control module is as follows: ; in, The objective function value; To control the cycle; Let be the CO2 concentration at time t; The target concentration of CO2; The power percentage of the ultraviolet disinfection lamp at time t, with a value ranging from 0 to 1; Let t be the fresh air volume at time t; Let be the aerosol particle number concentration at time t; , , , , These are weighting coefficients, with values ​​ranging from 0 to 10; The fresh air volume at time t-1.

[0010] More preferably, in the aerosol diffusion model constructed by the intelligent analysis module, the formula for calculating the aerosol particle diffusion coefficient is: ; in, The aerosol particle diffusion coefficient; The reference diffusion coefficient; CO2 concentration; The air velocity in the classroom; The temperature inside the classroom; Classroom height; The activity intensity coefficient of personnel in the classroom is calculated by combining personnel movement speed and activity frequency, and its value ranges from 0 to 1. , , The correction factor has a value range of 0.01-0.1.

[0011] A dynamic control method, applied to the classroom air quality management and epidemic early warning robot system described in any one of the above, includes the following steps: S1: The mobile monitoring module moves and samples in the classroom according to the preset path. The sampling point interval is 1.5 meters. It completes a full classroom coverage sampling every 3 minutes. The collected data includes the concentration of 16 VOCs components, the number and concentration of aerosol particles of different sizes, CO2 concentration, body temperature of people at each location, and temperature and humidity data. S2: The intelligent analysis module receives the sampled data, preprocesses the data including removing outliers and smoothing filtering, and then combines the laser particle counter data to construct an aerosol diffusion model. At the same time, it analyzes the correlation between formaldehyde concentration and CO2 concentration to determine the ventilation status. S3: The early warning decision module inputs the preprocessed data into the causal reasoning algorithm to calculate the probability of an influenza outbreak. When the probability is greater than or equal to 0.6, a Level 1 early warning is issued; when the probability is between 0.3 and 0.6, a Level 2 early warning is issued; and when the probability is less than 0.3, no early warning is issued. S4: The dynamic control module adjusts the fresh air volume and the power of the ultraviolet disinfection lamp of the fresh air system according to the warning level and the calculated objective function value. When there is a level 1 warning, the fresh air volume is adjusted to 80%-100% of the maximum rated value and the power of the ultraviolet disinfection lamp is adjusted to 70%-100% of the maximum rated value. When there is a level 2 warning, the fresh air volume is adjusted to 50%-80% of the maximum rated value and the power of the ultraviolet disinfection lamp is adjusted to 30%-70% of the maximum rated value. When there is no warning, the fresh air volume is maintained at 30%-50% of the maximum rated value and the ultraviolet disinfection lamp is turned off. S5: Repeat S1 to S4 every 20 minutes to update monitoring data and control strategies in real time.

[0012] Further preferably, the data preprocessing described in S2 specifically includes: for VOCs concentration data, wavelet transform is used for noise reduction, a db4 wavelet basis is selected, the number of decomposition layers is 3, and soft thresholding is applied to high-frequency coefficients, with the threshold being 0.02 times the maximum high-frequency coefficient; for aerosol particle number concentration data, moving average filtering is used; for CO2 concentration data, Kalman filtering is used, with the noise variance of the state equation set to 0.05 and the noise variance of the observation equation set to 0.1; for human body temperature data, outliers exceeding the range of 35℃-42℃ are removed, and then the average body temperature of each sampling point is calculated.

[0013] More preferably, in the process of control, the dynamic control module described in S4 turns off the ultraviolet disinfection lamp when it detects a class state, and turns on the ultraviolet disinfection lamp when it detects a break or post-class state and there is no one in the classroom. The fresh air system controls the wind speed below 0.3 m / s during class hours; during breaks, the wind speed can be increased to 0.5 m / s-0.8 m / s. The system monitors the equipment's operating status in real time. When the fresh air system or ultraviolet disinfection lamp malfunctions, it immediately issues an equipment malfunction alarm and switches to the backup control scheme. The backup control scheme includes: if the fresh air system malfunctions, opening classroom doors and windows for natural ventilation; if the ultraviolet disinfection lamp malfunctions, increasing the fresh air volume to 90%-100% of the maximum rated value.

[0014] Technical effects: This invention achieves simultaneous acquisition of multiple parameters through a mobile monitoring module, overcoming the limitations of single-parameter monitoring in existing technologies; it constructs a coupled model through an intelligent analysis module, breaking through the challenge of separating and analyzing air pollutants and pathogen transmission; it improves prediction accuracy through causal reasoning-based early warning, overcoming the shortcomings of single-threshold early warning; and it achieves coordinated linkage of equipment through a dynamic control module, solving the inefficiency of independent start-up and shutdown, effectively reducing the risk of respiratory disease transmission in classrooms, and realizing precise management of air quality and epidemic prevention and control. Attached Figure Description

[0015] Figure 1 This is a block diagram of the classroom air quality management and epidemic early warning robot system of the present invention; Figure 2 This is a flowchart of the dynamic control method of the present invention. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0017] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0018] Traditional technical solutions have the following technical problems: Traditional classroom air monitoring mostly uses fixed-point single-parameter sensors, which can only detect limited parameters such as CO2 or temperature and humidity. They cannot simultaneously monitor multi-dimensional data such as VOCs, aerosol particles and human body temperature, and lack the ability to couple and analyze formaldehyde concentration and pathogen transmission. Early warning mechanisms are mostly based on single indicator threshold judgments, which have low accuracy. Control measures are mostly single device start-stop, which cannot achieve coordinated linkage between fresh air and disinfection equipment, making it difficult to effectively reduce the risk of epidemic transmission.

[0019] Based on this, please refer to Figure 1 This embodiment provides a classroom air quality management and epidemic early warning robot system, including: The system includes a mobile monitoring module, an intelligent analysis module, an early warning and decision-making module, and a dynamic control module. The mobile monitoring module is used to collect air quality data in the classroom and feed it back to the intelligent analysis module; The mobile monitoring module is equipped with a miniature gas chromatograph module, a laser particle counter, a non-dispersive infrared CO2 sensor, an infrared thermal imager, a temperature and humidity sensor, and a gas sensor array. The intelligent analysis module fuses and processes the air quality data collected by the mobile monitoring module and constructs an aerosol diffusion model. The intelligent analysis module and the early warning decision module achieve data interaction through shared memory. The warning signal of the warning decision module is transmitted to the dynamic control module through an optocoupler isolation circuit. The warning decision module predicts the probability of an influenza outbreak by analyzing the geospatial correlation between abnormal formaldehyde concentration and infrared temperature measurement data and combining it with a causal reasoning algorithm. The dynamic control module uses a model predictive control algorithm and works in conjunction with the fresh air system and ultraviolet disinfection lamp for dynamic control.

[0020] This solution addresses the shortcomings of traditional monitoring methods, such as limited monitoring dimensions, insufficient early warning accuracy, and poor coordination in regulation, through multi-module collaboration. It achieves closed-loop management from multi-parameter monitoring to intelligent early warning and dynamic regulation. Notably, the mobile monitoring module employs a mobile design, overcoming the spatial limitations of fixed-point monitoring and covering all areas of the classroom; the intelligent analysis module's aerosol diffusion model combines multi-source data to reflect aerosol propagation patterns; the early warning decision-making module's causal reasoning algorithm enables correlation analysis between formaldehyde concentration and body temperature data; and the dynamic regulation module's model predictive control algorithm ensures precise linkage between fresh air systems and disinfection equipment.

[0021] The technical effects achieved by the above embodiments include: multi-parameter mobile monitoring can comprehensively capture classroom air quality and personnel status information, avoiding the one-sidedness of single-parameter monitoring; the fusion processing capability of the intelligent analysis module provides data support for the study of aerosol diffusion patterns; the causal reasoning algorithm improves the accuracy of epidemic early warning, making the early warning more in line with actual transmission risks; and the dynamic control and coordination strategy can effectively reduce the risk of disease transmission while ensuring air quality, providing an intelligent solution for classroom environment management.

[0022] Traditional technical solutions have the following technical problems: When traditional micro gas chromatography modules detect VOCs, the separation efficiency is low, the detection time is long, and the detection limit is high, making it impossible to accurately capture changes in low concentrations of VOCs; the particle size detection range of laser particle counters is limited, making it difficult to fully reflect the distribution of aerosol particles of different sizes, affecting the accurate judgment of pathogenic microorganism transmission carriers, and thus restricting the targeted nature of subsequent early warning and control.

[0023] Based on this, the mobile monitoring module's micro gas chromatography module uses a micro-column with a length of 1.2 meters and an inner diameter of 0.25 millimeters. The stationary phase of the micro-column is 5% phenyl-methyl polysiloxane, and the carrier gas is high-purity nitrogen. The mobile monitoring module is used to simultaneously detect the particle number concentration in six particle size channels: 0.3 μm, 0.5 μm, 1 μm, 2.5 μm, 5 μm, and 10 μm.

[0024] This solution addresses the problems of low separation efficiency, long detection time, and incomplete particle size coverage in traditional equipment by optimizing the structural parameters of the gas chromatography module and the detection settings of the laser particle counter, thereby improving the accuracy and comprehensiveness of VOCs and aerosol particle detection.

[0025] It is worth mentioning that the aspect ratio design of the micro-column, combined with specific stationary phase materials, enhances the separation capability of different VOCs components. The polarity of the 5% phenyl-methyl polysiloxane stationary phase is adapted to the polarity differences of 16 VOCs, achieving efficient separation. High-purity nitrogen carrier gas and precise flow rate control reduce interference factors and lower the detection limit. The six particle size channels of the laser particle counter cover the main range of inhalable particulate matter, especially the detection of small particles with a diameter of 0.3 μm, which can capture more potential pathogenic microorganism carriers.

[0026] The technical effects achieved by the above embodiments include: completing the detection of 16 VOCs within 8 minutes, improving detection efficiency and enabling timely reflection of changes in VOCs concentration in classrooms; a detection limit of 5 ppb, which can accurately capture abnormal fluctuations in low-concentration VOCs, providing early signals for problems such as insufficient ventilation; and particle detection through six particle size channels, which comprehensively reflects the distribution characteristics of aerosol particles, helps to deeply analyze the correlation between particles of different sizes and the spread of pathogenic microorganisms, provides a basis for developing targeted control strategies, and makes subsequent early warning and control more scientific.

[0027] Traditional technical solutions have the following technical problems: When processing multi-source data, traditional intelligent analysis modules use a single-threaded processing method, which results in low data processing efficiency, data backlog and delays, and a lack of targeted processing for different types of data, leading to low data quality. This makes it impossible to provide timely and reliable data support for subsequent early warning decisions and dynamic control, affecting the response speed and accuracy of the entire system.

[0028] Based on this, the intelligent analysis module adopts a multi-threaded parallel processing architecture, which includes a main thread and slave threads. The main thread receives real-time data from each sensor. The slave threads include a first slave thread, a second slave thread, and a third slave thread. The first slave thread performs baseline correction and peak identification on VOCs detection data, the second slave thread performs statistical analysis on aerosol particle data, and the third slave thread performs filtering processing on CO2 concentration, temperature, and humidity environmental parameters.

[0029] This solution addresses the problems of low efficiency, high data latency, and insufficient processing targeting in traditional single-threaded processing by designing a multi-threaded parallel processing architecture, thereby improving the efficiency and quality of data processing.

[0030] It is worth mentioning that: the 10Hz receiving frequency ensures real-time data acquisition, providing a guarantee for the main thread to quickly acquire raw data; the three slave threads have clear division of labor, adopting dedicated processing methods for the characteristics of different types of data; baseline correction and peak identification of VOCs data can eliminate the influence of baseline drift and improve the accuracy of component identification; statistical analysis of aerosol particle data can extract key distribution features; and filtering of CO2 and temperature and humidity data can reduce noise interference; the shared memory data interaction method enables rapid data transfer between threads, ensuring the continuity of data processing; and the 500ms latency control ensures that the processed data can be used for subsequent analysis in a timely manner.

[0031] The technical effects achieved by the above embodiments include: multi-threaded parallel processing significantly improves data processing efficiency and avoids congestion problems during single-threaded processing; targeted processing methods improve the quality of various types of data, making VOCs component identification more accurate, aerosol particle distribution characteristics clearer, and environmental parameters more stable; timely data interaction and low-latency processing provide fast and reliable data support for the early warning decision module and dynamic control module, making the entire system respond more quickly and decisions more in line with actual environmental changes.

[0032] Traditional technical solutions have the following technical problems: the sensitive element material of traditional gas sensor arrays is single and has poor specificity, which limits the ability to identify different VOCs and makes it difficult to accurately distinguish multiple VOCs components; the operating temperature of the sensitive element is fixed and cannot be adjusted according to the detection requirements, resulting in unstable detection accuracy under different temperature conditions, affecting the accurate capture of changes in VOCs concentration, and thus affecting the comprehensive assessment of classroom air quality.

[0033] Based on this, the gas sensor array includes 12 sensing elements, each of which is made of metal oxide nanomaterials. Six of the sensing elements are based on SnO2 and doped with Pt, Pd, and Au metal elements, respectively. The other six sensing elements are based on SnO2 and ZnO and doped with In, Ga, and Sb metal elements, respectively. The operating temperature of the sensing elements is controlled within the range of 200-400℃ by a temperature control circuit, and the specific identification of VOCs components is achieved by the response differences of different elements.

[0034] This solution addresses the issues of poor specificity and unstable detection accuracy in traditional sensor arrays by using diverse sensitive element materials and adjustable operating temperatures, thereby improving the ability to identify VOCs components.

[0035] It is worth mentioning that the 12 sensitive elements employ different material combinations. The SnO2 substrate is doped with Pt, Pd, and Au noble metals, utilizing the catalytic effect of noble metals to enhance the sensitivity to specific VOCs. The ZnO substrate is doped with In, Ga, and Sb metal elements, expanding the recognition range of VOCs through differences in elemental properties. The operating temperature range of 200-400℃ can be adapted to the optimal response temperature of different VOCs. For example, some VOCs are more sensitive at high temperatures, while others are more characteristic at medium and low temperatures. The ±1℃ adjustment accuracy ensures the accuracy of temperature control and reduces the impact of temperature fluctuations on the detection results. The response differences of different elements form a unique "fingerprint," which, through comparative analysis, enables the accurate identification of 16 VOCs.

[0036] The technical effects achieved by the above embodiments include: the diverse material combinations significantly improve the specificity of the sensor array, greatly enhancing the ability to distinguish different VOCs and accurately identifying 16 VOC components; the adjustable operating temperature allows the sensor to maintain high accuracy under different detection scenarios, ensuring accurate capture of changes in VOC concentration; the specific identification capability provides more detailed VOC component information for classroom air quality assessment, helps to determine the cause of abnormal air quality, provides a more accurate basis for ventilation control and other measures, and improves the overall system's ability to control air quality.

[0037] Traditional technical solutions have the following technical problems: When predicting the probability of influenza outbreaks, traditional epidemic early warning algorithms consider fewer influencing factors, often relying on only one or a few parameters, ignoring the impact of key factors such as population density and ventilation efficiency on disease transmission, resulting in a large deviation between the early warning results and the actual risk; moreover, the algorithm does not construct a reasonable correlation between various factors, making it difficult to accurately reflect the actual transmission pattern and affecting the reliability and guidance of the early warning.

[0038] Based on this, the formula for calculating the probability of an influenza outbreak in the causal reasoning algorithm of the early warning decision module is as follows: ; in, Indicates the probability of an influenza outbreak; Indicates formaldehyde concentration; Indicates CO2 concentration; Indicates aerosol particle number concentration; This indicates the abnormal body temperature rate, which is the ratio of the number of people with a body temperature exceeding 37.3℃ to the total number of people in the classroom. This indicates the density of people, which is the ratio of the total number of people in the classroom to the classroom area. The ventilation efficiency coefficient is calculated by the ratio of the actual ventilation volume to the theoretically required ventilation volume. , , , , , These are the weight coefficients of each influencing factor, determined through multiple linear regression analysis of historical data, with values ​​ranging from 0 to 1. This is a threshold parameter, with a value range of 5-10.

[0039] The formula adopts the form of the Sigmoid function, which maps the results of linear combination of multiple factors to the probability interval of 0-1. This not only conforms to the statistical characteristics of the probability of an epidemic outbreak, but also solves the problem that the traditional threshold judgment method cannot quantify the risk level, reflecting the creative idea of ​​multi-dimensional coupling analysis.

[0040] The probability of an influenza outbreak (0-1) is directly reflected in the risk level through the function output, providing a quantitative basis for early warning classification and replacing the traditional binary judgment mode of yes / no.

[0041] and : These represent formaldehyde concentration and CO2 concentration, respectively, weighted by... , These indicators, along with ventilation status, show that abnormally high formaldehyde concentrations indicate insufficient ventilation, while CO2 concentrations reflect population density. Together, they form a basic indicator of pathogen aggregation risk, addressing the shortcomings of traditional early warning systems that neglect the link between air pollutants and transmission.

[0042] and : Aerosol particle number concentration and abnormal body temperature rate, wherein the abnormal body temperature rate represents the percentage of people with a body temperature exceeding 37.3℃, by , After weighting, it directly correlates the number of pathogen vectors with potential sources of infection, filling the gap in traditional models that lack direct indicators of biological risk.

[0043] and Personnel density and ventilation efficiency coefficient, wherein the ventilation efficiency coefficient is the actual ventilation volume / theoretical requirement, through , By incorporating spatial transmission conditions into the analysis, it is found that higher population density leads to greater contact risk, while lower ventilation efficiency results in weaker pathogen clearance capacity. The coupling of these two factors enhances the model's adaptability to transmission scenarios.

[0044] - The weighting coefficients are obtained through multiple linear regression of historical data, dynamically reflecting the strength of the influence of different factors on transmission. For example, the weighting coefficients may increase during peak influenza seasons. Weights address the limitations of fixed parameters in traditional algorithms.

[0045] The threshold parameter can be adjusted to suit different classroom environments by changing the inflection point of the function. For example, it can be lowered in a preschool classroom. To improve the sensitivity of early warning and enhance the universality of the model.

[0046] This formula breaks through the limitations of traditional single-indicator early warning systems, constructing a four-element coupled risk assessment framework encompassing environmental pollutants, pathogen carriers, personnel status, and spatial conditions. For the first time, it incorporates ventilation efficiency coefficients and personnel density into probability calculations, achieving a leap from phenomenon monitoring to risk prediction. The nonlinear mapping characteristics of the Sigmoid function allow minute risk changes to be reflected through probability gradients, avoiding misjudgments due to abrupt changes near the threshold. The data-driven determination method of weighting coefficients ensures that the model can dynamically optimize with factors such as region and season, providing a scientific quantitative tool for early warning of epidemics and significantly improving the accuracy and foresight of early warnings.

[0047] This approach addresses the issues of incomplete consideration of factors and unreasonable correlations in traditional algorithms by introducing multi-dimensional influencing factors and constructing a reasonable formula model, thereby improving the accuracy of influenza outbreak probability prediction.

[0048] It is worth noting that the formula incorporates six factors that comprehensively cover key parameters related to the environment and personnel. Formaldehyde and CO2 concentrations reflect ventilation conditions, aerosol particle number concentration is an important indicator of pathogen transmission, abnormal body temperature rate is directly related to the status of potential patients, personnel density reflects the risk of transmission and contact, and the ventilation efficiency coefficient reflects the actual effectiveness of ventilation measures. The weighting coefficients are determined through multiple linear regression analysis, ensuring that the influence of each factor aligns with actual data patterns. The sigmoid function maps the linear combination results to a probability range of 0-1, consistent with the characteristics of probability prediction. Threshold parameters... It can be adjusted according to the characteristics of different classroom environments, which enhances the adaptability of the algorithm.

[0049] The technical effects achieved by the above embodiments include: comprehensive consideration of multiple factors makes the prediction of influenza outbreak probability more comprehensive, avoiding the one-sidedness of judgment based on a single factor; reasonable formula model and weight coefficient settings accurately reflect the correlation between each factor and disease transmission, improving the reliability of early warning results; the probabilistic output format makes the early warning results more intuitive, facilitating the formulation of different levels of response strategies based on the probability magnitude; the adaptive adjustment capability of the algorithm can meet the environmental characteristics of different classrooms, making the early warning more targeted and practical, and providing scientific decision support for epidemic prevention and control.

[0050] Traditional technical solutions have the following technical problems: When coordinating the fresh air system and the ultraviolet disinfection lamp, traditional dynamic control strategies often use single parameter threshold control, without comprehensively considering the balance between energy consumption costs, equipment wear and tear and air quality improvement effects. Furthermore, they do not introduce the influence weight of the time dimension on the control intensity, resulting in over-control or control lag. At the same time, they ignore the real-time feedback of aerosol concentration and early warning probability on the control strategy, making it difficult to achieve precise and low-cost joint control.

[0051] Based on this, the objective function calculation formula for the linkage strategy between the fresh air system and the ultraviolet disinfection lamp in the model predictive control algorithm of the dynamic control module is as follows: ; in, The objective function value; To control the cycle; Let be the CO2 concentration at time t; The target concentration of CO2; The power percentage of the ultraviolet disinfection lamp at time t, with a value ranging from 0 to 1; Let t be the fresh air volume at time t; Let be the aerosol particle number concentration at time t; , , , , These are weighting coefficients, with values ​​ranging from 0 to 10; The fresh air volume at time t-1.

[0052] The formula as a whole adopts an integral form to control the period. Cumulative target value within Minimization as the optimization objective reflects the idea of ​​dynamic control over time.

[0053] and : represents the real-time CO2 concentration and the target concentration at time t, respectively, and the squared term of the difference between the two. Through weighting coefficients Strengthen the control of core air quality indicators to ensure that CO2 concentration remains stable within a reasonable range, and address the problem of insufficient attention to key parameters in traditional regulation.

[0054] The power percentage of ultraviolet disinfection lamps, multiplied by the weight. Afterwards, the energy consumption cost of disinfection is quantified to avoid resource waste caused by indiscriminate high-intensity disinfection.

[0055] Time variable t and fresh air volume The product term, through By assigning a higher cost weight to long-term high-airflow operation, the ventilation effect and energy consumption are balanced, thus addressing the shortcomings of traditional strategies that ignore the impact of operating time on costs.

[0056] Disinfection power supplementary item ( ) and aerosol concentration The product of, by Adjusting the weights, when the disinfection power is insufficient, strengthens the control demand by increasing the aerosol concentration to ensure the disinfection priority in high-pollution scenarios.

[0057] : The absolute value of the change in fresh air volume between adjacent time points, after Weighted scaling suppresses frequent start-stop of equipment, extends its service life, and solves the problem of excessive equipment wear and tear in traditional control methods.

[0058] The formula, through multi-parameter coupled design, incorporates dimensions such as air quality, energy consumption, equipment wear and tear, and pollution levels into a unified optimization framework, overcoming the limitations of traditional single-threshold control. Weighting coefficients The strategy can be dynamically adjusted according to scenarios such as class / break time, achieving flexible adaptation. For example, the class period can be extended. To reduce fluctuations in fresh air volume and avoid interfering with teaching, an integral form ensures the continuity of the control strategy over time, enabling the system to respond quickly to environmental changes while maintaining an optimized state over the long term.

[0059] This solution addresses the problems of traditional control strategies, such as strong one-sidedness, insufficient cost control, and slow response, by constructing a multi-parameter coupled objective function, thereby achieving coordinated and optimized control of fresh air and disinfection equipment.

[0060] It is worth mentioning that the objective function incorporates the time variable t and the fresh air volume. The product term reflects the energy cost weighting for long-term high-airflow operation; aerosol concentration Supplementary item for disinfection power The product of these factors reinforces the disinfection priority in high-pollution scenarios; the absolute value of the change in fresh air volume... It is used to suppress the losses caused by frequent start-ups and shutdowns of equipment; the five weight coefficients can be dynamically adjusted according to the classroom usage scenario to achieve flexible adaptation of the strategy.

[0061] The technical effects achieved by the above embodiments include: the multi-parameter coupled objective function enables the control strategy to ensure that the CO2 concentration remains stable within the target range while dynamically allocating disinfection resources according to the degree of aerosol pollution; the time-dimensional cost weight design avoids indiscriminate high-energy-consumption operation and achieves a balance between energy saving and pollution control; the equipment operation smoothness constraint extends the service life of the fresh air system and disinfection lamps; and the scenario-based adjustment capability of the weight coefficients allows the control strategy to adapt to the actual needs of different teaching periods, providing a refined control solution for classroom environment management.

[0062] Traditional technical solutions have the following technical problems: Traditional aerosol diffusion models are mostly based on the ideal fluid assumption, which does not consider the coupling effect of CO2 concentration and air flow in the classroom, and ignores the influence of temperature gradient and human activities on the diffusion coefficient. This results in a large deviation between the model calculation results and the actual diffusion law, which cannot provide an accurate basis for the analysis of pathogenic microorganism transmission paths and affects the spatial targeting of early warning and control.

[0063] Based on this, the formula for calculating the aerosol particle diffusion coefficient in the aerosol diffusion model constructed by the intelligent analysis module is as follows: ;

[0064] in, The aerosol particle diffusion coefficient; The reference diffusion coefficient; CO2 concentration; The air velocity in the classroom; The temperature inside the classroom; Classroom height; The activity intensity coefficient of personnel in the classroom is calculated by combining personnel movement speed and activity frequency, and its value ranges from 0 to 1. , , The correction factor has a value range of 0.01-0.1.

[0065] The formula is based on the baseline diffusion coefficient and quantifies the impact of environmental and human factors on diffusion through three correction terms, breaking through the assumption limitations of traditional ideal models.

[0066] The baseline diffusion coefficient represents the aerosol diffusion capacity under standard conditions and provides a baseline value for subsequent corrections.

[0067] First amendment The coupling of CO2 concentration and air velocity, with exponents of 0.4 and 0.6 respectively reflecting the inhibitory effect of increased CO2 concentration on diffusion (higher concentration, greater diffusion resistance) and the promoting effect of increased flow velocity on diffusion (faster flow velocity, more significant diffusion), are shown. Correction coefficients are also provided. Calibrate the model in real-world environments to ensure it is suitable for different ventilation conditions.

[0068] Second Amendment The correlation between temperature and classroom height is as follows: an index of 0.3 reflects the promoting effect of increased temperature on molecular motion, while -0.2 reflects the dilution effect of increased spatial height on diffusion. It is used to balance the influence of vertical spatial gradients and solve the problem of traditional models ignoring temperature stratification.

[0069] Third Amendment Introducing the activity intensity coefficient of personnel An index of 0.2 quantifies the promoting effect of human disturbance on diffusion; the more frequent the activity, the stronger the airflow disturbance, and the faster the diffusion. It adapts to the personnel density characteristics of different classrooms.

[0070] The formula, through non-integer exponents and multi-factor coupling, overcomes the limitations of traditional linear models and better reflects the nonlinear characteristics of turbulent diffusion in classrooms. For example, the synergistic correction term for CO2 and flow velocity enables an innovative approach to indirectly infer the aerosol diffusion state through easily monitored CO2 concentration; the introduction of human activity factors allows the model to accurately capture the diffusion acceleration effect during peak periods between classes. The calibration mechanism of the correction coefficients ensures that the model can be adapted to different classroom building structures, such as differences in floor height and usage scenarios, such as small / large class teaching. It provides a quantitative tool for analyzing the transmission paths of pathogenic microorganisms, improving the spatial targeting of early warning and control.

[0071] This scheme improves the accuracy of aerosol diffusion law simulation by introducing a diffusion coefficient model based on multi-dimensional environmental and human factors. It solves the problems of traditional models ignoring actual influencing factors and having low calculation accuracy.

[0072] It is worth mentioning that the coupling term between CO2 concentration and air velocity... This reflects the direct impact of ventilation conditions on aerosol diffusion; the increased diffusion resistance with rising CO2 concentration is reflected in the exponential term; correction terms for temperature and classroom height are also included. The influence of buoyancy flow caused by the vertical spatial temperature gradient was considered; the personnel activity intensity coefficient was also taken into account. The introduction of this factor quantifies the promoting effect of human movement-induced airflow disturbance on diffusion; three correction coefficients , , The model can be calibrated in real-world environments to adapt to different classroom building structures and usage scenarios.

[0073] The technical effects achieved by the above embodiments include: the multi-factor coupled diffusion coefficient calculation enables the model to truly reflect the spatial propagation law of aerosols in the classroom, avoiding theoretical deviations of ideal models; the correlation design between CO2 and ventilation parameters provides a quantitative basis for indirectly inferring the aerosol diffusion state through CO2 monitoring; the inclusion of human activity factors enables the model to maintain high accuracy under different activity intensities such as breaks and classes; and the calibration mechanism of correction coefficients enhances the model's adaptability to different classroom environments, providing accurate spatial distribution data support for the analysis of epidemic transmission paths and disinfection of key areas.

[0074] Traditional technical solutions have the following technical problems: Traditional classroom environment control methods mostly adopt fixed-period sampling and single-mode control. The sampling interval is too long, resulting in insufficient data timeliness. The warning level classification is vague, the control strategy lacks pertinence, and the restrictions of teaching activities on equipment operation are not considered. This can easily lead to conflicts between control and teaching. At the same time, there is a lack of emergency response mechanism, and the continuity of environmental control cannot be guaranteed when equipment is abnormal.

[0075] Based on this, please refer to Figure 2 This embodiment provides a dynamic control method applied to the classroom air quality management and epidemic early warning robot system as described in any one of claims 1-7, characterized by comprising the following steps: S1: The mobile monitoring module moves and samples in the classroom according to the preset path. The sampling point interval is 1.5 meters. It completes a full classroom coverage sampling every 3 minutes. The collected data includes the concentration of 16 VOCs components, the number and concentration of aerosol particles of different sizes, CO2 concentration, body temperature of people at each location, and temperature and humidity data. S2: The intelligent analysis module receives the sampled data, preprocesses the data including removing outliers and smoothing filtering, and then combines the laser particle counter data to construct an aerosol diffusion model. At the same time, it analyzes the correlation between formaldehyde concentration and CO2 concentration to determine the ventilation status. S3: The early warning decision module inputs the preprocessed data into the causal reasoning algorithm to calculate the probability of an influenza outbreak. When the probability is greater than or equal to 0.6, a Level 1 early warning is issued; when the probability is between 0.3 and 0.6, a Level 2 early warning is issued; and when the probability is less than 0.3, no early warning is issued. S4: The dynamic control module adjusts the fresh air volume and the power of the ultraviolet disinfection lamp of the fresh air system according to the warning level and the calculated objective function value. When there is a level 1 warning, the fresh air volume is adjusted to 80%-100% of the maximum rated value and the power of the ultraviolet disinfection lamp is adjusted to 70%-100% of the maximum rated value. When there is a level 2 warning, the fresh air volume is adjusted to 50%-80% of the maximum rated value and the power of the ultraviolet disinfection lamp is adjusted to 30%-70% of the maximum rated value. When there is no warning, the fresh air volume is maintained at 30%-50% of the maximum rated value and the ultraviolet disinfection lamp is turned off. S5: Repeat S1 to S4 every 20 minutes to update monitoring data and control strategies in real time.

[0076] It is worth mentioning that the 1.5-meter sampling point interval and the 3-minute full classroom coverage design ensure the spatial integrity and temporal freshness of the data. Compared with the traditional sampling cycle of more than 10 minutes, it can quickly capture concentration fluctuations in a short period of time. The quantitative threshold of the three-level warning makes the warning level division clear and avoids the control deviation caused by vague descriptions. The range of equipment parameters corresponding to different warning levels provides flexible operation space for control and can be fine-tuned according to the real-time environment. The 20-minute strategy update cycle balances the control response speed and equipment operation stability, avoiding the energy waste caused by frequent adjustments and keeping up with environmental changes in a timely manner.

[0077] The technical effects achieved by the above embodiments include: high-frequency mobile sampling enables the system to monitor environmental differences in various areas of the classroom in real time, avoiding the local bias of fixed-point sampling; the hierarchical early warning mechanism ensures that the control strategy is precisely matched with the risk level, strengthening disinfection and ventilation during a level one early warning, appropriately intervening during a level two early warning, and operating at energy efficiency when there is no early warning, thus achieving efficient use of resources; the step-by-step process design ensures a closed-loop connection between monitoring, analysis, early warning, and control, improving the reliability of system operation; and the regular update mechanism enables the strategy to dynamically adapt to personnel flow and environmental changes in the classroom, ensuring the stability of long-term control effects and providing a friendly environment for teaching activities.

[0078] Traditional technical solutions have the following technical problems: Traditional data preprocessing methods use a uniform processing mode for data from different types of sensors without considering the differences in data characteristics, resulting in excessive noise reduction or loss of features for complex components such as VOCs, large statistical deviations in aerosol particle data, and rough handling of outliers in body temperature data, which affects the accuracy of subsequent analysis. Furthermore, they lack dedicated filtering parameters for classroom environments, resulting in poor adaptability of the processing effect.

[0079] Based on this, the data preprocessing described in S2 specifically includes: for VOCs concentration data, wavelet transform is used for noise reduction, with the db4 wavelet basis selected and a decomposition layer of 3 layers. Soft thresholding is applied to high-frequency coefficients, with the threshold being 0.02 times the maximum high-frequency coefficient. For aerosol particle number concentration data, moving average filtering is used. For CO2 concentration data, Kalman filtering is used, with the noise variance of the state equation set to 0.05 and the noise variance of the observation equation set to 0.1. For human body temperature data, outliers exceeding the 35℃-42℃ range are removed, and then the average body temperature at each sampling point is calculated.

[0080] This solution addresses the shortcomings of traditional processing methods, such as insufficient targeting and low data quality, by designing specific preprocessing methods for different data types, thereby improving the usability and analytical accuracy of the raw data.

[0081] It is worth mentioning that: the VOCs data adopted a 3-level decomposition using the db4 wavelet basis, which is adapted to the distribution characteristics of its high-frequency noise and low-frequency features. The soft threshold was set to 0.02 times the maximum high-frequency coefficient, which effectively suppressed instrument noise while preserving characteristic peaks. The 5-point moving average window for aerosol particle data balanced real-time performance with smoothing effect, avoiding interference from single-point fluctuations on the statistical distribution. The Kalman filter parameters for CO2 data, with a state noise variance of 0.05 and an observation noise variance of 0.1, were designed for its slowly varying characteristics, reducing jumps caused by instantaneous interference. The body temperature data was filtered within the range of 35℃-42℃, which not only excluded extreme values ​​caused by sensor failures but also preserved possible pathological high temperatures, and the average body temperature calculation enhanced spatial representativeness.

[0082] The technical effects achieved by the above embodiments include: the dedicated preprocessing method effectively suppresses noise in various types of data, making the characteristic peaks of VOCs clearer, the distribution of aerosol particles more stable, the trend of CO2 concentration change smoother, and the body temperature data more reliable; the multi-scale decomposition capability of wavelet transform preserves the weak signals of low-concentration VOCs, providing a data basis for early ventilation anomaly judgment; the combined application of moving average and Kalman filtering takes into account the rapid changes of aerosol particles and the slow changes of CO2, improving the time synchronization of multi-source data; the range screening and mean calculation of body temperature data reduces the impact of individual differences and measurement errors on the assessment of group status, providing high-quality basic data support for early warning decision-making.

[0083] Traditional technical solutions have the following technical problems: traditional control strategies do not consider the constraints of teaching activities on equipment operation; ultraviolet disinfection lamps may be turned on during class, causing personnel exposure risks; excessively high air velocity in the fresh air system interferes with teaching activities; and there is a lack of emergency plans for equipment failure. Once the fresh air or disinfection equipment malfunctions, the entire environmental control system will be paralyzed, and the continuous safety of the classroom environment cannot be guaranteed.

[0084] Based on this, the dynamic control module described in S4 will turn off the ultraviolet disinfection lamp when it detects a class period and turn on the ultraviolet disinfection lamp when it detects a break or post-class period and that no one is in the classroom. The fresh air system controls the wind speed below 0.3 m / s during class hours; during breaks, the wind speed can be increased to 0.5 m / s-0.8 m / s. The system monitors the equipment's operating status in real time. When the fresh air system or ultraviolet disinfection lamp malfunctions, it immediately issues an equipment malfunction alarm and switches to the backup control scheme. The backup control scheme includes: if the fresh air system malfunctions, opening classroom doors and windows for natural ventilation; if the ultraviolet disinfection lamp malfunctions, increasing the fresh air volume to 90%-100% of the maximum rated value.

[0085] This solution addresses the conflicts between traditional control and teaching methods, as well as the lack of support for equipment malfunctions, by introducing a teaching activity adaptation mechanism and a fault emergency strategy, thereby improving the system's practicality and reliability.

[0086] It's worth mentioning that: the UV disinfection lamps' time-limited operation is linked to personnel detection. By recognizing class status, if the personnel density remains consistently high, it automatically avoids operational risks, ensuring that disinfection operations are only performed during safe periods; the fresh air system's wind speed is controlled in stages: ≤0.3m / s during class and 0.5-0.8m / s between classes, balancing air renewal efficiency and the comfort of the teaching environment; a dual-path backup plan for equipment failure is implemented, with alternative solutions designed based on the functional characteristics of the fresh air and disinfection equipment. Natural ventilation compensates for insufficient mechanical fresh air, while high-volume replacement compensates for insufficient disinfection capacity; and a fault alarm mechanism ensures timely response from maintenance personnel, shortening equipment recovery time.

[0087] The technical effects achieved by the above embodiments include: the teaching activity adaptation mechanism synchronizes the control strategy with the teaching rhythm, avoids interference with classroom order caused by equipment operation, and ensures the safety of teachers and students; the wind speed graded control provides a quiet and comfortable environment for teaching activities while meeting air quality requirements; the fault emergency plan ensures the continuity of environmental management when equipment malfunctions, and reduces the risk accumulation caused by emergencies; the combination of alarm mechanism and backup plan improves the system's fault tolerance and operation and maintenance efficiency, makes classroom environment management more resilient, and provides a stable environmental guarantee for normalized teaching activities.

[0088] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A classroom air quality management and epidemic early warning robot system, characterized in that, include: The system includes a mobile monitoring module, an intelligent analysis module, an early warning and decision-making module, and a dynamic control module. The mobile monitoring module is used to collect air quality data in the classroom and feed it back to the intelligent analysis module; The mobile monitoring module is equipped with a miniature gas chromatograph module, a laser particle counter, a non-dispersive infrared CO2 sensor, an infrared thermal imager, a temperature and humidity sensor, and a gas sensor array. The intelligent analysis module fuses and processes the air quality data collected by the mobile monitoring module and constructs an aerosol diffusion model. The intelligent analysis module and the early warning decision module achieve data interaction through shared memory. The warning signal of the warning decision module is transmitted to the dynamic control module through an optocoupler isolation circuit. The warning decision module predicts the probability of an influenza outbreak by analyzing the geospatial correlation between abnormal formaldehyde concentration and infrared temperature measurement data and combining it with a causal reasoning algorithm. The dynamic control module uses a model predictive control algorithm and works in conjunction with the fresh air system and ultraviolet disinfection lamp for dynamic control.

2. The classroom air quality management and epidemic early warning robot system according to claim 1, characterized in that, The mobile monitoring module's micro gas chromatography module employs a micro-column with a length of 1.2 meters and an inner diameter of 0.25 millimeters. The stationary phase of the micro-column is 5% phenyl-methyl polysiloxane, and the carrier gas is high-purity nitrogen. The mobile monitoring module is used to simultaneously detect the particle number concentration in six particle size channels: 0.3 μm, 0.5 μm, 1 μm, 2.5 μm, 5 μm, and 10 μm.

3. The classroom air quality management and epidemic early warning robot system according to claim 1, characterized in that, The intelligent analysis module adopts a multi-threaded parallel processing architecture, which includes a main thread and slave threads. The main thread receives real-time data from each sensor. The slave threads include a first slave thread, a second slave thread, and a third slave thread. The first slave thread performs baseline correction and peak identification on VOCs detection data. The second slave thread performs statistical analysis on aerosol particle data. The third slave thread performs filtering on CO2 concentration, temperature, and humidity environmental parameters.

4. The classroom air quality management and epidemic early warning robot system according to claim 1, characterized in that, The gas-sensitive sensor array comprises 12 sensing elements, each made of metal oxide nanomaterials. Six of the sensing elements are based on SnO2 and doped with Pt, Pd, and Au metal elements, respectively. The other six sensing elements are based on SnO2 and ZnO and doped with In, Ga, and Sb metal elements, respectively. The operating temperature of the sensing elements is controlled within the range of 200-400℃ by a temperature control circuit, and the specific identification of VOCs components is achieved by the response differences of different elements.

5. The classroom air quality management and epidemic early warning robot system according to claim 1, characterized in that, In the causal reasoning algorithm of the early warning decision module, the formula for calculating the probability of an influenza outbreak is: ； in, Indicates the probability of an influenza outbreak; Indicates formaldehyde concentration; Indicates CO2 concentration; Indicates aerosol particle number concentration; This indicates the abnormal body temperature rate, which is the ratio of the number of people with a body temperature exceeding 37.3℃ to the total number of people in the classroom. This indicates the density of people, which is the ratio of the total number of people in the classroom to the classroom area. The ventilation efficiency coefficient is calculated by the ratio of the actual ventilation volume to the theoretically required ventilation volume. , , , , , These are the weight coefficients of each influencing factor, determined through multiple linear regression analysis of historical data, with values ​​ranging from 0 to 1. This is a threshold parameter, with a value range of 5-10.

6. The classroom air quality management and epidemic early warning robot system according to claim 1, characterized in that, The objective function calculation formula for the linkage strategy between the fresh air system and the ultraviolet disinfection lamp in the model predictive control algorithm of the dynamic control module is as follows: ； in, The objective function value; To control the cycle; Let be the CO2 concentration at time t; The target concentration of CO2; The power percentage of the ultraviolet disinfection lamp at time t, with a value ranging from 0 to 1; Let t be the fresh air volume at time t; Let be the aerosol particle number concentration at time t; , , , , These are weighting coefficients, with values ​​ranging from 0 to 10; The fresh air volume at time t-1.

7. The classroom air quality management and epidemic early warning robot system according to claim 1, characterized in that, In the aerosol diffusion model constructed by the intelligent analysis module, the formula for calculating the aerosol particle diffusion coefficient is as follows: ； in, The aerosol particle diffusion coefficient; The reference diffusion coefficient; CO2 concentration; The air velocity in the classroom; The temperature inside the classroom; Classroom height; The activity intensity coefficient of personnel in the classroom is calculated by combining personnel movement speed and activity frequency, and its value ranges from 0 to 1. , , The correction factor has a value range of 0.01-0.

1.

8. A dynamic control method, applied to the classroom air quality management and epidemic early warning robot system as described in any one of claims 1-7, characterized in that, Includes the following steps: S1: The mobile monitoring module moves and samples in the classroom according to the preset path. The sampling point interval is 1.5 meters. It completes a full classroom coverage sampling every 3 minutes. The collected data includes the concentration of 16 VOCs components, the number and concentration of aerosol particles of different sizes, CO2 concentration, body temperature of people at each location, and temperature and humidity data. S2: The intelligent analysis module receives the sampled data and preprocesses the data. The preprocessing includes removing outliers and smoothing filtering. Then, it combines the laser particle counter data to construct an aerosol diffusion model and analyzes the correlation between formaldehyde concentration and CO2 concentration to determine the ventilation status. S3: The early warning decision module inputs the preprocessed data into the causal reasoning algorithm to calculate the probability of an influenza outbreak. When the probability is greater than or equal to 0.6, a Level 1 early warning is issued; when the probability is between 0.3 and 0.6, a Level 2 early warning is issued; and when the probability is less than 0.3, no early warning is issued. S4: The dynamic control module adjusts the fresh air volume and the power of the ultraviolet disinfection lamp of the fresh air system according to the warning level and the calculated objective function value. When there is a level 1 warning, the fresh air volume is adjusted to 80%-100% of the maximum rated value and the power of the ultraviolet disinfection lamp is adjusted to 70%-100% of the maximum rated value. When there is a level 2 warning, the fresh air volume is adjusted to 50%-80% of the maximum rated value and the power of the ultraviolet disinfection lamp is adjusted to 30%-70% of the maximum rated value. When there is no warning, the fresh air volume is maintained at 30%-50% of the maximum rated value and the ultraviolet disinfection lamp is turned off. S5: Repeat S1 to S4 every 20 minutes to update monitoring data and control strategies in real time.

9. The dynamic control method according to claim 8, characterized in that, The data preprocessing described in S2 specifically includes: for VOCs concentration data, wavelet transform is used for noise reduction, with a db4 wavelet basis and a decomposition level of 3. Soft thresholding is applied to high-frequency coefficients, with a threshold value of 0.02 times the maximum high-frequency coefficient. For aerosol particle number concentration data, moving average filtering is used. For CO2 concentration data, Kalman filtering is used, with the noise variance of the state equation set to 0.05 and the noise variance of the observation equation set to 0.

1. For human body temperature data, outliers exceeding the 35℃-42℃ range are removed, and then the average body temperature at each sampling point is calculated.

10. The dynamic control method according to claim 8, characterized in that, In the dynamic control module described in S4, when a class is detected, the ultraviolet disinfection lamp is turned off; when a break or post-class period is detected and the classroom is empty, the ultraviolet disinfection lamp is turned on. The airflow speed of the fresh air system is controlled below 0.3 m / s during class hours; During breaks between classes, the wind speed can be increased to 0.5 m / s-0.8 m / s; The system monitors the equipment's operating status in real time. When the fresh air system or ultraviolet disinfection lamp malfunctions, it immediately issues an equipment malfunction alarm and switches to the backup control scheme. The backup control scheme includes: if the fresh air system malfunctions, opening classroom doors and windows for natural ventilation; if the ultraviolet disinfection lamp malfunctions, increasing the fresh air volume to 90%-100% of the maximum rated value.

Images