Device and method for monitoring and killing air pathogens in real time
Through the combination of double-layer generation adversarial network and fuzzy logic control technology, the hysteresis and precision of air pathogen monitoring in the prior art is solved, and real-time efficient identification and accuracy of air pathogens are achieved. Disinfect.
Patent Information
- Application Number
- CN202411200645.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-29
- Publication Date
- 2024-11-08
AI Technical Summary
The existing air pathogen monitoring technology has lag, insufficient detection accuracy, inability to distinguish multiple pathogens and lacks dynamic disinfection capabilities, resulting in an increase in public health risks.
The double-layer generative adversarial network and fuzzy logic control technology are adopted, combined with electrostatic adsorption cell and turbofan system, and real-time identification and precise disinfection of pathogens in the air through air circulation, temperature difference cultivation and fuzzy logic processing.
It significantly improves the ability to capture pathogen signals, reduces the false alarm rate and missed alarm rate, realizes efficient identification and detection of pathogens in the air, and improves detection accuracy and disinfection efficiency.
Smart Images

Figure CN118912663A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of pathogen monitoring technology, and in particular to a device and method for real-time monitoring and disinfection of airborne pathogens. Background Art
[0002] In the current environmental monitoring and air purification technology, especially in the field of air pathogen monitoring and disinfection, the existing technical means have many limitations. Usually, the existing air pathogen monitoring methods mostly rely on the analysis of human biological samples, and infer the distribution of pathogens in the environment by detecting individual infection conditions. However, this method has obvious lags. When an infection case is detected, the pathogen may have spread to a larger area, leading to regional or large-scale public health risks.
[0003] Existing air monitoring equipment usually only has a simple particle detection function and cannot effectively distinguish pathogens from other non-pathogenic particles. When monitoring, the equipment can only detect the total particle concentration in the air, but cannot identify the type and concentration of pathogens. This technical defect causes existing equipment to be unable to provide targeted detection and disinfection methods when facing multiple pathogens.
[0004] In terms of monitoring accuracy, existing technologies mainly rely on a single detection method, such as laboratory analysis after a single sampling, or simple optical or chemical sensors, which cannot achieve dynamic and continuous monitoring of pathogens in the air. At the same time, when the concentration of air pathogens is low or the environment is complex, the detection accuracy of existing equipment is often insufficient, and it is easy to miss or report a false alarm. In addition, it is difficult for existing technologies to distinguish and detect different types of pathogens. In the face of a complex environment where multiple pathogens coexist, existing technologies cannot respond effectively, thereby increasing the risk of airborne diseases.
[0005] In summary, the existing air pathogen monitoring and disinfection technology has many deficiencies, including the lag of monitoring methods, insufficient detection accuracy of equipment, difficulty in distinguishing and detecting multiple pathogens, and the non-targeted and lack of dynamic adjustment capabilities of disinfection methods. In the context of the increasingly severe global public health challenges, it exposes greater risks and deficiencies. Therefore, there is an urgent need for a real-time air pathogen monitoring and disinfection device and method to solve the problems in the existing technology. Summary of the invention
[0006] One purpose of the present invention is to provide a real-time monitoring and disinfection device and method for air pathogens. The present invention improves the ability to capture pathogen signals, greatly reduces the false alarm rate and missed alarm rate, and realizes efficient identification and detection of pathogens in the air.
[0007] According to an embodiment of the present invention, a real-time monitoring and disinfection device for air pathogens comprises an air filter base (1) and a replaceable detection chamber (2), wherein the replaceable detection chamber (2) is detachably mounted on the air filter base (1) via an outer quick positioning point, and comprises the following structure:
[0008] A purification unit housing (11), wherein an upper mounting cavity is provided in the purification unit housing (11), a mounting back plate (12) is provided on the top of the purification unit housing (11) to protect the upper mounting cavity, a filter air outlet (13) is provided on the side of the purification unit housing (11) to discharge air in the upper mounting cavity, and a replaceable filter element (18) is provided on the filter air outlet (13);
[0009] A detection chamber housing (23), a status display cover (22) for displaying the air pathogen monitoring and disinfection status is installed at the bottom of the detection chamber housing (23), a lower installation cavity is opened in the detection chamber housing (23), and a filter air inlet (21) is opened on the side of the detection chamber housing (23) for introducing air into the lower installation cavity;
[0010] The motor (14) is arranged at the upper part of the upper mounting cavity, and the rotating shaft end of the motor (14) is arranged downward, and the rotating shaft end of the motor (14) is sleeved with a first turbofan (15), a bearing (16) and a second turbofan (26) in sequence from top to bottom;
[0011] A first turbofan (15) and a second turbofan (26), wherein the first turbofan (15) and the second turbofan (26) have the same structure, and the first turbofan (15) and the second turbofan (26) are respectively arranged in the upper installation cavity and the lower installation cavity;
[0012] A bearing (16), wherein a power supply circuit board (17) is disposed on the outer casing of the bearing (16), and the power supply circuit board (17) provides power support for various components;
[0013] The electrostatic adsorption battery cell (27) is clamped at the bottom of the lower mounting cavity through a fixed clamping slot (272) arranged at the bottom, and the air duct (275) of the electrostatic adsorption battery cell (27) is arranged corresponding to the filter air inlet (21);
[0014] The circuit mounting seat (25) is arranged above the electrostatic adsorption battery core (27). A detection circuit (24) is arranged on the circuit mounting seat (25). The electrostatic adsorption battery core (27) adsorbs pathogen particles and is monitored by the detection circuit (24).
[0015] Optionally, the first turbofan (15) includes a turbofan body (151), the center of the turbofan body (151) is provided with a bearing terminal (152) which is sleeved on the end of the rotating shaft of the motor (14), and a plurality of arc-shaped blades (153) are arranged in an array at equal intervals outside the turbofan body (151), and the blades (153) are fixed by a blade clamp.
[0016] Optionally, the electrostatic adsorption battery cell (27) comprises a cross-shaped core seat (271), a plurality of silicon steel sheets (274) are arranged in parallel at the side ends of the cross-shaped core seat (271), an air duct (275) is formed between the silicon steel sheets (274), and an electrode (273) is installed on the top of the silicon steel sheet (274).
[0017] A monitoring and disinfection method of an air pathogen real-time monitoring and disinfection device comprises the following steps:
[0018] S1, using an air collection unit disposed inside the replaceable detection chamber (2), and using a second turbofan (26) to introduce air from the area to be detected into the replaceable detection chamber (2) to form an air circulation;
[0019] S2, using the electrostatic adsorption cell (27) and the second turbofan (26) to vibrate and separate the aerosols and suspended particles in the air, so as to achieve preliminary separation of pathogens and other non-pathogenic particles in the air;
[0020] S3, a temperature difference incubation unit is also provided in the replaceable detection chamber (2), and the air separated by vibration is passed through the temperature difference incubation unit provided inside the device, so that the pathogen is incubated for a predetermined time under specific temperature conditions to obtain a pathogen signal;
[0021] S4, the detection circuit (24) is built with a monitoring internal reference unit and a monitoring internal reference unit, and the collected pathogen signals are pre-processed by fuzzy logic in the monitoring internal reference unit, and pathogen signals of different concentrations and types are converted into fuzzy sets using fuzzy rules, and preliminary screening and fuzzy decision making are performed on the pathogen signals;
[0022] S5, inputting the pathogen signal after fuzzy logic processing into the first-layer generation network of the two-layer generative adversarial network to generate hypothetical pathogen samples, and the first-layer discriminant network discriminates the generated hypothetical pathogen samples, and optimizes the ability to capture pathogen signals through iterative training;
[0023] S6, inputting the signal output by the first-layer generative adversarial network into the second-layer generative adversarial network to generate a pathogen characteristic signal, and the second discriminant network verifies and optimizes the generated signal;
[0024] S7, introducing the pathogen signal processed by the double-layer generative adversarial network into the monitoring internal reference unit, reacting the internal reference agent with the pathogen to generate a detectable signal, and capturing and analyzing the pathogen through the detection circuit (24), thereby realizing the monitoring of pathogens in the air;
[0025] S8. After the monitoring is completed, the air flowing out of the replaceable detection chamber (2) is filtered and disinfected through the filter element (18) to remove the toxic and harmful gases and other pathogens remaining in the air, and is discharged from the air filter base (1) through the filter outlet (13);
[0026] S9, dynamically adjusting the response parameters of the internal reference agent in the monitoring internal reference unit according to the detected pathogen signal to adapt to pathogens of different types and concentrations;
[0027] S10. According to the type of pathogen to be monitored, select an appropriate monitoring internal reference agent and set it in the monitoring internal reference unit for targeted monitoring and detection of different pathogens.
[0028] Optionally, S2 includes the following steps:
[0029] S21, after the air circulation is formed, the second turbofan (26) is started, and the turbofan body (151) generates a vortex through the arc-shaped blades (153), and the aerosol and suspended particles in the air are concentratedly guided along the direction of the air duct (275);
[0030] S22, using the electrostatic field generated by the electrostatic adsorption cell (27) to adsorb and separate the passing aerosols and suspended particles using the silicon steel sheet (274) on the cross-shaped core seat (271), the electrostatic field strength is E, E represents the electric field strength generated by the electrostatic adsorption cell (27), and the electric field strength is dynamically adjusted according to the charged characteristics of the aerosols and suspended particles to be separated;
[0031] S23, when the aerosol and suspended particles pass through the electrostatic adsorption cell (27), the voltage difference V between the electrodes (273) is used to make the particles with different charges adsorbed on the surface of the silicon steel sheet (274) according to the charge polarity of the particles, wherein V is the voltage across the electrodes, and the voltage value is set according to the charge amount of the particles to be separated;
[0032] S24, through the air duct (275) of the electrostatic adsorption battery (27), the separated aerosol and suspended particles are separated in different directions, the larger pathogenic particles are adsorbed to the silicon steel sheet (274), and the non-pathogenic particles are directed to the filter outlet (13) for subsequent treatment;
[0033] S25. During the adsorption process of the electrostatic adsorption battery cell (27), the second turbofan (26) and the electrostatic adsorption battery cell (27) are operated synchronously to maintain continuous air flow.
[0034] Optionally, S3 includes the following steps:
[0035] S31, after the pathogens and non-pathogenic particles are initially separated, the air is guided to a temperature difference culture unit provided in the replaceable detection chamber (2), the temperature difference culture unit comprising a temperature controller and a culture unit, the temperature controller forms a desired temperature gradient by adjusting the power of a heating element and a cooling element;
[0036] S32, the pathogens in the air stay in the temperature difference culture unit for a preset time, and the temperature in the culture unit is maintained within the temperature range T that is optimal for the growth of the pathogens by controlling the temperature gradient. opt , T opt Indicates the optimal culture temperature set in the temperature difference culture unit, which is dynamically adjusted according to the growth characteristics of different pathogens;
[0037] S33, when the pathogens are cultured within a predetermined time, the pathogens in the culture unit are detected by the sensor and the generated pathogen signal S is recorded. signal .
[0038] Optionally, S4 includes the following steps:
[0039] S41, in pathogen signal S signal After being generated, the pathogen signal is input into the monitoring internal reference unit built into the detection circuit (24). The monitoring internal reference unit uses the collected pathogen signal S signal Perform preliminary pretreatment to obtain the preliminary pretreated pathogen signal S pre (t);
[0040] S42, in the pre-processing process, the pathogen signal S is converted into pre (t) According to the concentration and type, it is transformed into a fuzzy set F set , where the fuzzy set F set It is determined according to the fuzzy membership function μ(x) of the pathogen signal, which represents the membership of the signal intensity x at a specific concentration and type:
[0041]
[0042] Among them, μ(x) is the membership function of the pathogen signal x, c is the central value of the pathogen signal intensity, σ is the standard deviation of the membership function, and X is the complete set of pathogen signals;
[0043] S43, for the fuzzy set F set Fuzzy rule reasoning is performed on pathogen signals in the i , the input signals of different fuzzy sets are screened and judged to form fuzzy decision Dfuzzy :
[0044]
[0045] Among them, D fuzzy As a result of fuzzy decision making, and are the membership functions of pathogen signals in different fuzzy rules, respectively, and x and y are variables of different input signals;
[0046] S44, based on fuzzy decision D fuzzy , preliminarily screen out pathogen signals that meet the detection criteria, and pass the pathogen signals to the two-layer generative adversarial network unit:
[0047]
[0048] Among them, S filtered represents the pathogen signal after screening, γ is the target intensity of the pathogen signal, δ is the width parameter of the signal, and D fuzzy is the fuzzy decision result, and x and y are the input variables of the pathogen signal.
[0049] Optionally, S5 includes the following steps:
[0050] S51, the pathogen signal S after fuzzy decision filtered Input to the first-layer generative network of the two-layer generative adversarial network in real time, and generate a hypothetical pathogen sample G1(z) through the first-layer generative network, where G1(z) is the pathogen sample generated by the first-layer generative network in the current time period, and z is the dynamic noise vector. The goal of the first-layer generative network is to adjust the parameters in real time to make the generated hypothetical pathogen sample G1(z) close to the real pathogen sample;
[0051] S52, inputting the generated hypothetical pathogen sample G1(z) into the first-layer discriminant network in real time, the first-layer discriminant network evaluates the authenticity of the sample by calculating the confidence of the hypothetical pathogen sample, and the output of the first-layer discriminant network is D1(G1(z)), where D1(G1(z)) represents the confidence of the first-layer discriminant network for the hypothetical pathogen sample under the current conditions, and the confidence is calculated based on the difference function between the hypothetical pathogen sample and the real signal collected in real time;
[0052] S53, through the online iterative training of the first layer generation network and the first layer discrimination network, dynamically optimize the parameters θ of the generation network g and the parameters θ of the first layer of the discriminant network d , to achieve the minimization of the difference function between the hypothetical pathogen sample and the real pathogen sample signal collected in real time
[0053]
[0054] in, represents the real-time expected value, p data (x) is the distribution of real pathogen signals collected in real time, p z (z) is the input noise distribution of the first layer of the generative network, θ g and θ d They are the real-time parameter sets of the first-layer generating network and the first-layer discriminative network respectively;
[0055] S54. Through multiple rounds of online iterations, the optimized first-layer generative network generates hypothetical pathogen samples that are closer to the real pathogen signal. Confidence function of the first layer of discriminant network Towards real-time confidence in true pathogen signals.
[0056] Optionally, the S6 includes the following steps:
[0057] S61. Generate the pathogen hypothesis signal output by the first layer of the adversarial network Input to the second-layer generation network of the two-layer generation adversarial network in real time. The second-layer generation network generates and processes the input pathogen hypothesis signal to generate a pathogen characteristic signal G2(z′), wherein G2(z′) is the pathogen characteristic signal generated by the second-layer generation network in the current time period, and z' is the corrected noise vector received by the second-layer generation network. The goal of the second-layer generation network is to optimize the authenticity and specificity of the pathogen characteristic signal G2(z′);
[0058] S62, inputting the generated pathogen characteristic signal G2(z′) into the second-layer discriminant network in real time, and the second-layer discriminant network verifies the signal by calculating the confidence D2(G2(z′)) of the pathogen characteristic signal, wherein D2(G2(z′)) represents the confidence of the second-layer discriminant network in the pathogen characteristic signal under the current conditions, and the confidence is calculated based on the difference function between the pathogen characteristic signal and the expected true pathogen signal;
[0059] S63, through the online iterative training of the second layer generation network and the second layer discrimination network, optimize the parameter θ of the second layer generation network g2 and the parameters θ of the second layer discriminant network d2 , so that the difference function between the pathogen characteristic signal and the real pathogen signal Minimize:
[0060]
[0061] in, represents the real-time expected value, p data(x) is the expected true pathogen signal distribution, p z′ (z′) is the input noise distribution of the second layer generator network, θ g2 and θ d2 They are the real-time parameter sets for the second-layer generating network and the second-layer discriminative network respectively;
[0062] S64, after multiple rounds of online iterations, the pathogen characteristic signal is finally generated Closer to the real pathogen signal, while completing the real-time pathogen signal capture, the final generated pathogen characteristic signal Transmitted to the disinfection unit.
[0063] Optionally, the S7 includes the following steps:
[0064] S71, the pathogen characteristic signal finally generated The monitoring internal reference unit is introduced in real time. The monitoring internal reference unit reacts specifically with the characteristic signal of the pathogen through the internal reference agent to generate a detectable signal S detect :
[0065]
[0066] Among them, α, β, γ and λ are the reaction parameters of the internal reference agent, α represents the initial reaction intensity of the agent, β represents the attenuation coefficient of the reaction, γ is the enhancement factor of the agent, λ represents the attenuation rate of the signal over time, t0 is the initial time point of the reaction, and τ is the integral variable;
[0067] S72, generating a detectable signal S detect After that, the signal is transmitted to the detection circuit (24) in real time, and the detection circuit amplifies and filters the signal. The processed signal is recorded as S amplified (t):
[0068]
[0069] Where η is the amplifier gain, σ is the filter gain factor, δ is the filter time constant, κ is the signal nonlinear correction coefficient, and t1 is the starting time point of signal processing;
[0070] S73, the amplified signal S is analyzed by the analysis module in the detection circuit (24). amplified (t) is captured and analyzed, and the analysis result is recorded as the pathogen detection value D pathogen (t):
[0071]
[0072] Where ω is the signal modulation frequency, ζ· is the nonlinear analysis factor, ρ is the signal modulation amplitude, and t2 is the analysis start time point;
[0073] S74, according to pathogen detection value D pathogen (t) changes in real time, monitor the internal reference unit to dynamically adjust the reaction parameters of the internal reference agent, including agent concentration, reaction time and temperature, and monitor pathogens of different types and concentrations:
[0074]
[0075] Among them, θ up and θ down are the upper and lower thresholds of the detection value change rate.
[0076] The beneficial effects of the present invention are:
[0077] (1) The present invention significantly improves the real-time monitoring and disinfection capabilities of airborne pathogens by introducing a two-layer generative adversarial network and fuzzy logic control technology. Through the two-layer generative adversarial network structure, the collected pathogen signals can be deeply modeled and feature extracted. The first layer of the generative adversarial network generates and discriminates hypothetical samples of preliminary pathogen signals. Through multiple iterative training, the parameters of the generation network and the discrimination network are continuously optimized to ensure that the generated pathogen hypothesis samples have a high degree of authenticity and availability. The second layer of the generative adversarial network further refines and optimizes the characteristics of the pathogen signals, and uses the optimized pathogen characteristic signals for subsequent precise monitoring and disinfection processing, thereby improving the capture capability of pathogen signals, greatly reducing the false alarm rate and missed alarm rate, and realizing efficient identification and detection of pathogens in the air.
[0078] (2) The present invention preprocesses and screens pathogen signals through a fuzzy logic controller, so that pathogen signals of different concentrations and types can be optimized before entering the generative adversarial network. The fuzzy logic controller converts pathogen signals into fuzzy sets and forms preliminary fuzzy decisions through fuzzy rule reasoning, effectively solving the problem that traditional methods are difficult to accurately detect in complex environments.
[0079] (3) The present invention combines the characteristic signal of the pathogen with the specific reaction of the internal reference agent, generates a detectable signal through the dynamic reaction of the internal reference agent with the characteristic signal of the pathogen, and amplifies, filters and analyzes the signal through the detection circuit to ensure high-sensitivity detection of the pathogen signal, real-time monitoring of the pathogen characteristics, and dynamic adjustment of the reaction parameters of the internal reference agent to achieve accurate monitoring and disinfection of different pathogens.
[0080] (4) The real-time air pathogen monitoring and disinfection device of the present invention significantly improves the efficiency and accuracy of air pathogen detection and disinfection. The device adopts a replaceable detection chamber structure, which can flexibly adapt to the air quality monitoring needs in different environments. The electrostatic adsorption cell and turbofan system in the detection chamber are used in combination to effectively capture and separate pathogen particles in the air. BRIEF DESCRIPTION OF THE DRAWINGS
[0081] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:
[0082] Figure 1 This is a device structure diagram of a device and method for real-time monitoring and disinfection of air pathogens proposed by the present invention;
[0083] Figure 2 This is a front view of a device and method for real-time monitoring and disinfection of airborne pathogens proposed by the present invention;
[0084] Figure 3 A cross-sectional view of a device and method for real-time monitoring and disinfection of airborne pathogens proposed by the present invention;
[0085] Figure 4 This is a detection circuit structure diagram of a device and method for real-time monitoring and disinfection of air pathogens proposed by the present invention;
[0086] Figure 5 A structural diagram of the connection between the second turbofan and the electrostatic adsorption battery core of a device and method for real-time monitoring and disinfection of air pathogens proposed by the present invention;
[0087] Figure 6 This is a structural diagram of an electrostatic adsorption battery cell of a device and method for real-time monitoring and disinfection of air pathogens proposed by the present invention;
[0088] Figure 7 This is a second turbofan structure diagram of a device and method for real-time monitoring and disinfection of air pathogens proposed by the present invention.
[0089] In the figure: 1-air filter base, 11-purification unit housing, 12-installation back plate, 13-filter outlet, 14-motor, 15-first turbofan, 151-turbofan body, 152-bearing terminal, 153-blade, 154-blade clamp, 16-bearing, 17-power supply circuit board, 18-filter element, 2-replaceable detection compartment, 21-filter air inlet, 22-status display cover, 23-detection compartment housing, 24-detection circuit, 25-circuit mounting seat, 26-second turbofan, 27-electrostatic adsorption battery, 271-cross core seat, 272-fixed card slot, 273-electrode, 274-silicon steel sheet, 275-air duct. DETAILED DESCRIPTION
[0090] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.
[0091] refer to Figure 1-7 , a real-time monitoring and disinfection device for air pathogens, comprising an air filter base 1 and a replaceable detection chamber 2, wherein the replaceable detection chamber 2 is detachably mounted on the air filter base 1 through an outer quick positioning point, and comprises the following structure:
[0092] A purification unit housing 11, an upper mounting cavity is provided in the purification unit housing 11, a mounting back plate 12 is provided on the top of the purification unit housing 11 to protect the upper mounting cavity, a filter air outlet 13 is provided on the side of the purification unit housing 11 to discharge the air in the upper mounting cavity, and a replaceable filter element 18 is provided on the filter air outlet 13;
[0093] A detection chamber housing 23, a status display cover 22 for displaying the air pathogen monitoring and disinfection status is installed at the bottom of the detection chamber housing 23, a lower installation cavity is opened in the detection chamber housing 23, and a filter air inlet 21 is opened on the side of the detection chamber housing 23 for introducing air into the lower installation cavity;
[0094] The motor 14 is arranged at the upper part of the upper mounting cavity, and the rotating shaft end of the motor 14 is arranged downward, and the rotating shaft end of the motor 14 is sleeved with the first turbofan 15, the bearing 16 and the second turbofan 26 in sequence from top to bottom;
[0095] The first turbofan 15 and the second turbofan 26 have the same structure, and the first turbofan 15 and the second turbofan 26 are respectively arranged in the upper mounting cavity and the lower mounting cavity;
[0096] The bearing 16 has a power supply circuit board 17 disposed on its outer cover, which provides power support for various components;
[0097] The electrostatic adsorption cell 27 is clamped at the bottom of the lower mounting cavity through a fixed clamping slot 272 provided at the bottom, and the air duct 275 of the electrostatic adsorption cell 27 is provided corresponding to the filter air inlet 21;
[0098] The circuit mounting seat 25 is arranged above the electrostatic adsorption battery cell 27 . A detection circuit 24 is arranged on the circuit mounting seat 25 . The electrostatic adsorption battery cell 27 adsorbs pathogen particles and monitors them through the detection circuit 24 .
[0099] In this embodiment, the first turbofan 15 includes a turbofan body 151, and a bearing terminal 152 is arranged at the center of the turbofan body 151 and is sleeved on the rotating shaft end of the motor 14. A plurality of arc-shaped blades 153 are arranged in an array at equal intervals outside the turbofan body 151, and the blades 153 are fixed by blade clamps.
[0100] In this embodiment, the electrostatic adsorption battery cell 27 includes a cross-shaped core base 271 , and a plurality of silicon steel sheets 274 are arranged in parallel at the side ends of the cross-shaped core base 271 , and air ducts 275 are formed between the silicon steel sheets 274 . Electrodes 273 are installed on the tops of the silicon steel sheets 274 .
[0101] A monitoring and disinfection method of an air pathogen real-time monitoring and disinfection device comprises the following steps:
[0102] S1, through the air collection unit arranged inside the replaceable detection chamber 2, the air in the area to be detected is introduced into the replaceable detection chamber 2 by using the second turbofan 26 to form an air circulation;
[0103] S2, using the electrostatic adsorption cell 27 and the second turbofan 26 to vibrate and separate the aerosols and suspended particles in the air, so that the pathogens in the air are initially separated from other non-pathogenic particles;
[0104] S3, a temperature difference incubation unit is also provided in the replaceable detection chamber 2, and the air after vibration separation passes through the temperature difference incubation unit provided inside the device, so that the pathogen is incubated for a predetermined time under specific temperature conditions to obtain a pathogen signal;
[0105] S4, the detection circuit 24 is built with a monitoring internal reference unit and a monitoring internal reference unit, and the collected pathogen signals are preprocessed by the fuzzy logic in the monitoring internal reference unit, and the pathogen signals of different concentrations and types are converted into fuzzy sets by using fuzzy rules, and the pathogen signals are preliminarily screened and fuzzy decisions are made;
[0106] S5, inputting the pathogen signal after fuzzy logic processing into the first-layer generation network of the two-layer generative adversarial network to generate hypothetical pathogen samples, and the first-layer discriminant network discriminates the generated hypothetical pathogen samples, and optimizes the ability to capture pathogen signals through iterative training;
[0107] S6, inputting the signal output by the first-layer generative adversarial network into the second-layer generative adversarial network to generate a pathogen characteristic signal, and the second discriminant network verifies and optimizes the generated signal;
[0108] S7, introducing the pathogen signal processed by the double-layer generative adversarial network into the monitoring internal reference unit, reacting the internal reference agent with the pathogen to generate a detectable signal, and capturing and analyzing the pathogen through the detection circuit 24 to achieve monitoring of pathogens in the air;
[0109] S8. After the monitoring is completed, the air flowing out of the replaceable detection chamber 2 is filtered and disinfected through the filter element 18 to remove the toxic and harmful gases and other pathogens remaining in the air, and then discharged from the air filter base 1 through the filter outlet 13;
[0110] S9, dynamically adjusting the response parameters of the internal reference agent in the monitoring internal reference unit according to the detected pathogen signal to adapt to pathogens of different types and concentrations;
[0111] S10. According to the type of pathogen to be monitored, select an appropriate monitoring internal reference agent and set it in the monitoring internal reference unit for targeted monitoring and detection of different pathogens.
[0112] In this implementation, S2 includes the following steps:
[0113] S21, after the air circulation is formed, the second turbofan 26 is started, and the turbofan body 151 generates a vortex through the arc-shaped blades 153, and the aerosols and suspended particles in the air are concentratedly guided along the direction of the air duct 275;
[0114] S22, the electrostatic field generated by the electrostatic adsorption cell 27 is used to adsorb and separate the passing aerosols and suspended particles using the silicon steel sheet 274 on the cross-shaped core seat 271. The electrostatic field strength is E, where E represents the electric field strength generated by the electrostatic adsorption cell 27. The electric field strength is dynamically adjusted according to the charged characteristics of the aerosols and suspended particles to be separated;
[0115] S23, when the aerosol and suspended particles pass through the electrostatic adsorption cell 27, the voltage difference V between the electrodes 273 is used to make the particles with different charges be adsorbed on the surface of the silicon steel sheet 274 according to the charge polarity of the particles, wherein V is the voltage across the electrodes, and the voltage value is set according to the charge amount of the particles to be separated;
[0116] S24, through the air duct 275 of the electrostatic adsorption battery 27, the separated aerosols and suspended particles are separated in different directions, the larger pathogenic particles are adsorbed to the silicon steel sheet 274, and the non-pathogenic particles are directed to the filter outlet 13 for subsequent processing;
[0117] S25. During the adsorption process of the electrostatic adsorption battery cell 27, the second turbofan 26 and the electrostatic adsorption battery cell 27 are operated synchronously to maintain continuous air flow.
[0118] In this implementation, S3 includes the following steps:
[0119] S31, after the pathogens and non-pathogenic particles are initially separated, the air is guided to the temperature difference culture unit provided in the replaceable detection chamber 2, the temperature difference culture unit includes a temperature controller and a culture unit, and the temperature controller forms a required temperature gradient by adjusting the power of the heating element and the cooling element;
[0120] The culture unit is used to provide an optimal proliferation environment for pathogens in the air, thereby enhancing the strength and reliability of pathogen signals. The core design concept of the culture unit in this embodiment is to provide stable and controllable culture conditions while ensuring the activity of the pathogen. The culture unit includes a temperature control system and a constant temperature culture chamber, the size of which is designed according to the replaceable detection chamber 2, which is not shown in the figure. The specific culture unit is set in the replaceable detection chamber 2 below the detection circuit 24. The temperature control system consists of a temperature sensor that can respond quickly and a set of integrated heating and cooling elements. The elements are monitored and adjusted in real time by the central control module to maintain the temperature in the culture chamber within the most suitable growth range for the pathogen. The interior of the culture chamber uses high conductivity materials to ensure uniform temperature distribution and slow internal air flow to avoid interfering with the natural proliferation process of the pathogen. The culture unit is also equipped with an automatic cleaning function. After each culture cycle, the system will automatically start the cleaning program to disinfect the culture chamber using ultraviolet irradiation and ozone treatment to ensure the pollution-free nature of the next culture cycle.
[0121] S32, the pathogens in the air stay in the temperature difference culture unit for a preset time, and the temperature in the culture unit is maintained within the temperature range T that is optimal for the growth of the pathogens by controlling the temperature gradient. opt , T opt Indicates the optimal culture temperature set in the temperature difference culture unit, which is dynamically adjusted according to the growth characteristics of different pathogens;
[0122] S33, when the pathogens are cultured within a predetermined time, the pathogens in the culture unit are detected by the sensor and the generated pathogen signal S is recorded. signal .
[0123] In this implementation, S4 includes the following steps:
[0124] S41, in pathogen signal S signal After being generated, the pathogen signal is input into the monitoring internal reference unit built into the detection circuit 24. The monitoring internal reference unit uses the collected pathogen signal S signal Perform preliminary pretreatment to obtain the preliminary pretreated pathogen signal S pre (t);
[0125] S42, in the pre-processing process, the pathogen signal S is converted into pre (t) According to the concentration and type, it is transformed into a fuzzy set F set, where the fuzzy set F set It is determined according to the fuzzy membership function μ(x) of the pathogen signal, which represents the membership of the signal intensity x at a specific concentration and type:
[0126]
[0127] Among them, μ(x) is the membership function of the pathogen signal x, c is the central value of the pathogen signal intensity, σ is the standard deviation of the membership function, and X is the complete set of pathogen signals;
[0128] S43, for the fuzzy set F set Fuzzy rule reasoning is performed on pathogen signals in the i , the input signals of different fuzzy sets are screened and judged to form fuzzy decision D fuzzy :
[0129]
[0130] Among them, D fuzzy As a result of fuzzy decision making, and are the membership functions of pathogen signals in different fuzzy rules, respectively, and x and y are variables of different input signals;
[0131] S44, based on fuzzy decision D fuzzy , preliminarily screen out pathogen signals that meet the detection criteria, and pass the pathogen signals to the two-layer generative adversarial network unit:
[0132]
[0133] Among them, S filtered represents the pathogen signal after screening, γ is the target intensity of the pathogen signal, δ is the width parameter of the signal, and D fuzzy is the fuzzy decision result, and x and y are the input variables of the pathogen signal.
[0134] In this implementation, S5 includes the following steps:
[0135] S51, the pathogen signal S after fuzzy decision filtered Input to the first-layer generative network of the two-layer generative adversarial network in real time, and generate a hypothetical pathogen sample G1(z) through the first-layer generative network, where G1(z) is the pathogen sample generated by the first-layer generative network in the current time period, and z is the dynamic noise vector. The goal of the first-layer generative network is to adjust the parameters in real time to make the generated hypothetical pathogen sample G1(z) close to the real pathogen sample;
[0136] S52, inputting the generated hypothetical pathogen sample G1(z) into the first-layer discriminant network in real time, the first-layer discriminant network evaluates the authenticity of the sample by calculating the confidence of the hypothetical pathogen sample, and the output of the first-layer discriminant network is D1(G1(z)), where D1(G1(z)) represents the confidence of the first-layer discriminant network for the hypothetical pathogen sample under the current conditions, and the confidence is calculated based on the difference function between the hypothetical pathogen sample and the real signal collected in real time;
[0137] S53, through the online iterative training of the first layer generation network and the first layer discrimination network, dynamically optimize the parameters θ of the generation network g and the parameters θ of the first layer of the discriminant network d , to achieve the minimization of the difference function between the hypothetical pathogen sample and the real pathogen sample signal collected in real time
[0138]
[0139] in, represents the real-time expected value, p data (x) is the distribution of real pathogen signals collected in real time, p z (z) is the input noise distribution of the first layer of the generative network, θ g and θ d They are the real-time parameter sets of the first-layer generating network and the first-layer discriminative network respectively;
[0140] S54. Through multiple rounds of online iterations, the optimized first-layer generative network generates hypothetical pathogen samples that are closer to the real pathogen signal. Confidence function of the first layer of discriminant network Towards real-time confidence in true pathogen signals.
[0141] In this implementation, S6 includes the following steps:
[0142] S61. Generate the pathogen hypothesis signal output by the first layer of the adversarial network The second-layer generative network of the two-layer generative adversarial network is input in real time. The second-layer generative network generates and processes the input pathogen hypothesis signal to generate the pathogen characteristic signal G2(z′), where G2(z ′ ) is the pathogen characteristic signal generated by the second-layer generation network in the current time period, z' is the corrected noise vector received by the second-layer generation network, and the goal of the second-layer generation network is to optimize the authenticity and specificity of the pathogen characteristic signal G2(z');
[0143] S62, inputting the generated pathogen characteristic signal G2(z′) into the second-layer discriminant network in real time, and the second-layer discriminant network verifies the signal by calculating the confidence D2(G2(z′)) of the pathogen characteristic signal, wherein D2(G2(z′)) represents the confidence of the second-layer discriminant network in the pathogen characteristic signal under the current conditions, and the confidence is calculated based on the difference function between the pathogen characteristic signal and the expected true pathogen signal;
[0144] S63, through the online iterative training of the second layer generation network and the second layer discrimination network, optimize the parameter θ of the second layer generation network g2 and the parameters θ of the second layer discriminant network d2 , so that the difference function between the pathogen characteristic signal and the real pathogen signal Minimize:
[0145]
[0146] in, represents the real-time expected value, p data (x) is the expected true pathogen signal distribution, p z′ (z′) is the input noise distribution of the second layer generator network, θ g2 and θ d2 They are the real-time parameter sets for the second-layer generating network and the second-layer discriminative network respectively;
[0147] S64, after multiple rounds of online iterations, the pathogen characteristic signal is finally generated Closer to the real pathogen signal, while completing the real-time pathogen signal capture, the final generated pathogen characteristic signal Transmitted to the disinfection unit.
[0148] In this implementation, S7 includes the following steps:
[0149] S71, the pathogen characteristic signal finally generated The monitoring internal reference unit is introduced in real time. The monitoring internal reference unit reacts specifically with the characteristic signal of the pathogen through the internal reference agent to generate a detectable signal S detect :
[0150]
[0151] Among them, α, β, γ and λ are the reaction parameters of the internal reference agent, α represents the initial reaction intensity of the agent, β represents the attenuation coefficient of the reaction, γ is the enhancement factor of the agent, λ represents the attenuation rate of the signal over time, t0 is the initial time point of the reaction, and τ is the integral variable;
[0152] S72, generating a detectable signal S detectAfter that, the signal is transmitted to the detection circuit 24 in real time, and the detection circuit amplifies and filters the signal. The processed signal is recorded as S amplified (t):
[0153]
[0154] Where η is the amplifier gain, σ is the filter gain factor, δ is the filter time constant, κ is the signal nonlinear correction coefficient, and t1 is the starting time point of signal processing;
[0155] S73, the analysis module in the detection circuit 24 analyzes the amplified signal S amplified (t) is captured and analyzed, and the analysis result is recorded as the pathogen detection value D pathogen (t):
[0156]
[0157] Where ω is the signal modulation frequency, ζ· is the nonlinear analysis factor, ρ is the signal modulation amplitude, and t2 is the analysis start time point;
[0158] S74, according to pathogen detection value D pathogen (t) changes in real time, monitor the internal reference unit to dynamically adjust the reaction parameters of the internal reference agent, including agent concentration, reaction time and temperature, and monitor pathogens of different types and concentrations:
[0159]
[0160] Among them, θ up and θ down are the upper and lower thresholds of the detection value change rate.
[0161] Embodiment 1:
[0162] On a winter morning when the flu epidemic was high, the air in the emergency department of a large general hospital in the city center was filled with a variety of pathogens. In order to prevent the spread of the flu virus in the hospital, the hospital decided to deploy the air pathogen real-time monitoring and disinfection device of the present invention in the emergency department ward area to test its actual effect in a high-risk environment. The device was installed on the ceiling of the central area of the emergency department, covering the entire area including the waiting area and wards.
[0163] At 9:30 am on the same day, the device started and began to automatically collect air samples in the emergency department. The suspended particles and pathogens in the air were introduced into the replaceable detection chamber with the air flow. The second turbofan circulated the air in the detection chamber to make the collected air samples representative. The aerosols and suspended particles in the air were effectively separated through the 3.5kV / cm electrostatic field generated by the electrostatic adsorption cell. The detected pathogens included influenza virus H3N2 and some common bacteria.
[0164] Within the first 30 minutes after sampling, the temperature difference culture unit in the detection chamber maintained the temperature at 33°C. At this time, the influenza virus in the air proliferated in the culture unit and generated a stable pathogen signal. After the culture was completed, the monitoring internal reference unit pre-processed the pathogen signal through the fuzzy logic controller and identified the characteristic signal of the H3N2 influenza virus. Subsequently, the two-layer generative adversarial network started to run. The first-layer generative network generated a hypothetical H3N2 virus sample and verified it through the discriminant network. After 20 iterative training, the generated hypothetical pathogen sample matched the actual signal to more than 95%.
[0165] At 10:15, the monitoring system detected that the concentration of influenza virus in the air of the emergency department increased rapidly, reaching 150 virus particles per cubic meter. The system immediately generated a detailed environmental report, recording the virus concentration, distribution area and change trend. Then, the device triggered the alarm mechanism and transmitted the data to the hospital's central control room. Based on the real-time alarm signal, the hospital management immediately initiated corresponding prevention and control measures, including restricting the flow of patients in the emergency department and arranging for medical staff to wear higher-level protective equipment.
[0166] At 10:30, the disinfection unit started operating. Based on the captured H3N2 virus characteristic signal, the concentration of the internal reference agent was dynamically adjusted, and the system began to accurately disinfect the pathogens in the air. The entire disinfection process lasted 15 minutes. After the disinfection, the device sampled and monitored the air again. The results showed that the H3N2 virus concentration in the emergency department had dropped to 5 virus particles per cubic meter, returning to a safe level. In the following 30 minutes, the system continued to monitor the air quality and found no signs of a further increase in the virus concentration.
[0167] To evaluate the effectiveness of the device of the present invention, the hospital simultaneously used traditional air purifiers in other ward areas for comparative testing. Under the same environmental conditions, the traditional air purifier reduced the H3N2 virus concentration in the ward from 150 virus particles per cubic meter to 50 virus particles within 40 minutes of operation. In comparison, the device of the present invention significantly reduced the virus concentration in the air in a shorter time, and provided timely and reliable data support for the hospital's prevention and control measures through automatic alarm and report generation mechanisms.
[0168] Experimental data further show that in different areas of the emergency department, the monitoring and disinfection effects of the device of the present invention are better than those of traditional equipment. In the waiting area of the emergency department, the device reduced the virus concentration from 120 virus particles per cubic meter to 10 virus particles within 20 minutes, while the traditional air purifier took 40 minutes to achieve a similar effect. In addition, the device of the present invention can also distinguish and process different types of pathogen signals in real time, avoiding the problem of improper handling that may occur when traditional air purifiers face multiple pathogens.
[0169] In this experiment, the device of the present invention successfully avoided the large-scale spread of influenza viruses in the hospital through precise pathogen monitoring and rapid disinfection. Hospital managers highly recognized the performance of the device and plan to further promote its use in other high-risk areas to enhance the safety and prevention and control capabilities of the overall environment. The successful application of the present invention not only proves its potential in the field of public health and safety, but also demonstrates its outstanding performance in complex and changeable air environments.
[0170] The present invention significantly improves the real-time monitoring and disinfection capabilities of airborne pathogens by introducing a two-layer generative adversarial network and fuzzy logic control technology. Through the two-layer generative adversarial network structure, the collected pathogen signals can be deeply modeled and feature extracted. The first-layer generative adversarial network generates and discriminates hypothetical samples of preliminary pathogen signals. Through multiple iterative training, the parameters of the generation network and the discrimination network are continuously optimized to ensure that the generated pathogen hypothesis samples are highly authentic and usable. The second-layer generative adversarial network further refines and optimizes the characteristics of the pathogen signals, and uses the optimized pathogen characteristic signals for subsequent precise monitoring and disinfection. This improves the capture capability of pathogen signals, significantly reduces the false alarm rate and missed alarm rate, and realizes efficient identification and detection of pathogens in the air.
[0171] The present invention preprocesses and screens pathogen signals through a fuzzy logic controller, so that pathogen signals of different concentrations and types can be optimized before entering the generative adversarial network. The fuzzy logic controller converts pathogen signals into fuzzy sets and forms preliminary fuzzy decisions through fuzzy rule reasoning, effectively solving the problem that traditional methods are difficult to accurately detect in complex environments.
[0172] The present invention combines the characteristic signal of the pathogen with the specific reaction of the internal reference agent, generates a detectable signal through the dynamic reaction of the internal reference agent with the characteristic signal of the pathogen, and amplifies, filters and analyzes the signal through the detection circuit to ensure high-sensitivity detection of the pathogen signal, real-time monitoring of the pathogen characteristics, dynamic adjustment of the reaction parameters of the internal reference agent, and realization of accurate monitoring and disinfection of different pathogens.
[0173] The real-time monitoring and disinfection device for air pathogens of the present invention significantly improves the efficiency and accuracy of air pathogen detection and disinfection. The device adopts a replaceable detection chamber structure, which can flexibly adapt to air quality monitoring needs in different environments. The electrostatic adsorption battery and turbofan system in the detection chamber are used in combination to effectively capture and separate pathogen particles in the air.
[0174] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.
Claims
1. A real-time monitoring and disinfection device for air pathogens, comprising an air filter base (1) and a replaceable detection chamber (2), wherein the replaceable detection chamber (2) is detachably mounted on the air filter base (1) via an outer quick positioning point, and is characterized in that: Includes the following structure: A purification unit housing (11), wherein an upper mounting cavity is provided in the purification unit housing (11), a mounting back plate (12) is provided on the top of the purification unit housing (11) to protect the upper mounting cavity, a filter air outlet (13) is provided on the side of the purification unit housing (11) to discharge air in the upper mounting cavity, and a replaceable filter element (18) is provided on the filter air outlet (13); A detection chamber housing (23), a status display cover (22) for displaying the air pathogen monitoring and disinfection status is installed at the bottom of the detection chamber housing (23), a lower installation cavity is opened in the detection chamber housing (23), and a filter air inlet (21) is opened on the side of the detection chamber housing (23) for introducing air into the lower installation cavity; The motor (14) is arranged at the upper part of the upper mounting cavity, and the rotating shaft end of the motor (14) is arranged downward, and the rotating shaft end of the motor (14) is sleeved with a first turbofan (15), a bearing (16) and a second turbofan (26) in sequence from top to bottom; A first turbofan (15) and a second turbofan (26), wherein the first turbofan (15) and the second turbofan (26) have the same structure, and the first turbofan (15) and the second turbofan (26) are respectively arranged in the upper installation cavity and the lower installation cavity; A bearing (16), wherein a power supply circuit board (17) is disposed on the outer casing of the bearing (16), and the power supply circuit board (17) provides power support for various components; The electrostatic adsorption battery cell (27) is clamped at the bottom of the lower mounting cavity through a fixed clamping slot (272) arranged at the bottom, and the air duct (275) of the electrostatic adsorption battery cell (27) is arranged corresponding to the filter air inlet (21); The circuit mounting seat (25) is arranged above the electrostatic adsorption battery core (27). A detection circuit (24) is arranged on the circuit mounting seat (25). The electrostatic adsorption battery core (27) adsorbs pathogen particles and is monitored by the detection circuit (24).
2. The device for real-time monitoring and disinfection of airborne pathogens according to claim 1, characterized in that: The first turbofan (15) comprises a turbofan body (151), the center of which is provided with a bearing terminal (152) sleeved on the end of the rotating shaft of the motor (14), and a plurality of arc-shaped blades (153) are arranged in an array at equal intervals outside the turbofan body (151), and the blades (153) are fixed by a blade clamp.
3. The device for real-time monitoring and disinfection of airborne pathogens according to claim 1, characterized in that: The electrostatic adsorption battery cell (27) comprises a cross-shaped core seat (271), a plurality of silicon steel sheets (274) are arranged in parallel at the side ends of the cross-shaped core seat (271), an air duct (275) is formed between the silicon steel sheets (274), and an electrode (273) is installed on the top of the silicon steel sheet (274).
4. The monitoring and disinfection method of the air pathogen real-time monitoring and disinfection device according to claim 1, characterized in that: The following steps are involved: S1, using an air collection unit disposed inside the replaceable detection chamber (2), and using a second turbofan (26) to introduce air from the area to be detected into the replaceable detection chamber (2) to form an air circulation; S2, using the electrostatic adsorption cell (27) and the second turbofan (26) to vibrate and separate the aerosols and suspended particles in the air, so as to achieve preliminary separation of pathogens and other non-pathogenic particles in the air; S3, a temperature difference incubation unit is also provided in the replaceable detection chamber (2), and the air separated by vibration is passed through the temperature difference incubation unit provided inside the device, so that the pathogen is incubated for a predetermined time under specific temperature conditions to obtain a pathogen signal; S4, the detection circuit (24) is built with a monitoring internal reference unit and a monitoring internal reference unit, and the collected pathogen signals are pre-processed by fuzzy logic in the monitoring internal reference unit, and pathogen signals of different concentrations and types are converted into fuzzy sets using fuzzy rules, and preliminary screening and fuzzy decision making are performed on the pathogen signals; S5, inputting the pathogen signal after fuzzy logic processing into the first-layer generation network of the two-layer generative adversarial network to generate hypothetical pathogen samples, and the first-layer discriminant network discriminates the generated hypothetical pathogen samples, and optimizes the ability to capture pathogen signals through iterative training; S6, inputting the signal output by the first-layer generative adversarial network into the second-layer generative adversarial network to generate a pathogen characteristic signal, and the second discriminant network verifies and optimizes the generated signal; S7, introducing the pathogen signal processed by the double-layer generative adversarial network into the monitoring internal reference unit, reacting the internal reference agent with the pathogen to generate a detectable signal, and capturing and analyzing the pathogen through the detection circuit (24), thereby realizing the monitoring of pathogens in the air; S8. After the monitoring is completed, the air flowing out of the replaceable detection chamber (2) is filtered and disinfected through the filter element (18) to remove the toxic and harmful gases and other pathogens remaining in the air, and is discharged from the air filter base (1) through the filter outlet (13); S9, dynamically adjusting the response parameters of the internal reference agent in the monitoring internal reference unit according to the detected pathogen signal to adapt to pathogens of different types and concentrations; S10. According to the type of pathogen to be monitored, select an appropriate monitoring internal reference agent and set it in the monitoring internal reference unit for targeted monitoring and detection of different pathogens.
5. The monitoring and disinfection method of the air pathogen real-time monitoring and disinfection device according to claim 4 is characterized in that: The S2 comprises the following steps: S21, after the air circulation is formed, the second turbofan (26) is started, and the turbofan body (151) generates a vortex through the arc-shaped blades (153), and the aerosol and suspended particles in the air are concentratedly guided along the direction of the air duct (275); S22, using the electrostatic field generated by the electrostatic adsorption cell (27) to adsorb and separate the passing aerosols and suspended particles using the silicon steel sheet (274) on the cross-shaped core seat (271), the electrostatic field strength is E, E represents the electric field strength generated by the electrostatic adsorption cell (27), and the electric field strength is dynamically adjusted according to the charged characteristics of the aerosols and suspended particles to be separated; S23, when the aerosol and suspended particles pass through the electrostatic adsorption cell (27), the voltage difference V between the electrodes (273) is used to make the particles with different charges adsorbed on the surface of the silicon steel sheet (274) according to the charge polarity of the particles, wherein V is the voltage across the electrodes, and the voltage value is set according to the charge amount of the particles to be separated; S24, through the air duct (275) of the electrostatic adsorption battery (27), the separated aerosol and suspended particles are separated in different directions, the larger pathogenic particles are adsorbed to the silicon steel sheet (274), and the non-pathogenic particles are directed to the filter outlet (13) for subsequent treatment; S25. During the adsorption process of the electrostatic adsorption battery cell (27), the second turbofan (26) and the electrostatic adsorption battery cell (27) are operated synchronously to maintain continuous air flow.
6. The monitoring and disinfection method of the air pathogen real-time monitoring and disinfection device according to claim 4 is characterized in that: The S3 comprises the following steps: S31, after the pathogens and non-pathogenic particles are initially separated, the air is guided to a temperature difference culture unit provided in the replaceable detection chamber (2), the temperature difference culture unit comprising a temperature controller and a culture unit, the temperature controller forms a desired temperature gradient by adjusting the power of a heating element and a cooling element; S32, the pathogens in the air stay in the temperature difference culture unit for a preset time, and the temperature in the culture unit is maintained within the temperature range T that is optimal for the growth of the pathogens by controlling the temperature gradient. opt , T opt Indicates the optimal culture temperature set in the temperature difference culture unit, which is dynamically adjusted according to the growth characteristics of different pathogens; S33, when the pathogens are cultured within a predetermined time, the pathogens in the culture unit are detected by the sensor and the generated pathogen signal S is recorded. signal .
7. The monitoring and disinfection method of the air pathogen real-time monitoring and disinfection device according to claim 4 is characterized in that: The S4 comprises the following steps: S41, in pathogen signal S signal After being generated, the pathogen signal is input into the monitoring internal reference unit built into the detection circuit (24). The monitoring internal reference unit uses the collected pathogen signal S signal Perform preliminary pretreatment to obtain the preliminary pretreated pathogen signal S pre (t); S42, in the pre-processing process, the pathogen signal S is converted into pre (t) According to the concentration and type, it is transformed into a fuzzy set F set , where the fuzzy set F set It is determined according to the fuzzy membership function μ(x) of the pathogen signal, which represents the membership of the signal intensity x at a specific concentration and type: Among them, μ(x) is the membership function of the pathogen signal x, c is the central value of the pathogen signal intensity, σ is the standard deviation of the membership function, and X is the complete set of pathogen signals; S43, for the fuzzy set F set Fuzzy rule reasoning is performed on pathogen signals in the i , the input signals of different fuzzy sets are screened and judged to form fuzzy decision D fuzzy : Among them, D fuzzy As a result of fuzzy decision making, and are the membership functions of pathogen signals in different fuzzy rules, respectively, and x and y are variables of different input signals; S44, based on fuzzy decision D fuzzy , preliminarily screen out pathogen signals that meet the detection criteria, and pass the pathogen signals to the two-layer generative adversarial network unit: Among them, S filtered represents the pathogen signal after screening, γ is the target intensity of the pathogen signal, δ is the width parameter of the signal, and D fuzzy is the fuzzy decision result, and x and y are the input variables of the pathogen signal.
8. The monitoring and disinfection method of the air pathogen real-time monitoring and disinfection device according to claim 4 is characterized in that: The S5 comprises the following steps: S51, the pathogen signal S after fuzzy decision filtered Input to the first-layer generative network of the two-layer generative adversarial network in real time, and generate a hypothetical pathogen sample G1(z) through the first-layer generative network, where G1(z) is the pathogen sample generated by the first-layer generative network in the current time period, and z is the dynamic noise vector. The goal of the first-layer generative network is to adjust the parameters in real time to make the generated hypothetical pathogen sample G1(z) close to the real pathogen sample; S52, inputting the generated hypothetical pathogen sample G1(z) into the first-layer discriminant network in real time, the first-layer discriminant network evaluates the authenticity of the sample by calculating the confidence of the hypothetical pathogen sample, and the output of the first-layer discriminant network is D1(G1(z)), where D1(G1(z)) represents the confidence of the first-layer discriminant network for the hypothetical pathogen sample under the current conditions, and the confidence is calculated based on the difference function between the hypothetical pathogen sample and the real signal collected in real time; S53, through the online iterative training of the first layer generation network and the first layer discrimination network, dynamically optimize the parameters θ of the generation network g and the parameters θ of the first layer of the discriminant network d , to achieve the minimization of the difference function between the hypothetical pathogen sample and the real pathogen sample signal collected in real time in, represents the real-time expected value, p data (x) is the distribution of real pathogen signals collected in real time, p z (z) is the input noise distribution of the first layer of the generative network, θ g and θ d They are the real-time parameter sets of the first-layer generating network and the first-layer discriminative network respectively; S54. Through multiple rounds of online iterations, the optimized first-layer generative network generates hypothetical pathogen samples that are closer to the real pathogen signal. Confidence function of the first layer of discriminant network Towards real-time confidence in true pathogen signals.
9. The monitoring and disinfection method of the air pathogen real-time monitoring and disinfection device according to claim 4 is characterized in that: The S6 comprises the following steps: S61. Generate the pathogen hypothesis signal output by the first layer of the adversarial network Input to the second-layer generation network of the two-layer generation adversarial network in real time. The second-layer generation network generates and processes the input pathogen hypothesis signal to generate a pathogen characteristic signal G2(z′), wherein G2(z′) is the pathogen characteristic signal generated by the second-layer generation network in the current time period, and z' is the corrected noise vector received by the second-layer generation network. The goal of the second-layer generation network is to optimize the authenticity and specificity of the pathogen characteristic signal G2(z′); S62, inputting the generated pathogen characteristic signal G2(z′) into the second-layer discriminant network in real time, and the second-layer discriminant network verifies the signal by calculating the confidence D2(G2(z′)) of the pathogen characteristic signal, wherein D2(G2(z′)) represents the confidence of the second-layer discriminant network in the pathogen characteristic signal under the current conditions, and the confidence is calculated based on the difference function between the pathogen characteristic signal and the expected true pathogen signal; S63, through the online iterative training of the second-layer generation network and the second-layer discriminant network, optimize the parameter θ of the second-layer generation network g2 and the parameters θ of the second layer discriminant network d2 , so that the difference function between the pathogen characteristic signal and the real pathogen signal Minimize: in, represents the real-time expected value, p data (x) is the expected true pathogen signal distribution, p z′ (z′) is the input noise distribution of the second layer generator network, θ g2 and θ d2 They are the real-time parameter sets for the second-layer generating network and the second-layer discriminative network respectively; S64, after multiple rounds of online iterations, the pathogen characteristic signal is finally generated Closer to the real pathogen signal, while completing the real-time pathogen signal capture, the final generated pathogen characteristic signal Transmitted to the disinfection unit.
10. The monitoring and disinfection method of the air pathogen real-time monitoring and disinfection device according to claim 4, characterized in that: The S7 comprises the following steps: S71, the pathogen characteristic signal finally generated The monitoring internal reference unit is introduced in real time. The monitoring internal reference unit reacts specifically with the characteristic signal of the pathogen through the internal reference agent to generate a detectable signal S detect : Among them, α, β, γ and λ are the reaction parameters of the internal reference agent, α represents the initial reaction intensity of the agent, β represents the attenuation coefficient of the reaction, γ is the enhancement factor of the agent, λ represents the attenuation rate of the signal over time, t0 is the initial time point of the reaction, and τ is the integral variable; S72, generating a detectable signal S detect After that, the signal is transmitted to the detection circuit (24) in real time, and the detection circuit amplifies and filters the signal. The processed signal is recorded as S amplified (t): Where η is the amplifier gain, σ is the filter gain factor, δ is the filter time constant, κ is the signal nonlinear correction coefficient, and t1 is the starting time point of signal processing; S73, the amplified signal S is analyzed by the analysis module in the detection circuit (24). amplified (t) is captured and analyzed, and the analysis result is recorded as the pathogen detection value D pathogen (t): Where ω is the signal modulation frequency, ζ· is the nonlinear analysis factor, ρ is the signal modulation amplitude, and t2 is the analysis start time point; S74, according to pathogen detection value D pathogen (t) changes in real time, monitor the internal reference unit to dynamically adjust the reaction parameters of the internal reference agent, including agent concentration, reaction time and temperature, and monitor pathogens of different types and concentrations: Among them, θ up and θ down are the upper and lower thresholds of the detection value change rate.
