CN-115116547-B - Construction method of air microorganism real-time monitoring model

Abstract

The invention discloses a construction method of a real-time air microorganism monitoring model, which comprises the following steps of collecting plankton in air of a place to be monitored, calculating plankton concentration, monitoring and collecting air environment quality data of the place to be monitored in the same time period, collecting plankton concentration and air environment quality data of the place to be monitored in different time periods to form a multi-dimensional data table, analyzing the correlation between plankton concentration and air environment quality data of the place to be monitored, selecting air environment quality with the best correlation with plankton concentration in air as a correlation parameter, modeling by using a Gaussian mixture model, optimizing the K value of the Gaussian mixture model to obtain a final Gaussian mixture model, and selecting the long axis of a corresponding ellipse as a construction line segment of a piecewise function to obtain a continuous piecewise linear function, namely the real-time air microorganism monitoring model. The invention has higher accuracy, low cost and easy implementation and popularization.

Inventors

• XIA PENG
• WANG JIANGBIN
• ZHONG XISHENG
• CHENG YUANHUO
• GU DAPENG
• CHEN HONGYIN

Assignees

• 中科爱芯微环境信息科技(深圳)有限公司
• 中科爱芯微环境信息科技(深圳)有限公司

Dates

Publication Date

20260421

Application Date

20220805

Priority Date

20220805

Claims (6)

1. 1. The method for constructing the air microorganism real-time monitoring model is characterized by comprising the following steps of: Collecting plankton in the air of a place to be monitored, and calculating by adopting a culture counting method to obtain the concentration of plankton in the air of the place to be monitored; Monitoring and collecting air environment quality data in the air of a place to be monitored in the same time period, wherein the air environment quality data is one or a combination of more than two of particle concentration, carbon dioxide concentration, temperature data and humidity data; step (3), repeating the step (1) and the step (2), and collecting the concentration of planktonic bacteria in the air and the air environment quality data of the to-be-monitored place in different time periods to form a multi-dimensional data table; step (4), analyzing the correlation between the concentration of plankton in the air of the place to be monitored and the air environment quality data according to the multidimensional data table obtained in the step (3); Step (5), according to the correlation analysis data result of the step (4), selecting air environment quality data with higher correlation with concentration of plankton in the air as vectors in a Gaussian mixture model, and modeling by using the Gaussian mixture model to obtain a Gaussian mixture model with a K value greater than or equal to 2, wherein the K value represents the number of ellipses in the Gaussian mixture model, and the K value is a positive integer; step (6), optimizing the K value of the Gaussian mixture model to obtain a final Gaussian mixture model; step (7), selecting a long axis of a corresponding ellipse as a forming line segment of the piecewise function, wherein the slope of the long axis is smaller than or equal to 1, so as to obtain a continuous piecewise linear function, namely a real-time air microorganism monitoring model; In the step (5), when modeling the Gaussian mixture model, the probability function is calculated as follows: (equation 2); In the formula 2, x is air environment quality data, p (x) is a weighted sum of k p (x|mu, sigma), k is the number of components, namely the number corresponds to the number of ellipses, alpha is a weight coefficient of each component, and p (x|mu, sigma) is a probability density function of an n-dimensional random vector x obeying Gaussian distribution; Thus, the first and second substrates are bonded together, (Equation 3); Equation 3 is a multidimensional gaussian distribution function, n is a vector of x, μ is a desired value μ=e (x) of x, T represents a transpose, Σ is a covariance matrix, and in equation 3 Σ=cov (x) =e { (x- μ) (x- μ) T }; The specific method of the step (6) is as follows: setting the K value of the Gaussian mixture model as K, wherein K is more than or equal to 2 and less than or equal to n/4, and K is a positive integer; step (6-2), carrying out Gaussian mixture modeling according to the value of k, and calculating the slope of a long axis by using two focuses of the obtained k ellipses to obtain k linear equations; Substituting the original data of the air environment quality into the linear equation obtained in the step (6-2), and calculating to obtain the calculated concentration of the plankton; Setting the K value of the Gaussian mixture model as k+1, and repeating the step (6-2) and the step (6-3) to obtain the sum of the calculated concentration of the plankton and the absolute value of the error of the actual measured concentration of the plankton corresponding to the calculated concentration of the plankton when the K value is k+1, namely T k+1 ; Setting the K value of the Gaussian mixture model as k+2, repeating the step (6-2) and the step (6-3) to obtain T k+2 , setting the K value of the Gaussian mixture model as n/4, rounding n/4, repeating the step (6-2) and the step (6-3) to obtain T n/4 , comparing the sizes of T k 、T k+1 、T k+2 ……T n/4 , and determining the K value corresponding to the minimum sum of the absolute values of the errors as the optimal K value of the Gaussian mixture model.
2. 2. The method for constructing the air microorganism real-time monitoring model according to claim 1 is characterized in that in the step (1), a six-level anderson impact sampler is utilized to collect plankton in air at a place to be monitored, then six dishes of the six-level anderson sampler are cultivated according to a national standard method, the colony numbers of the six dishes are recorded respectively, and the plankton concentration in the air at the place to be monitored is calculated.
3. 3. The method for constructing an airborne microorganism real-time monitoring model according to claim 1, wherein in the step (2), air environmental quality data in the air of the site to be monitored is monitored and collected by using an air quality monitor.
4. 4. The method for constructing an air microorganism real-time monitoring model according to claim 3, wherein the air quality monitor is one or a combination of more than two of a laser particle counting sensor, a carbon dioxide sensor, a temperature sensor and a humidity sensor.
5. 5. The method of claim 1, wherein in the step (3), the total number of the data sets of the multidimensional data is n, and n is a positive integer and greater than or equal to 45.
6. 6. The method for constructing an air microorganism real-time monitoring model according to claim 5, wherein in the step (4), the correlation is represented by a correlation coefficient r, and the calculation formula of r is as follows: (equation 1); In the formula 1, x i is the total number of colonies sampled for the ith time, and y i is air environment quality data; Average of the total number of n colonies; Is the average of n sets of air environment quality data.

Description

Construction method of air microorganism real-time monitoring model Technical Field The invention relates to the technical field of microorganism monitoring. In particular to a method for constructing an air microorganism real-time monitoring model. Background Airborne microorganisms refer to microorganisms that are present in the air. The presence of airborne microorganisms and their harmfulness are confirmed by on-site sampling tests and epidemiological investigation. The contamination level is often characterized in public place health monitoring by the total number of bacteria in the air. The monitoring technology of air microorganism is developed in various ways, but the following main values are provided: 1. Culture counting technology, which is a technology for hundreds of years to sample microorganisms in the air or on the surface of objects, is also the most reliable method for identifying pathogens. The natural sedimentation method is the most traditional culture sampling method, namely, a sampling plate is cultured for 12-48 hours after being exposed in the environment for 10 minutes, grown colonies are only identified and counted by naked eyes, and the pollution degree of environmental microorganisms can be evaluated by referring to corresponding evaluation standards. The method is simple and convenient, but has long operation time and low sampling efficiency, and cannot accurately reflect the pollution level. The use of an automatic sampling system and a multi-purpose microbiological sampler overcomes this disadvantage, as is typical of devices such as the anderson 6-stage sampler. The air sampling culture technique has the advantages that the detection result can be reported only after sampling, the timeliness is poor, and common airborne microorganisms can be collected and obtained at the same time, so that the method is favorable for establishing an archive record of indoor air quality or microorganism concentration in the air, and is favorable for timely taking protective and remedial measures. 2. The laser scattering particle counting technology is to identify the physical parameters of air particles by utilizing photoelectric technology and develop a particle counter based on the technology. The microbial aerosol is distributed in a biased state in the air, most of the environment is concentrated in a range of 5-10 microns, the particle counter can identify the particle size in the range of 0.3-20 microns by utilizing the laser scattering principle, and the particle size range comprises most bacteria and all spores, so that the microbial aerosol is a feasible alternative method for detecting airborne pathogens. The particle counter is low-cost aerosol monitoring equipment and can be used for monitoring and alarming the microorganism aerosol in special environments, such as the sudden and obvious change of the concentration of the environmental particles when the biological warfare agent attacks. The disadvantage of this technique is that only the total number of air particles can be identified, and the lack of specificity. 3. Laser-induced fluorescent biological particle counting technique the laser-induced fluorescent biological particle counting technique can be used for characteristic identification of biological characteristics of particles. The physical parameters of particle size are identified by the light scattering principle, the biological intrinsic characteristics of the microorganism particles are identified by ultraviolet excited biological fluorescence, and the characteristic counting of the microorganism concentration is realized by the laser-induced fluorescence detection technology at the two-dimensional level of morphological and biological parameters. A high-power laser or a xenon lamp is used as an ultraviolet excitation light source, mobile biological warfare agent detection alarm systems (TSl 3312, BAWS, veroTect and the like) under field conditions are developed in countries such as America, english and the like, a UV-LED is used as an ultraviolet excitation light source, a TSI company miniaturizes and portability the biological particle counting technical device, and the biological safety monitoring alarm in a closed environment building is realized, but because non-microorganism particles such as cigarettes, pollen and the like in the actual environment can also generate fluorescence under the induction of ultraviolet laser, the technology is easily interfered by the non-microorganism particles such as cigarettes, pollen and the like, and meanwhile, even the miniaturized detection device is still more than hundred thousand RMB in price, so that the technology is unfavorable for market popularization. 4. Other techniques, mass spectrometry, which measure the spectrum of a compound or organic compound with variable wavelengths, can be used for detection of a class of pathogens by identifying several primary, or a group of compounds, of a microorganism, but primarily