A system and method for measuring phosgene content based on network digital technology

Abstract

The invention relates to a method for measuring phosgene content based on network digital technology, comprising S1, preparing a 1,8-naphthalimide ratio fluorescent probe; S2, using the 1,8-naphthalimide ratio fluorescent probe to prepare a phosgene detection test paper loaded with a probe molecule; S3, reacting the phosgene detection test paper with a phosgene gas sample to be measured to obtain a phosgene reaction test paper; S4, photographing the phosgene reaction test paper to obtain a preprocessed image; S5, sending the preprocessed image to an Internet cloud platform through a network; S6, the Internet cloud platform using a computer image analysis method to analyze the preprocessed image to obtain phosgene content data of the phosgene gas sample to be measured; S7, the Internet cloud platform or the Internet cloud server compares the phosgene content data with preset standard data to obtain result information; S8, obtaining the result information of the Internet cloud platform or the Internet cloud server through a mobile network and / or the Internet; the invention has the advantages of low detection cost, convenient operation, high sensitivity and not being easily interfered with.

Description

Technical Field

[0001] The present invention relates to the technical field of environmental safety detection, and in particular to a system and method for measuring phosgene content based on network digitalization technology. Background Art

[0002] Phosgene, also known as carbonyl chloride, has a chemical formula of COCl2. As an important organic intermediate, phosgene has the characteristics of high reactivity, short reaction route for synthesizing downstream products, clean production and less pollution, and plays an important role in the pharmaceutical, pesticide, dye and other industries. However, phosgene is also a highly toxic gas. Inhaling excessive phosgene can cause lung damage or even death.

[0003] Factories involved in the production or use of phosgene will be equipped with relevant phosgene detection devices to monitor the phosgene concentration in the factory area in real time. However, due to the particularity of the device being installed in a fixed position, it is impossible to conduct individualized and differentiated phosgene concentration monitoring for workers in different positions, and it is also impossible to promptly determine the amount of phosgene inhaled by the patient after the danger occurs, thereby delaying the treatment of the disease. At the same time, for operating workers who do not have obvious symptoms of acute poisoning but may be exposed to phosgene, it is also a guarantee for the health and safety of workers to clarify their own cumulative phosgene intake, stop work and rest before reaching the intake warning line, or seek medical treatment in advance. Therefore, how to accurately monitor the individualized and differentiated phosgene intake of workers is a problem that needs to be solved at present.

[0004] In the prior art, methods for detecting phosgene include gas chromatography, colorimetry, electrochemical measurement, fluorescent probe detection, etc. Among them, the fluorescent probe detection method is widely used in phosgene detection due to its advantages of low cost, good selectivity, high sensitivity, and simple operation.

[0005] However, the fluorescent probe detection method in the prior art requires the use of special detection instruments in the laboratory for detection, and cannot achieve portable or movable detection. Summary of the invention

[0006] In order to solve the technical problem that the fluorescent probe detection method in the prior art needs to use a special detection instrument to perform detection in the laboratory, and cannot realize portable detection or mobile detection, the present invention provides a system and method for determining phosgene content based on network digital technology.

[0007] The technical solution of the present invention to solve the above technical problems is as follows:

[0008] A method for determining phosgene content based on network digital technology comprises the following steps:

[0009] S1, preparation of 1,8-naphthalimide ratiometric fluorescent probe;

[0010] S2, using the 1,8-naphthalimide ratio fluorescent probe to prepare a phosgene detection test paper;

[0011] S3, reacting the phosgene detection test paper loaded with the 1,8-naphthaleneimide ratio fluorescent probe with a phosgene gas sample to be tested to obtain a phosgene reaction test paper;

[0012] S4, photographing the phosgene reaction test paper to obtain a preprocessed image;

[0013] S5, sending the pre-processed image to an Internet cloud platform or an Internet cloud server via a mobile network and / or the Internet;

[0014] S6. Analyzing the pre-processed image using a computer image analysis method in the Internet cloud platform or the Internet cloud server to obtain phosgene content data of the phosgene gas sample to be measured;

[0015] S7, using the Internet cloud platform or the Internet cloud server to compare the phosgene content data with preset standard data to obtain result information;

[0016] S8. Obtain the result information in the Internet cloud platform or the Internet cloud server through a mobile network and / or the Internet.

[0017] The beneficial effects of the present invention are as follows: after the present invention takes a photo to obtain the color, the image is processed and analyzed by an Internet cloud platform or an Internet cloud server, and the result and corresponding feedback information are sent to the user, so that the applicable groups are more diversified. The obtained result is more intuitive, the error caused by artificial fluorescence observation is avoided, and the accuracy of the detection result is guaranteed. At the same time, the present invention is applicable to various mobile terminals with shooting functions, such as smart phones. Therefore, the present invention has the advantages of low detection cost, convenient operation, high sensitivity and not easy to be interfered with.

[0018] Based on the above technical solution, the present invention can also be improved as follows.

[0019] Further, the method for preparing 1,8-naphthalimide ratio fluorescent probe comprises the following steps:

[0020] S10, dissolving N-n-butyl-4-bromo-1,8-naphthalimide and ethylenediamine in a 2-methoxyethanol solution to obtain a solution A, wherein the molar ratio of the N-n-butyl-4-bromo-1,8-naphthalimide to the ethylenediamine is in the range of 1:4 to 1:6;

[0021] S11, subjecting the solution A to reflux reaction to obtain solution B;

[0022] S12, extracting the solution B with ethyl acetate to obtain a solution C;

[0023] S13, washing the solution C with saturated brine to obtain the washed solution C;

[0024] S14, drying the washed solution C with anhydrous sodium sulfate to obtain crystal D;

[0025] S15. Recrystallize the crystal D with ethyl acetate to obtain the 1,8-naphthalimide ratio fluorescent probe.

[0026] Further, the method for preparing a phosgene detection test paper using the 1,8-naphthalimide ratio fluorescent probe comprises the following steps:

[0027] S20, adding the polystyrene crystal and the 1,8-naphthaleneimide ratio fluorescent probe into a dichloromethane solution, and stirring evenly to obtain a solution E, wherein the mass ratio of the polystyrene crystal to the 1,8-naphthaleneimide ratio fluorescent probe is in a range of 1000:1 to 1500:1;

[0028] S21, soaking the filter paper in the solution E;

[0029] S22, drying the filter paper soaked in the solution E to obtain the phosgene detection test paper.

[0030] Furthermore, the method of reacting the phosgene detection test paper loaded with the 1,8-naphthaleneimide ratio fluorescent probe with a phosgene gas sample to be tested comprises placing the phosgene detection test paper in the phosgene gas sample to be tested for 3-7 minutes to obtain the phosgene reaction test paper.

[0031] Furthermore, the method of photographing the phosgene reaction test paper to obtain the preprocessed image is specifically to photograph the phosgene reaction test paper under irradiation with a 365nm ultraviolet lamp using a mobile communication terminal with a photographing function to obtain the preprocessed image.

[0032] Further, the method of sending the pre-processed image to an Internet cloud platform or an Internet cloud server via a mobile network and / or the Internet comprises the following steps:

[0033] S40, using a median filtering method to remove the salt and pepper noise in the preprocessed image to obtain a denoised image;

[0034] S41, performing adaptive threshold segmentation on the denoised image to obtain a local threshold image;

[0035] S42. Statistically analyze the color system values of all pixels in the local threshold image, and perform a linear mean calculation on the six color system values that appear the most frequently in the local threshold image, using the number of occurrences corresponding to the six color system values as weights, to obtain the weighted average color system values of the six color system values in the local threshold image;

[0036] S43. Connect the mobile communication terminal with a shooting function to a mobile network and / or the Internet;

[0037] S44. Use the mobile communication terminal to send the weighted average color system values to the Internet cloud platform or the Internet cloud server.

[0038] Furthermore, the computer image analysis method is to input the weighted average color system values into the BP neural network model in the Internet cloud platform or the Internet cloud server.

[0039] Furthermore, the weight and threshold adjustment formula of the BP neural network model is

[0040]

[0041] where t is the number of training times, W is the weight coefficient matrix, X is the input vector, θ is the threshold coefficient matrix, η is the learning rate, δ is the output error, α is the momentum coefficient, and α ∈ (0, 1);

[0042] The adjustment formula of the momentum coefficient α is

[0043]

[0044] where E(t + 1) is the mean square error at the (t + 1)th moment, E(t) is the mean square error at the tth moment. When ΔE < 0, the learning rate η at the tth moment is increased. When ΔE ≥ 0, the learning rate η at the tth moment is decreased. When E(t + 1) ≤ 1.05E or E(t + 1) ≥ E(t), the momentum coefficient α remains unchanged, and its learning rate η also remains unchanged;

[0045] The method of dynamically adjusting the learning rate is used to introduce a proportionality factor η1 to adaptively change the learning rate. The adjustment formula of the learning rate is as follows

[0046]

[0047] where η1 ∈ (0, 1], η(t) is the learning rate at the tth moment. When E(t + 1) < E(t), the learning efficiency is improved by dividing the learning rate at the tth moment by η1. When E(t + 1) ≥ E(t), the learning efficiency is reduced by multiplying the learning rate at the tth moment by η1.

[0048] The beneficial effect of adopting the above further scheme is that the BP neural network is improved. The traditional BP neural network has the problems of being easily trapped in the local minimum and failing to obtain the global optimum, and the convergence speed is slowed down and the learning efficiency is reduced when the number of training times is large. In view of these problems, the present invention has made some improvements to the BP neural network algorithm. The BP algorithm is based on the gradient descent method when adjusting the weights, so it is easy to oscillate and the convergence speed is slowed down. In order to improve the convergence speed of the BP network, the method of increasing the momentum term and adaptively adjusting the learning rate is used to avoid the formation of local minima, and to improve the convergence speed and learning efficiency.

[0049] Based on the above technical solution, the present invention also provides a system for measuring phosgene content based on network digital technology, and the specific solution is as follows:

[0050] A system for measuring phosgene content based on network digital technology, comprising:

[0051] A detection module, used for reacting a test paper loaded with a 1,8-naphthalimide ratio fluorescent probe with a gas to be tested to obtain a phosgene reaction test paper;

[0052] An image acquisition module, used for photographing the phosgene reaction test paper to obtain a pre-processed image, and sending and receiving data via a mobile network and / or the Internet;

[0053] An Internet cloud platform or Internet cloud server is used to receive the pre-processed image sent by the image acquisition module through a mobile network and / or the Internet, and analyze the pre-processed image using a computer image analysis method to obtain the phosgene content data of the gas to be measured, and then compare the phosgene content data with preset standard data to obtain result information, and send the result information to the image acquisition module through a mobile network and / or the Internet.

[0054] Further, the image acquisition module is specifically used to use a median filtering method to remove the salt and pepper noise in the preprocessed image to obtain a denoised image; perform adaptive threshold segmentation on the denoised image to obtain a local threshold image; perform statistical analysis on the color system values ​​of all pixels in the local threshold image, and perform linear mean calculation on the six color system values ​​that appear most frequently in the local threshold image using the number of occurrences corresponding to the six color system values ​​as weights to obtain a weighted average color system value of the six color system values ​​in the local threshold image;

[0055] The Internet cloud platform or Internet cloud server is specifically used to receive the weighted average color system value sent by the image acquisition module through the mobile network and / or the Internet, input the weighted average color system value into the BP neural network model in the Internet cloud platform or the Internet cloud server, obtain the phosgene content data of the gas to be tested, and then compare the phosgene content data with the preset standard data to obtain result information, and send the result information to the image acquisition module through the mobile network and / or the Internet. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a flowchart of the present invention;

[0057] Figure 2 This is a schematic diagram of the preparation of phosgene detection test paper;

[0058] Figure 3 It is a system structure block diagram of the present invention. DETAILED DESCRIPTION

[0059] The principles and features of the present invention are described below in conjunction with the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention.

[0060] The computer professional terms involved in the present invention are explained as follows:

[0061] The emergence of BP neural network is a multi-layer feedforward neural network trained according to the error back propagation algorithm. It is one of the most widely used neural network models. The emergence of BP (Back Propagtion) neural network systematically solves the learning problem of the connection weights of the hidden layer of the multi-layer neural network and gives a complete mathematical derivation. BP neural network is the core part of the forward neural network and the essence of the entire artificial neural network. In actual pattern recognition applications, about 80% of neural network models use BP neural network or its variations. Compared with other classification methods, BP neural network has the advantages of simple model structure and clear mathematical principles of network algorithm.

[0062] like Figures 1 to 2 As shown, this embodiment provides a method for measuring phosgene content based on network digital technology, comprising the following steps:

[0063] S1, preparation of 1,8-naphthalimide ratiometric fluorescent probe;

[0064] S2, using the 1,8-naphthalimide ratio fluorescent probe to prepare a phosgene detection test paper; and placing the test paper loaded with the 1,8-naphthalimide ratio fluorescent probe in a clamping device to make a phosgene detection card;

[0065] S3, reacting the phosgene detection plate loaded with the 1,8-naphthaleneimide ratio fluorescent probe with a phosgene gas sample to be tested to obtain a phosgene reaction test paper;

[0066] S4, remove the phosgene reaction test paper on the phosgene detection card, and obtain the image phosgene reaction test paper through the camera of the smart phone to obtain a pre-processed image;

[0067] S5, sending the pre-processed image to an Internet cloud platform or an Internet cloud server via a mobile network and / or the Internet;

[0068] S6. The Internet cloud platform or the Internet cloud server analyzes the pre-processed image using a computer image analysis method to obtain phosgene content data of the phosgene gas sample to be measured;

[0069] S7, the Internet cloud platform or the Internet cloud server compares the phosgene content data with preset standard data to obtain result information;

[0070] S8. Obtain the result information of the Internet cloud platform or the Internet cloud server through a mobile network and / or the Internet.

[0071] like Figure 2 As shown, the method for preparing 1,8-naphthalimide ratio fluorescent probe comprises the following steps:

[0072] S10, mixing an N-n-butyl-4-bromo-1,8-naphthalimide solution, an ethylenediamine solution and a 2-methoxyethanol solution to obtain a solution A, wherein a molar ratio of the N-n-butyl-4-bromo-1,8-naphthalimide to the ethylenediamine is in a range of 1:4 to 1:6;

[0073] S11, subjecting the solution A to reflux reaction to obtain solution B;

[0074] S12, extracting the solution B with ethyl acetate to obtain a solution C;

[0075] S13, washing the solution C with saturated brine to obtain the washed solution C;

[0076] S14, drying the washed solution C with anhydrous sodium sulfate to obtain crystal D;

[0077] S15. Recrystallize the crystal D with ethyl acetate to obtain the 1,8-naphthalimide ratio fluorescent probe.

[0078] The specific scheme is as follows: Synthesis of fluorescent probe molecules: 200 mg of N-n-butyl-4-bromo-1,8-naphthalimide with a substance amount of 0.72 mmol and 174 mg of ethylenediamine with a substance amount of 2.88 mmol were added to 10 mL of 2-methoxyethanol solvent, and after reflux for 8 hours, the mixture was extracted with ethyl acetate and washed three times with saturated brine. Then, the mixture was dried with anhydrous sodium sulfate and recrystallized with ethyl acetate to obtain an orange target compound, namely 1,8-naphthalimide ratio fluorescent probe crystals, with a yield of 89%.

[0079] Among them, H NMR (600MHz, CDCl3): δ (ppm) = 8.58 (d, J = 6Hz, 1H), 8.46 (d, J = 6Hz, 1H), 8.18 (d, J = 12Hz, 1H), 7.62 (t, J = 9Hz, 1H), 6.70 (d, J = 12Hz, 1H) ,6.18(s,1H),4.16(t,J=9Hz,2H),3.42(t,J=6Hz,2H),3.18(t,J=6Hz,2H),1.74-1.69(m,2H),1.48-1.42(m,2H),0.97(t,J=6Hz,3H).

[0080] C NMR (150MHz, DMSO-d6): δ (ppm) = 164.1, 163.2, 151.1, 134.5, 130.9, 129.7, 1 28.9,124.4,122.1,120.4,107.9,104.1,46.7,40.4,39.3,30.2,20.3,14.2. HR-MS(ESI):calcd for[C18H21N3O2+H]+312.1707, found 312.1706.

[0081] The method for preparing a phosgene detection test paper using the 1,8-naphthalimide ratio fluorescent probe comprises the following steps:

[0082] S20, adding the polystyrene crystal and the 1,8-naphthaleneimide ratio fluorescent probe into a dichloromethane solution, and stirring evenly to obtain a solution E, wherein the mass ratio of the polystyrene crystal to the 1,8-naphthaleneimide ratio fluorescent probe is in a range of 1000:1 to 1500:1;

[0083] S21, soaking the filter paper in the solution E;

[0084] S22, drying the filter paper soaked in the solution E to obtain the phosgene detection test paper.

[0085] The specific configuration is to add 3.0g of polystyrene and 2.0mg of 1,8-naphthalimide ratio fluorescent probe to 60mL of dichloromethane. Stir and disperse thoroughly until a transparent and uniform solution is formed. Cut the filter paper into 1.5×2.0cm size and soak it in the above solution for 10 minutes. Then take out the filter paper and dry it in the air to obtain the test paper for detecting gas phase phosgene.

[0086] Specifically, the method of reacting the phosgene detection test paper loaded with the 1,8-naphthalimide ratio fluorescent probe with a phosgene gas sample to be tested comprises placing the phosgene detection test paper in the phosgene gas sample to be tested for 3-7 minutes to obtain the phosgene reaction test paper.

[0087] Fluorescent probe molecules (compounds) and phosgene in dichloromethane solutions were prepared respectively, and fluorescence detection was performed at λex=385nm in the solution state to obtain a fluorescence spectrum.

[0088] Prepare seven phosgene solutions of different concentrations, wherein the concentrations of the seven phosgene solutions are 1ppm, 3ppm, 5ppm, 10ppm, 15ppm, 20ppm and 30ppm respectively: firstly, the seven phosgene solutions of 1ppm, 3ppm, 5ppm, 10ppm, 15ppm, 20ppm and 30ppm are respectively prepared in dichloromethane according to the concentration gradient of 6.75mM, 20.25mM, 33.75mM, 67.50mM, 101.25mM, 135.00mM and 202.50mM, then a predetermined amount of 10μL of each solution is placed in a 5mL centrifuge tube, and 10μL of the prepared dichloromethane solution containing 0.1% triethylamine is added to each tube. Then, the test piece is placed in the centrifuge tube and the lid is immediately closed. After 5 minutes, the change of the fluorescence color of the test paper is recorded under a 365nm ultraviolet lamp.

[0089] Specifically, the method of photographing the phosgene reaction test paper to obtain the preprocessed image is to photograph the phosgene reaction test paper under irradiation with a 365nm ultraviolet lamp using a mobile communication terminal with a photographing function to obtain the preprocessed image.

[0090] Specifically, the method of sending the pre-processed image to an Internet cloud platform or an Internet cloud server via a mobile network and / or the Internet comprises the following steps:

[0091] S40, using a median filtering method to remove the salt and pepper noise in the preprocessed image to obtain a denoised image;

[0092] S41, performing adaptive threshold segmentation on the denoised image to obtain a local threshold image;

[0093] S42. Statistically analyze the color system values of all pixels in the local threshold image, and perform a linear mean calculation on the six color system values that appear the most frequently in the local threshold image, using the number of occurrences corresponding to the six color system values as weights, to obtain the weighted average color system values of the six color system values in the local threshold image;

[0094] S43. Connect the mobile communication terminal with a photographing function to a mobile network and / or the Internet;

[0095] S44. Use the mobile communication terminal to send the weighted average color system values to the Internet cloud platform or the Internet cloud server. Among them, the computer image analysis method is to input the weighted average color system values into the BP neural network model in the Internet cloud platform or the Internet cloud server.

[0096] Specifically, the weight and threshold adjustment formula of the BP neural network model is as follows:

[0097]

[0098] Among them, t is the number of training times, W is the weight coefficient matrix, X is the input vector, θ is the threshold coefficient matrix, η is the learning rate, δ is the output error, α is the momentum coefficient, and α ∈ (0, 1);

[0099] The adjustment formula of the momentum coefficient α is as follows:

[0100]

[0101] Among them, E(t + 1) is the mean square error at the (t + 1)th moment, E(t) is the mean square error at the tth moment. When ΔE < 0, the learning rate η at the tth moment is increased. When ΔE ≥ 0, the learning rate η at the tth moment is decreased. When E(t + 1) ≤ 1.05E or E(t + 1) ≥ E(t), the momentum coefficient α remains unchanged, and its learning rate η also remains unchanged;

[0102] The method of dynamically adjusting the learning rate is adopted to introduce a proportionality factor η1 to adaptively change the learning rate. The adjustment formula of the learning rate is as follows:

[0103]

[0104] Among them, η1 ∈ (0, 1], η(t) is the learning rate at the tth moment. When E(t + 1) < E(t), the learning efficiency is improved by dividing the learning rate at the tth moment by η1; when E(t + 1) ≥ E(t), the learning efficiency is reduced by multiplying the learning rate at the tth moment by η1.

[0105] Such as Figure 3As shown, based on Example 1, the present invention also provides a system for measuring phosgene content based on network digital technology, comprising:

[0106] A detection module, used for reacting a test paper loaded with a 1,8-naphthalimide ratio fluorescent probe with a gas to be tested to obtain a phosgene reaction test paper;

[0107] An image acquisition module, used for photographing the phosgene reaction test paper to obtain a pre-processed image, and sending and receiving data via a mobile network and / or the Internet;

[0108] An Internet cloud platform or Internet cloud server is used to receive the pre-processed image sent by the image acquisition module through a mobile network and / or the Internet, and analyze the pre-processed image using a computer image analysis method to obtain the phosgene content data of the gas to be measured, and then compare the phosgene content data with preset standard data to obtain result information, and send the result information to the image acquisition module through a mobile network and / or the Internet.

[0109] Specifically, the image acquisition module is specifically used to use a median filtering method to remove salt and pepper noise in the preprocessed image to obtain a denoised image; perform adaptive threshold segmentation on the denoised image to obtain a local threshold image; perform statistical analysis on the color system values ​​of all pixels in the local threshold image, and perform linear mean calculation on the six color system values ​​that appear most frequently in the local threshold image using the number of occurrences corresponding to the six color system values ​​as weights to obtain a weighted average color system value of the six color system values ​​in the local threshold image; the image acquisition module is preferably a mobile communication terminal, such as a smart phone.

[0110] The Internet cloud platform or Internet cloud server is specifically used to receive the weighted average color system value sent by the image acquisition module through the mobile network and / or the Internet, input the weighted average color system value into the BP neural network model in the Internet cloud platform or the Internet cloud server, obtain the phosgene content data of the gas to be tested, and then compare the phosgene content data with the preset standard data to obtain result information, and send the result information to the image acquisition module through the mobile network and / or the Internet.

[0111] Specifically, the image acquisition module is a smartphone capable of taking photos, which is used to capture image data of the fluorescence color change of the ratio fluorescence probe test paper before and after the reaction with phosgene;

[0112] The Internet cloud platform or Internet cloud server includes a computing and analysis unit, a data storage unit, and a data feedback unit.

[0113] The calculation and analysis unit is specifically composed of a small program front-end logic layer and a back-end BP neural network model, which is used to digitally analyze the color of the acquired image to obtain the color information of the image, process the picture data transmitted by the image acquisition unit, analyze the RGB values ​​of all pixels of the image, and input the trained BP neural network model after operations such as median filtering, adaptive threshold segmentation, and feature pixel RGB extraction. The model outputs the phosgene content after calculation.

[0114] The detection result storage unit is used to store the processing results of the data processing unit each time and show the cumulative absorption of phosgene. It is specifically composed of a backend server database and cloud storage, and the development technology framework is spring boot.

[0115] The data feedback unit has a preset cumulative absorption value in the system, uses a comparison module to compare the measured data with the preset standard, and presents the cumulative absorption of phosgene and corresponding feedback to the user based on the comparison result. Its specific structure is the result display page in the applet, and the page display technology is composed of HTML5, CSS and JS.

[0116] The specific steps are as follows:

[0117] Step 1: According to the instructions of the mini program, the user interacts with the mini program, and the mini program obtains a picture of the phosgene reaction test paper by calling the smartphone camera or the smartphone album;

[0118] Step 2: The image is preprocessed in the applet logic layer, including:

[0119] ① Use the median filter algorithm to remove salt and pepper noise in the image.

[0120] ② Adaptive threshold segmentation of images.

[0121] ③ By performing statistical analysis on the RGB values ​​of all pixels in the image, the six RGB values ​​that appear most frequently are linearly averaged using their number of occurrences as weights, and the final weighted average RGB value is used as the feature value of the image.

[0122] Step 3: The image’s feature values ​​and other data are uploaded to the cloud-based mini program backend. After its feature values ​​are input into the BP neural network model, the backend returns the analysis results to the mini program frontend.

[0123] Step 4: Display the results returned by the backend to the user, who decides to save the results or retest. Among them, at different phosgene cumulative absorption amounts, the test results will be recorded through the test result storage unit and presented in the form of a line graph in the result display interface.

[0124] Example 3

[0125] Based on the above-mentioned Example 1 and Example 2, this embodiment provides a small program operation process of a portable network digital phosgene content detection system based on 1,8-naphthalimide ratio fluorescence probe.

[0126] The specific operation process of the smartphone applet is as follows:

[0127] 1) Open the applet page on your smartphone;

[0128] 2) Click on Phosgene Detection and authorize to log in to the mini program as WeChat;

[0129] 3) Enter the camera interface of the smartphone as an image acquisition unit;

[0130] 4) Obtaining image data of the fluorescence color change of the test paper loaded with ratio fluorescent probe before and after the reaction with different concentrations of phosgene;

[0131] 5) The image is preprocessed in the applet logic layer, including: using the median filter algorithm to remove salt and pepper noise in the image, adaptive threshold segmentation of the image, and statistical analysis of the RGB values ​​of all pixels in the image. The six RGB values ​​with the most occurrences are linearly averaged using their occurrence times as weights, and the final weighted average RGB value is used as the feature value of the image;

[0132] 6) Use the comparison module to compare the measured data with the system preset cumulative absorption value standard, and present the cumulative absorption of phosgene and corresponding feedback to the user based on the comparison result. At the same time, the test results will be recorded through the result storage unit and presented in the form of a line graph in the result display interface. The portable network digital phosgene content detection system based on 1,8-naphthalimide ratio-type fluorescent probe provided by the present invention can realize rapid and accurate detection of toxic phosgene using a smart phone platform.

[0133] The present invention uses the traditional BP neural network and the improved BP neural network in the present invention to perform 1000 training and learning on 4500 sets of training samples. The final MSE of the traditional BP neural network is lower than the pre-set MSE. After 550 iterations of the improved BP neural network, the MSE approaches the expected error setting value. Under the same training error, the improved BP neural network has the least number of iterations and requires a shorter training time. Therefore, the improved BP neural network converges faster.

[0134] Sensitivity Se, specificity Sp and accuracy Acc are used as indicators to evaluate the classification effect of test strip images. The higher the three indicators are, the better the classification effect is. The calculation formula is:

[0135] SE=TP / (TP+FN)

[0136] SN=TN / (TN+FN)

[0137] Acc=(TN+FN) / (TN+TP+FN+FP)

[0138] Among them, TP, FP, TN and FN represent the average number of correctly identified samples, the average number of incorrectly identified samples, the average number of correctly rejected samples and the average number of incorrectly rejected samples respectively. The trained BP neural network and the improved BP neural network were used to classify and identify 2500 test set samples multiple times. The classification results are compared as shown in the following table:

[0139]

[0140]

[0141] It can be seen from the above table that the improved BP neural network algorithm has a significant classification effect in the classification of test strip images.

[0142] The present invention has the advantages of low detection cost, convenient operation, high sensitivity and not susceptible to interference. The present invention uses a small program to take pictures and pick colors, and sends the results and corresponding feedback to users after simple data processing, making the applicable groups more diversified. The results are more intuitive, avoiding the errors caused by artificial fluorescence observation, and ensuring the accuracy of the detection results.

[0143] The present invention improves the BP neural network. The traditional BP neural network has the problems of being easily trapped in the local minimum and failing to obtain the global optimal value, and the convergence speed will be slowed down and the learning efficiency will be reduced when the number of training times is large. In view of these problems, the present invention makes some improvements to the BP neural network algorithm. The BP algorithm is based on the gradient descent method when adjusting the weights, so it is easy to oscillate and the convergence speed is slowed down. In order to improve the convergence speed of the BP network, the method of increasing the momentum term and adaptively adjusting the learning rate is used to avoid the formation of local minima, and to improve the convergence speed and learning efficiency.

[0144] The improved BP neural network algorithm provided by the present invention improves the convergence speed and accuracy of the neural network and shortens the training time. In an experiment with a sample size of 2500, the classification accuracy rate reaches 99.67%.

[0145] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the concept and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for determining phosgene content based on network digital technology, characterized in that: The following steps are included: S1, preparation of 1,8-naphthalimide ratiometric fluorescent probe; S2, using the 1,8-naphthalimide ratio fluorescent probe to prepare a phosgene detection test paper; S3, reacting the phosgene detection test paper loaded with the 1,8-naphthaleneimide ratio fluorescent probe with a phosgene gas sample to be tested to obtain a phosgene reaction test paper; S4, photographing the phosgene reaction test paper to obtain a preprocessed image; S5, sending the pre-processed image to an Internet cloud platform or an Internet cloud server via a mobile network and / or the Internet; S6. Analyzing the pre-processed image using a computer image analysis method in the Internet cloud platform or the Internet cloud server to obtain phosgene content data of the phosgene gas sample to be measured; S7, using the Internet cloud platform or the Internet cloud server to compare the phosgene content data with preset standard data to obtain result information; S8. Obtain the result information in the Internet cloud platform or the Internet cloud server through a mobile network and / or the Internet; The method for preparing 1,8-naphthalimide ratio fluorescent probe comprises the following steps: S10, dissolving N-n-butyl-4-bromo-1,8-naphthalimide and ethylenediamine in a 2-methoxyethanol solution to obtain a solution A, wherein the molar ratio of the N-n-butyl-4-bromo-1,8-naphthalimide to the ethylenediamine is in the range of 1:4 to 1:6; S11, subjecting the solution A to reflux reaction to obtain solution B; S12, extracting the solution B with ethyl acetate to obtain a solution C; S13, washing the solution C with saturated brine to obtain the washed solution C; S14, drying the washed solution C with anhydrous sodium sulfate to obtain crystal D; S15, recrystallizing the crystal D with ethyl acetate to obtain the 1,8-naphthalimide ratio fluorescent probe; The method of sending the pre-processed image to an Internet cloud platform or an Internet cloud server via a mobile network and / or the Internet comprises the following steps: S40, using a median filtering method to remove the salt and pepper noise in the preprocessed image to obtain a denoised image; S41, performing adaptive threshold segmentation on the denoised image to obtain a local threshold image; S42, performing statistical analysis on the color system values ​​of all pixels in the local threshold image, and performing linear mean calculation on the six color system values ​​that appear most frequently in the local threshold image, respectively using the number of occurrences corresponding to the six color system values ​​as weights, to obtain a weighted average color system value of the six color system values ​​in the local threshold image; S43, connecting the mobile communication terminal with a photographing function to a mobile network and / or the Internet; S44, using the mobile communication terminal to send the weighted average color system value to the Internet cloud platform or the Internet cloud server; The computer image analysis method is to input the weighted average color system value into the BP neural network model in the Internet cloud platform or the Internet cloud server; The weight and threshold adjustment formula of the BP neural network model is as follows: where t is the number of training times, W is the weight coefficient matrix, X is the input vector, θ is the threshold coefficient matrix, η is the learning rate, δ is the output error, α is the momentum coefficient, and α ∈ (0, 1); The adjustment formula of the momentum coefficient α is as follows: where E(t + 1) is the mean square error at time (t + 1), E(t) is the mean square error at time t. When ΔE < 0, the learning rate η at time t is increased; when ΔE ≥ 0, the learning rate η at time t is decreased; The method of dynamically adjusting the learning rate introduces a proportionality factor η1 to adaptively change the learning rate. The adjustment formula of the learning rate is as follows: where η1 ∈ (0, 1], η(t) is the learning rate at time t. When E(t + 1) < E(t), the learning efficiency is improved by dividing the learning rate at time t by η1; when E(t + 1) ≥ E(t), the learning efficiency is reduced by multiplying the learning rate at time t by η1.

2. The method for determining phosgene content based on network digital technology according to claim 1, characterized in that: The method for preparing a phosgene detection test strip using the 1,8-naphthalimide ratio fluorescent probe includes the following steps: S20. Add polystyrene crystals and the 1,8-naphthalimide ratio fluorescent probe to dichloromethane solution, and stir evenly to obtain solution E. Among them, the mass ratio range of the polystyrene crystals to the 1,8-naphthalimide ratio fluorescent probe is 1000:1 to 1500:1; S21. Immerse filter paper in solution E; S22. After drying the filter paper soaked in solution E, obtain the phosgene detection test strip.

3. The method for determining phosgene content based on network digital technology according to claim 1, characterized in that: The method for reacting the phosgene detection test strip loaded with the 1,8-naphthalimide ratio fluorescent probe with a phosgene gas sample to be detected includes placing the phosgene detection test strip in the phosgene gas sample to be detected for 3 - 7 minutes to obtain the phosgene reaction test strip.

4. The method for determining phosgene content based on network digital technology according to claim 1, characterized in that: The method for photographing the phosgene reaction test strip to obtain a preprocessed image is specifically to photograph the phosgene reaction test strip under the irradiation of a 365 nm ultraviolet lamp using a mobile communication terminal with a photographing function to obtain the preprocessed image.

5. A system for measuring phosgene content based on network digital technology, wherein the phosgene content is measured by the method according to any one of claims 1 to 4, characterized in that: Including: A detection module for reacting a detection test strip loaded with a 1,8-naphthalimide ratio fluorescent probe with a gas to be detected to obtain a phosgene reaction test strip; An image acquisition module for photographing the phosgene reaction test strip to obtain a preprocessed image, and sending and receiving data through a mobile network and / or the Internet; An Internet cloud platform or an Internet cloud server for receiving the preprocessed image sent by the image acquisition module through a mobile network and / or the Internet, analyzing the preprocessed image using a computer image analysis method to obtain the phosgene content data of the gas to be detected, comparing the phosgene content data with preset standard data to obtain result information, and sending the result information to the image acquisition module through a mobile network and / or the Internet.

6. The system for measuring phosgene content based on network digital technology according to claim 5, characterized in that: The image acquisition module is specifically used to use a median filtering method to remove the salt and pepper noise in the preprocessed image to obtain a denoised image; perform adaptive threshold segmentation on the denoised image to obtain a local threshold image; perform statistical analysis on the color system values ​​of all pixels in the local threshold image, and perform linear mean calculation on the six color system values ​​that appear most frequently in the local threshold image, using the number of occurrences corresponding to the six color system values ​​as weights, to obtain a weighted average color system value of the six color system values ​​in the local threshold image; The Internet cloud platform or Internet cloud server is specifically used to receive the weighted average color system value sent by the image acquisition module through the mobile network and / or the Internet, input the weighted average color system value into the BP neural network model in the Internet cloud platform or the Internet cloud server, obtain the phosgene content data of the gas to be tested, and then compare the phosgene content data with the preset standard data to obtain result information, and send the result information to the image acquisition module through the mobile network and / or the Internet.

Images