Foam Fire Extinguishing Agent Quality Detection Method, System and Storage Medium

Abstract

The present invention discloses a method, a system and a storage medium for detecting the quality of a foam fire extinguishing agent, comprising the following steps: S1, selecting a plurality of qualified samples and samples to be tested of the foam fire extinguishing agent, collecting their near-infrared absorption spectra and processing them to obtain preprocessed spectra; S2, calculating the overall absorbance difference of all samples according to the preprocessed spectra, and using the Grubbs test method with the overall absorbance difference as the benchmark to screen out the qualified samples that meet the requirements, and using their preprocessed spectra as a verification spectrum set; S3, separately adding the near-infrared absorption spectrum of each sample to be tested into the verification spectrum set, and calculating the Grubbs statistic of the sample to be tested; S4, calculating the matching degree between the sample to be tested and the qualified samples according to the maximum Grubbs statistic of the qualified samples screened in step S2 and the Grubbs statistic of the sample to be tested in step S3, and determining that the quality of the sample to be tested is qualified if the set conditions are met; the present invention can quickly and accurately detect whether the quality of the foam fire extinguishing agent is qualified.

Description

Technical Field

[0001] The present invention relates to the quality detection of foam fire extinguishing agents, in particular to a method, a system and a storage medium for detecting the quality of foam fire extinguishing agents. Background Art

[0002] The fixed fire extinguishing system is the last safety barrier for large oil-filled equipment such as transformers. The quality of the foam fire extinguishing agent used in the fire extinguishing system directly determines the fire extinguishing efficiency of the system, which is crucial for the operation safety of substations (converter stations). However, the national standard detection method for foam fire extinguishing agents generally requires a detection cycle of 2 months, which is difficult to match the on-site engineering progress. There is even a situation where the fire extinguishing agent is in service with problems. At the same time, the mechanism of performance degradation of foam fire extinguishing agents is not yet clear, and environmental factors such as ultraviolet rays, high temperature, and humidity often cause the standby fire extinguishing agent to fail in advance.

[0003] Although there has been research on using near-infrared spectroscopy technology to detect the quality of fire extinguishing agents, the difference in the quality of foam fire extinguishing agents in the near-infrared spectrum is small. Directly judging whether a foam fire extinguishing agent is qualified through the difference in the near-infrared spectrum often easily leads to many unqualified fire extinguishing agent samples being misjudged as qualified samples, thus causing uncontrollable hazards. Summary of the Invention

[0004] Object of the Invention: The object of the present invention is to provide a method, a system and a storage medium for detecting the quality of foam fire extinguishing agents, which can quickly and accurately detect whether the quality of the foam fire extinguishing agent is qualified.

[0005] Technical Solution: The method for detecting the quality of foam fire extinguishing agents according to the present invention includes the following steps:

[0006] S1. Select a plurality of qualified samples and samples to be tested of the foam fire extinguishing agent, collect their near-infrared absorption spectra and perform smoothing processing to obtain preprocessed spectra;

[0007] S2. Calculate the overall absorbance difference of all samples according to the preprocessed spectra, and use the Grubbs test method to screen out the qualified samples that meet the requirements based on the overall absorbance difference, and use their preprocessed spectra as the verification spectrum set;

[0008] S3. Individually add the near-infrared absorption spectra of each sample to be tested to the verification spectrum set to calculate their overall absorbance difference, and calculate the Grubbs statistic of the sample to be tested using the average value and standard deviation of the overall absorbance difference of all qualified samples in the original verification spectrum set;

[0009] S4. Calculate the matching degree between the sample to be tested and the qualified samples according to the maximum Grubbs statistic of the qualified samples after screening in step S2 and the Grubbs statistic of the sample to be tested in step S3, and determine the sample to be tested with the matching degree meeting the set conditions as having qualified quality.

[0010] Based on the above technical method, in steps S1 and S2, the near-infrared absorption spectra of the foam fire extinguishing agent samples are collected and the overall difference in absorbance of all samples is calculated. The overall difference in absorbance can reflect the fluctuation of the absorbance of the samples at each wavelength, which is equivalent to a characteristic value of each sample's spectrum. Using the Grubbs test method with the overall difference in absorbance as the benchmark, qualified samples that meet the requirements are selected, and the qualified samples with significantly abnormal overall differences in absorbance among the qualified samples can be excluded. This can ensure that the overall differences in absorbance, that is, the spectral characteristics, of all qualified samples corresponding to the verification spectral set are within a set reasonable range, which also ensures that the quality of these qualified samples does not vary much. Only in this way can the quality of the test sample be determined based on the deviation degree between the test sample and all qualified samples in the verification spectral set. In step S3, each test sample is separately added to the verification spectral set as a new whole, and the overall difference in absorbance of the test sample in this new whole is calculated. By adding the average value and standard deviation of the overall differences in absorbance of all the original qualified samples in the verification spectral set, the Grubbs statistic of the test sample is calculated. This statistic reflects the E of the test sample. j The degree of deviation from the average value of the overall differences in absorbance of all qualified samples in the verification spectral set, which is also equivalent to the degree of deviation of the spectral characteristics of the test sample from the average value of the spectral characteristics of all qualified samples in the verification spectral set. And the spectral characteristics correspond to the quality of the foam fire extinguishing agent, which also reflects the degree of deviation between the quality of the test sample and the average value of the quality of the qualified samples. Similarly, the Grubbs statistic of each qualified sample in the verification spectral set actually also reflects the degree of deviation between the quality of this qualified sample and the average value of the quality of all qualified samples.

[0011] In the whole process of the above method, only by collecting the near-infrared absorption spectra of the samples and performing relevant calculations can it be judged whether the quality of the test sample is qualified. The process is fast and simple, without the need for a long wait. Moreover, it does not directly compare the differences in the spectra of the samples, but calculates the Grubbs statistics of the qualified samples and the test samples and then calculates the matching degree between the two, and uses the matching degree to judge whether the quality of the test sample is qualified. Compared with the direct visual comparison of the spectra, it is more accurate and reliable. Because simply visually comparing the spectra cannot make an effective difference judgment as a whole and is prone to missing small differences and misjudgments. Generally speaking, this method can quickly and accurately detect whether the quality of the foam fire extinguishing agent is qualified.

[0012] Preferably, the near-infrared absorption spectra of each sample in step S1 are collected multiple times, and the collected spectra are averaged and then smoothed to obtain the preprocessed spectrum of the sample.

[0013] Collecting the near-infrared absorption spectra of each sample multiple times and then averaging them to obtain a spectrum can reduce the errors existing in single collection through averaging compared with single collection.

[0014] Preferably, in step S1, the Savitzky-Golay filter, Adjacent Averaging algorithm or FFT Filtering algorithm is used to smooth the near-infrared absorption spectrum.

[0015] Preferably, the overall absorbance difference in step S2 is calculated according to the following formula

[0016]

[0017] where is the overall absorbance difference of the i-th sample, represents the absorbance value of the i-th sample at the k-th wavelength, is for calculating the average absorbance value of all samples in the sample set used for calculating

[0018] at the k-th wavelength, and n is the total number of wavelengths of the light used for measurement.

[0019] S2.1. Calculate the Grubbs statistic of the qualified samples according to the following formula

[0020]

[0021] where is the Grubbs statistic of the i-th qualified sample, is the overall absorbance difference of the i-th qualified sample in the set composed of all qualified samples, is the average value of the overall absorbance differences of all qualified samples, is the standard deviation of the overall absorbance differences of all qualified samples;

[0022] S2.2. Confirm whether there is greater than the critical value , if so, remove these qualified samples and return to step S2.1 until there is no greater than the critical value .

[0023] Preferably, the Grubbs statistic of the sample to be tested in step S3 is calculated according to the following formula

[0024]

[0025] where is the Grubbs statistic of the j-th sample to be tested, is the overall absorbance difference of the j-th sample to be tested in the validation spectral set

[0026] Preferably, the matching degree between the preprocessed spectrum of the sample to be tested and the validation spectral set in step S4 is calculated according to the following formula

[0027]

[0028] where is the matching degree between the preprocessed spectrum of the j-th sample to be tested and the validation spectral set, ;

[0029] When it is determined that the sample to be tested is unqualified, and when it is determined that the sample to be tested needs further testing, and when it is determined that the sample to be tested is qualified

[0030] Preferably, the material of the cuvette used for spectral collection in step S1 is quartz or optical glass

[0031] Preferably, the spectral collection wavelength range in step S1 is 800~3300nm

[0032] The quality detection system for foam fire extinguishing agent of the present invention includes

[0033] Spectral collection and processing module: used to collect the near-infrared absorption spectra of multiple qualified samples and samples to be tested of the foam fire extinguishing agent and perform smoothing processing to obtain preprocessed spectra

[0034] Spectral screening module: used to calculate the overall absorbance difference of all samples according to the preprocessed spectra, and use the Grubbs test method based on the overall absorbance difference to screen out the qualified samples that meet the requirements and use their preprocessed spectra as the validation spectral set

[0035] Grubbs statistic calculation module for samples to be tested: used to calculate the overall absorbance difference of each sample to be tested by adding it to the validation spectral set alone, and calculate the Grubbs statistic of the sample to be tested using the average value and standard deviation of the overall absorbance differences of all qualified samples in the original validation spectral set

[0036] Detection module: It is used to calculate the matching degree between the maximum Grubbs statistic of the qualified samples after screening by the spectral screening module and the Grubbs statistic of the sample to be tested in step S3, and determine that the sample to be tested with the matching degree meeting the set conditions is qualified in quality.

[0037] The computer-readable storage medium storing one or more programs according to the present invention, wherein the one or more programs include instructions that, when executed by a computing device, cause the computing device to execute any one of the above methods.

[0038] Beneficial effects: Compared with the prior art, the significant effects of the present invention are as follows: By collecting the near-infrared absorption spectrum of the sample and calculating the overall difference in absorbance of the sample, which is equivalent to extracting the characteristic value of the overall spectrum, and then screening the sample and calculating the matching degree based on the overall difference to determine whether the quality of the sample to be tested is qualified. The whole process is fast and simple, and compared with directly comparing the spectrograms, the result determined by the overall characteristics extracted from the spectrum is more accurate. Description of the Drawings

[0039] Figure 1 It is a spectral comparison diagram of the sample to be tested and the qualified samples after screening in Example 1;

[0040] Figure 2 It is a spectral comparison diagram of the sample to be tested and the qualified samples after screening in Example 2. Detailed Embodiments

[0041] As shown in the figure, the foam fire extinguishing agent quality detection method described in the present invention includes the following steps:

[0042] S1. Select multiple qualified samples and samples to be tested of the foam fire extinguishing agent, collect their near-infrared absorption spectra and perform smoothing processing to obtain preprocessed spectra;

[0043] The cuvette material used for collecting the spectrum is quartz or optical glass, and the spectral collection wavelength range is 800~3300nm. The wavelength range does not affect the detection effect of this method, and other wavelength ranges can be used for collection according to actual needs.

[0044] The near-infrared absorption spectrum of each sample is collected multiple times, and the spectra collected multiple times are averaged and then smoothed to obtain the preprocessed spectrum of the sample; The Savitzky-Golay filter, Adjacent Averaging algorithm or FFT Filtering algorithm is used to smooth the near-infrared absorption spectrum.

[0045] S2. Calculate the overall difference in absorbance of all samples according to the preprocessed spectra, and use the Grubbs test method based on the overall difference in absorbance to screen out the qualified samples that meet the requirements, and use their preprocessed spectra as the verification spectrum set;

[0046] The overall absorbance difference is calculated according to the following formula

[0047]

[0048] where is the overall absorbance difference of the i-th sample, represents the absorbance value of the i-th sample at the k-th wavelength, is for calculating the average absorbance value of all samples in the sample set used for calculation at the k-th wavelength, and n is the total number of wavelengths of the light used for measurement; here, the sample set used for calculating can only contain qualified samples, or can be a sample combination of qualified samples and samples to be tested mixed together.

[0049] The overall absorbance difference reflects the overall deviation degree of the absorbance of a certain sample from the average absorbance of all other samples under the light of all wavelengths. It is equivalent to a characteristic value of the near-infrared absorption spectrum of a certain sample, and numerically reflects the overall characteristics of the spectrum.

[0050] Using the Grubbs test method to screen out qualified samples that meet the requirements based on the overall absorbance difference includes the following sub-steps

[0051] S2.1. Calculate the Grubbs statistic of the qualified samples according to the following formula

[0052]

[0053] where is the Grubbs statistic of the i-th qualified sample, is the overall absorbance difference of the i-th qualified sample in the set composed of all qualified samples. Here, the sample set used for calculating the overall absorbance difference is the set composed of all qualified samples; is the average value of the overall absorbance differences of all qualified samples, is the standard deviation of the overall absorbance differences of all qualified samples;

[0054] S2.2. Confirm whether there is greater than the critical value . If so, eliminate these qualified samples and return to step S2.1 until there is no greater than the critical value . After returning to step S2.1, the sample set used for calculating the overall absorbance difference should remove those qualified samples that have been eliminated to form a new sample set for calculation, including and are all calculated with the new sample set.

[0055] The near-infrared absorption spectra of the qualified samples screened through steps S2.1 and S2.2 are constructed into a validation spectral set.

[0056] S3. Individually add the near-infrared absorption spectrum of each sample to be tested to the validation spectral set to calculate the overall absorbance difference. Use the average value and standard deviation of the overall absorbance differences of all qualified samples in the original validation spectral set to calculate the Grubbs statistic of the sample to be tested;

[0057] To better illustrate step S3, assume that individually adding the near-infrared absorption spectrum of the j-th sample to be tested to the validation spectral set is equivalent to constructing an independent sample set J for the j-th sample to be tested. The set J consists of the near-infrared absorption spectra of all qualified samples in the validation spectral set and the j-th sample to be tested. The overall absorbance difference of the j-th sample to be tested can be calculated using the set J. .

[0058] Using the average value and standard deviation of the overall absorbance differences of all qualified samples in the original validation spectral set to calculate the Grubbs statistic of the sample to be tested means that when calculating the Grubbs statistic of the j-th sample to be tested, the and of all qualified samples in the original validation spectral set are still used to calculate the Grubbs statistic of this sample to be tested, rather than using the newly constructed set J to calculate the new and , so as to more truly reflect and the of the validation spectral set.

[0059] Specifically, the Grubbs statistic of the sample to be tested is calculated according to the following formula

[0060]

[0061] where, is the Grubbs statistic of the j-th sample to be tested, is the overall absorbance difference of the j-th sample to be tested in the validation spectral set, that is, the overall absorbance difference of the j-th sample to be tested in the set J.

[0062] S4. Calculate the matching degree between the sample to be tested and the qualified samples based on the maximum Grubbs statistic of the qualified samples screened in step S2 and the Grubbs statistic of the sample to be tested in step S3. Determine that the sample to be tested with a matching degree meeting the set conditions is of qualified quality.

[0063] The matching degree between the preprocessed spectrum of the sample to be tested and the validation spectral set is calculated according to the following formula

[0064]

[0065] where, is the matching degree between the preprocessed spectrum of the j-th sample to be tested and the verification spectrum set, , that is is twice the maximum overall absorbance difference of qualified samples in the verification spectrum set.

[0066] When , it is determined that the sample to be tested is unqualified. When , it is determined that the sample to be tested needs further testing. When , it is determined that the sample to be tested is qualified.

[0067] The foam fire extinguishing agent quality detection system described in the present invention includes:

[0068] Spectrum acquisition and processing module: used to collect the near-infrared absorption spectra of multiple qualified samples and samples to be tested of the foam fire extinguishing agent and perform smoothing processing to obtain preprocessed spectra;

[0069] Spectrum screening module: used to calculate the overall absorbance difference of all samples based on the preprocessed spectra, and use the Grubbs test method to screen out qualified samples that meet the requirements based on the overall absorbance difference, and use their preprocessed spectra as the verification spectrum set;

[0070] Grubbs statistic calculation module for samples to be tested: used to add each sample to be tested into the verification spectrum set separately, and calculate the Grubbs statistic of the sample to be tested using the average value and standard deviation of the overall absorbance difference of all qualified samples in the verification spectrum set;

[0071] Detection module: used to calculate the matching degree between the maximum Grubbs statistic of the qualified samples screened by the spectrum screening module and the Grubbs statistic of the sample to be tested in step S3, and determine that the quality of the sample to be tested is qualified if the matching degree meets the set conditions.

[0072] The computer-readable storage medium storing one or more programs described in the present invention, the one or more programs include instructions, and when the instructions are executed by a computing device, the computing device is caused to execute the above method.

[0073] To better illustrate the method of the present invention, the following will be described with two specific embodiments:

[0074] Embodiment 1: Establishment of an optimal verification set spectral library based on 5 qualified 6% AFFF samples and quality evaluation of unknown samples to be tested

[0075] The specific steps are as follows:

[0076] (1) Use an absorption cell with an optical path length of 0.1 cm, and use a near-infrared spectrometer to test the near-infrared absorption spectra of the samples between 1200 nm and 2400 nm. Set the scanning interval to 1.0 nm, with a total of 1201 wavelengths. Each sample is sampled three times repeatedly (to reduce the error caused by internal inhomogeneity of the sample), and three near-infrared absorption spectra are continuously measured for each sampled sample. Finally, all the spectra are averaged and converted into one spectrum, and the Savitzky-Golay convolution smoothing method is used to smooth the original spectrum, with the window point number set to 10 and the polynomial order to 2, to obtain the preprocessing standard spectrum of the qualified sample; using the same test and preprocessing conditions as in step one, obtain the preprocessing spectra of 5 qualified 6%AFFF samples and use them as the spectra of the initial validation set.

[0077] (2) Calculate the overall difference in near-infrared spectral absorbance of the samples in the initial validation set , and the calculation model is as follows:

[0078]

[0079] where i is a natural number within [1, +∞), is the overall difference in near-infrared spectral absorbance of the i-th qualified sample, is the absorbance value of the i-th qualified sample under the k-th wavelength condition, is the average absorbance value of all qualified samples in the initial validation set under the k-th wavelength condition, and n is the total number of wavelengths of the light used for measurement, n = 1201.

[0080] After calculation, the overall differences in near-infrared spectral absorbance of the 5 qualified 6% AFFF samples are shown in Table 1:

[0081] Table 1 Overall differences in near-infrared spectral absorbance of 5 qualified 6%AFFF samples

[0082] Sample Number 1 2 3 4 5 <![CDATA[E i > 0.2717 0.3698 0.3269 0.3669 0.4223

[0083] (3) Calculate the Grubbs statistic of the samples in the initial validation set , and the calculation model is as follows:

[0084]

[0085] where, is the Grubbs statistic of the i-th qualified sample, is the average value of the overall differences in near-infrared spectral absorbance of all qualified samples, and S is the standard deviation of the overall differences in near-infrared spectral absorbance of all qualified samples.

[0086] Because in this embodiment The sample set used only contains qualified samples, so there is no representation for distinction here.

[0087] The calculation formula of S is as follows:

[0088]

[0089] where N is the total number of qualified samples in the initial validation set.

[0090] After calculation, the Grubbs statistic of the qualified samples in 5 initial validation sets of 6% AFFF is shown in Table 2:

[0091] Table 2 Grubbs statistics of qualified samples in 5 initial validation sets of 6% AFFF of the qualified samples

[0092] Sample Number 1 2 3 4 5 <![CDATA[G i > 1.5890 0.3654 0.4908 0.3077 1.4101

[0093] After analysis, the Grubbs statistic of 5 qualified samples is less than the critical value , , so no further optimization is required and it can be used as the optimal validation spectral set;

[0094] Using the same near-infrared spectrum test and pretreatment method for qualified samples, the pretreated near-infrared absorption spectrum of test sample 6 is obtained. The near-infrared absorption spectrum of test sample 6 is added to the optimal validation spectral set, and the overall difference in near-infrared spectral absorbance of test sample 6 is calculated using the same calculation method as the samples in the initial validation set. Only during the calculation process needs to be replaced with the average absorbance value of all samples including the test sample under the k-th wavelength condition, and finally is obtained.

[0095] From this, the Grubbs statistic of the test sample is calculated according to the following formula

[0096]

[0097] where, is the average value of the overall differences in near-infrared spectral absorbance of all qualified samples in the optimal validation spectral set, S is the standard deviation of the overall differences in near-infrared spectral absorbance of all qualified samples in the optimal validation spectral set, and the calculation gives .

[0098] (5) Calculate the consistency spectral threshold of the optimal validation spectral set , and the calculation model is as follows:

[0099] , Max(G i ) = 1.5890,

[0100] After calculation, = 3.1780;

[0101] (6) Calculate the matching degree between the near-infrared spectrum of the sample to be tested 6 and the optimal validation set spectrum. The calculation model is as follows:

[0102]

[0103] After calculation, the matching degree R of the sample to be tested 6 is 54.3186. The matching degree R is much less than 90%, so it is determined that the sample to be tested 6 is a non-conforming sample. By Figure 1 directly comparing the near-infrared absorption spectrum of the sample to be tested 6 and the spectra of 5 qualified samples in the validation set, there are obvious differences between the two spectra.

[0104] Example 2: Establishment of an optimal validation set spectral library based on 36 3% AFFF qualified samples and quality evaluation of unknown samples to be tested

[0105] Specifically, it includes the following steps:

[0106] (1) Use an absorption cell with an optical path of 0.2 cm, and use a near-infrared spectrometer to measure the near-infrared absorption spectrum of the sample between 1200 nm and 2400 nm. Set the scanning interval to 1.0 nm, with a total of 1201 wavelengths. The near-infrared spectrum of each sample is collected three times, and three spectra are continuously collected each time. Finally, all spectra are averaged and converted into one spectrum, and the Savitzky-Golay convolution smoothing method is used to smooth the original spectrum. Set the window points to 10 and the polynomial order to 2 to obtain the preprocessed standard spectrum of the qualified sample; using the same test and preprocessing conditions as in step one, obtain the preprocessed spectral diagrams of 36 3% AFFF qualified samples and use them as the initial validation set spectra;

[0107] (2) Calculate the overall difference in the near-infrared spectrum absorbance of the initial validation set samples , and the calculation model is as follows:

[0108]

[0109] Among them, i is a natural number within [1, +∞), is the overall difference in the near-infrared spectrum absorbance of the i-th qualified sample, is the absorbance value of the i-th qualified sample under the k-th wavelength condition, is the average absorbance value of all qualified samples in the initial validation set at the k-th wavelength condition, n is the total number of wavelengths of the light used for measurement, and n = 1201.

[0110] After calculation, the overall difference in the near-infrared spectral absorbance of 39 qualified 3% AFFF samples As shown in Table 3:

[0111] Table 3 Overall difference in the near-infrared spectral absorbance of 39 qualified 3% AFFF samples

[0112] Sample Number <![CDATA[E i > Sample Number <![CDATA[E i > Sample Number <![CDATA[E i > Sample Number <![CDATA[E i > 1 0.2197 11 0.1521 21 0.1396 31 0.1419 2 0.2321 12 0.1786 22 0.0841 32 0.1533 3 0.1832 13 0.0985 23 0.1289 33 0.1425 4 0.2046 14 0.1704 24 0.1673 34 0.1919 5 0.162 15 0.1187 25 0.1271 35 0.1488 6 0.1983 16 0.1532 26 0.1752 36 0.1896 7 0.1863 17 0.1869 27 0.1125 37 0.1932 8 0.1339 18 0.1082 28 0.1433 38 0.1585 9 0.1358 19 0.1614 29 0.1102 39 0.1843 10 0.1286 20 0.1246 30 0.1539

[0113] (3)Calculate the Grubbs statistic of the samples in the initial validation set near The calculation model is as follows:

[0114]

[0115] Among them, is the Grubbs statistic of the i-th qualified sample, is the average value of the overall difference in the near-infrared spectral absorbance of all qualified samples, and S is the standard deviation of the overall difference in the near-infrared spectral absorbance of all qualified samples.

[0116] The calculation formula of S is as follows:

[0117]

[0118] where N is the total number of qualified samples in the initial validation set.

[0119] After calculation, the Grubbs statistic of 39 qualified samples in the initial validation set of 3% AFFF is As shown in Table 4:

[0120] Table 4 Grubbs statistic of 39 qualified samples in the initial validation set of 3% AFFF is

[0121] Sample Number <![CDATA[G i > Sample Number <![CDATA[G i > Sample Number <![CDATA[G i > Sample Number <![CDATA[G i > 1 1.9009 11 0.1152 21 0.4897 31 0.422 2 2.2719 12 0.6756 22 2.1468 32 0.0815 3 0.812 13 1.7179 23 0.8103 33 0.4045 4 1.4506 14 0.4311 24 0.3368 34 1.072 5 0.1802 15 1.1128 25 0.862 35 0.2155 6 1.2631 16 0.0831 26 0.5744 36 1.0024 7 0.9036 17 0.9231 27 1.2988 37 1.1099 8 0.6594 18 1.4266 28 0.3801 38 0.0754 9 0.6041 19 0.1609 29 1.3684 39 0.8455 10 0.8189 20 0.9385 30 0.0632

[0122] After analysis, among the 39 qualified samples, the Grubbs statistics of samples No. 2 and No. 22 are both greater than the critical value of 2.0 and should be discarded; after discarding samples No. 2 and No. 22, further optimization is performed on the remaining 37 qualified sample validation set;

[0123] After calculation, the overall difference in the near-infrared spectral absorbance of 37 qualified 3% AFFF samples As shown in Table 5:

[0124] Table 5 Overall difference in near-infrared spectral absorbance of 37 qualified 3% AFFF samples

[0125] Sample Number <![CDATA[E i > Sample Number <![CDATA[E i > Sample Number <![CDATA[E i > Sample Number <![CDATA[E i > 1 0.2226 11 0.1555 21 0.1413 31 0.1411 2 Rejected 12 0.1802 22 Rejected 32 0.1498 3 0.1867 13 0.0993 23 0.1273 33 0.141 4 0.2073 14 0.1699 24 0.1646 34 0.1886 5 0.1657 15 0.1213 25 0.1269 35 0.1474 6 0.2011 16 0.1527 26 0.1727 36 0.1872 7 0.1899 17 0.1875 27 0.1099 37 0.189 8 0.1365 18 0.1086 28 0.1433 38 0.1557 9 0.1383 19 0.1604 29 0.1079 39 0.1807 10 0.1298 20 0.1243 30 0.1505

[0126] The Grubbs statistic of 37 qualified 3% AFFF samples in the initial validation set was calculated as follows: As shown in Table 6:

[0127] Table 6 Grubbs statistic of 37 qualified 3% AFFF samples in the initial validation set as follows:

[0128] Sample Number <![CDATA[G i > Sample Number <![CDATA[G i > Sample Number <![CDATA[G i > Sample Number <![CDATA[G i > 1 2.2283 11 0.0093 21 0.4821 31 0.4879 2 Rejected 12 0.8163 22 Rejected 32 0.1981 3 1.0332 13 1.8825 23 0.9468 33 0.493 4 1.7166 14 0.4718 24 0.2961 34 1.0939 5 0.3328 15 1.1492 25 0.9604 35 0.2798 6 1.5131 16 0.1006 26 0.5663 36 1.0481 7 1.1396 17 1.0577 27 1.5276 37 1.1092 8 0.6408 18 1.5707 28 0.416 38 0.003 9 0.5806 19 0.1566 29 1.5958 39 0.8309 10 0.8633 20 1.0472 30 0.176

[0129] After analysis, among the 37 qualified samples, the Grubbs statistic of Sample No. 1 was greater than the critical value of 2.0 and should be discarded; after discarding Sample No. 1, the operations in (2) and (3) were repeated to further optimize the remaining 36 qualified sample validation set; After calculation, the overall difference in near-infrared spectral absorbance of the remaining 36 qualified 3% AFFF samples

[0130] is shown in Table 7: As shown in Table 7:

[0131] Table 7 Overall difference in near-infrared spectral absorbance of 36 qualified 3% AFFF samples

[0132] Sample Number <![CDATA[E i > Sample Number <![CDATA[E i > Sample Number <![CDATA[E i > Sample Number <![CDATA[E i > 1 Rejected 11 0.1578 21 0.1431 31 0.1405 2 Rejected 12 0.1838 22 Rejected 32 0.1474 3 0.19 13 0.1001 23 0.1258 33 0.1405 4 0.2103 14 0.1674 24 0.165 34 0.1858 5 0.1681 15 0.1234 25 0.1269 35 0.1443 6 0.2025 16 0.1548 26 0.1719 36 0.186 7 0.1927 17 0.1857 27 0.1088 37 0.1857 8 0.1384 18 0.1094 28 0.1433 38 0.1532 9 0.1399 19 0.1586 29 0.1062 39 0.1766 10 0.1303 20 0.1245 30 0.1476

[0133] The Grubbs statistic of 36 qualified 3% AFFF samples in the initial validation set was calculated as follows: As shown in Table 8:

[0134] Table 8 Grubbs statistic of 36 qualified 3% AFFF samples in the initial validation set as follows:

[0135] Sample Number <![CDATA[G i > Sample Number <![CDATA[G i > Sample Number <![CDATA[G i > Sample Number <![CDATA[G i > 1 Rejected 11 0.1395 21 0.3756 31 0.4694 2 Rejected 12 1.0563 22 Rejected 32 0.2267 3 1.2754 13 1.8945 23 0.986 33 0.4704 4 1.9931 14 0.4799 24 0.3942 34 1.1265 5 0.5034 15 1.0734 25 0.9469 35 0.3353 6 1.718 16 0.0359 26 0.6373 36 1.1358 7 1.3715 17 1.1257 27 1.5862 37 1.1258 8 0.5425 18 1.5664 28 0.3712 38 0.0203 9 0.4895 19 0.1679 29 1.6787 39 0.8048 10 0.8275 20 1.0337 30 0.2183

[0136] After analysis, the Grubbs statistic of the 36 qualified samples was less than the critical value of 2.0. Therefore, no further optimization was required, and the near-infrared spectra of the remaining 36 qualified samples were selected as the optimal validation spectral set;

[0137] ​(5) The preprocessed near-infrared spectrogram of the sample to be measured X1 is obtained by using the same near-infrared spectrum test and preprocessing method as the spectral set qualified samples, and the overall difference in near-infrared spectrum absorbance of the sample to be measured X1 is calculated by using the same calculation method as the validation set samples. , except that during the calculation process it needs to be replaced with the average absorbance value of all 37 samples including the sample to be measured X1 and 36 qualified samples under the k-th wavelength condition, and finally

[0138] Thus, the Grubbs statistic of the sample to be measured is calculated according to the following formula

[0139]

[0140] where is the average value of the overall difference in near-infrared spectrum absorbance of all 36 qualified samples in the optimal validation spectral set, S is the standard deviation of the overall difference in near-infrared spectrum absorbance of all 36 qualified samples in the optimal validation spectral set, and the calculation gives .

[0141] (6) Calculate the consistency spectrogram threshold of the qualified sample validation spectral set , and the calculation model is as follows:

[0142] , Max(Ti) = 1.9931,

[0143] After calculation, = 3.9862;

[0144] (7) Calculate the matching degree between the near-infrared spectrum of the sample to be measured X1 and the optimal validation set spectrum, and the calculation model is as follows:

[0145]

[0146] After calculation, the matching degree R of the sample to be measured X1 is 90.24. When the matching degree R ∈ [90% - 95%], the sample needs to be further tested for fire extinguishing performance; by directly comparing the near-infrared spectrogram of the sample to be measured X1 and the spectrograms of 36 validation set qualified samples, the difference between the two spectrograms is small, and only after magnifying the local spectrogram can the tiny differences be seen (attached Figure 2 ), and it is easy to misjudge if directly judging whether the sample is qualified through the spectrogram.

[0147] Analyze the various indicators of sample X1 using the national standard method, and the specific results are shown in Table 9:

[0148] Table 9 Physical indicators and fire extinguishing performance of 3% AFFF unknown sample X1

[0149] Serial Number Test Items Test Data of Nonconforming Samples 1 Density 1.056 2 Surface Tension 22.0 3 Interfacial Tension 3.5 4 Diffusion Coefficient -0.5 5 Freezing Point -10 6 Foaming Ratio (11.4 L Pipe Gun) 7.5 7 25% Liquid Separation Time (11.4L Pipe Gun) 3’10’’ 8 90% Fire Control Time 2’40’’ 9 Fire Extinguishing Time Incomplete Fire Extinguishing at 3' 10 Fire Resistance Time /

[0150] The analysis results of national standards show that the physical indexes of "surface tension" and "diffusion coefficient", and the fire extinguishing performance indexes of "90% fire control time" and "fire extinguishing time" of the 3% AFFF unknown sample X1 are unqualified items, which are consistent with the prediction results provided by this method, indicating that this method has high accuracy.

Claims

1. A method for testing the quality of a foam fire extinguishing agent, characterized in that: The following steps are involved: S1, selecting a plurality of qualified samples of foam fire extinguishing agent and samples to be tested, collecting their near infrared absorption spectra and performing smoothing to obtain pre-processed spectra; S2. Calculate the overall absorbance difference of all qualified samples according to the preprocessed spectra using the following formula: Among them, E i is the overall absorbance difference of the i-th qualified sample, x ik represents the absorbance value of the i-th qualified sample at the k-th wavelength, μ k For calculating E i The average absorbance value of all qualified samples in the sample set at the kth wavelength, where n is the total number of wavelengths of light used for measurement; Based on the overall difference in absorbance, the Grubbs test method is used to screen out qualified samples that meet the requirements and use their pre-processed spectra as the verification spectrum set. The screening of qualified samples specifically includes the following steps: S2.

1. Calculate the Grubbs statistic of qualified samples according to the following formula: Among them, G i is the Grubbs statistic of the i-th qualified sample, is the average value of the overall absorbance difference of all qualified samples, and S is the standard deviation of the overall absorbance difference of all qualified samples; S2.

2. Confirm whether G exists i Greater than the critical value G p If there are any, remove these qualified samples and return to step S2.1 until there is no G i Greater than the critical value G p ; S3, add the near infrared absorption spectrum of each sample to be tested to the verification spectrum set to calculate the overall difference in absorbance. The Grubbs statistic of the sample to be tested is calculated according to the following formula: Among them, G j is the Grubbs statistic of the jth sample to be tested, E j is the overall difference in absorbance of the jth sample to be tested in the validation spectrum set; S4. Calculate the matching degree between the sample to be tested and the qualified sample according to the following formula: Among them, R j is the matching degree between the jth sample to be tested and the qualified sample, γ=2Max(G i ); When R j ∈(0,0.9) when the sample is judged as unqualified. j ∈[0.9,0.95], it is determined that the sample needs further testing. j ∈(0.95,1], the sample to be tested is considered qualified.

2. The foam fire extinguishing agent quality detection method according to claim 1, characterized in that: In step S1, the near-infrared absorption spectrum of each sample is repeatedly collected multiple times, and the spectra collected multiple times are averaged and smoothed to obtain a pre-processed spectrum of the sample.

3. The foam fire extinguishing agent quality detection method according to claim 1, characterized in that: In the step S1, a Savitzky-Golay filter, an AdjacentAveraging algorithm or an FFTFiltering algorithm is used to smooth the near-infrared absorption spectrum.

4. The foam fire extinguishing agent quality detection method according to claim 1, characterized in that: The material of the cuvette used for collecting the spectrum in step S1 is quartz or optical glass.

5. A foam fire extinguishing agent quality detection system using the method of claim 1, characterized in that: The system comprises: Spectrum acquisition and processing module: used to collect near-infrared absorption spectra of multiple qualified samples and samples to be tested of foam fire extinguishing agent and perform smoothing to obtain pre-processed spectra; Spectral screening module: used to calculate the overall absorbance difference of all samples based on the pre-processed spectra, and use the Grubbs test method to screen out qualified samples that meet the requirements and use their pre-processed spectra as the verification spectrum set; Grubbs statistic calculation module for samples to be tested: used to add each sample to be tested to the verification spectrum set to calculate its overall absorbance difference, and use the average value and standard deviation of the overall absorbance difference of all qualified samples in the original verification spectrum set to calculate the Grubbs statistic of the sample to be tested; Detection module: used to calculate the matching degree between the tested sample and the qualified sample by using the maximum Grubbs statistic of the qualified samples after screening by the spectral screening module and the Grubbs statistic of the tested sample in step S3, and judge the tested sample whose matching degree meets the set conditions as qualified.

6. A computer-readable storage medium storing one or more programs, characterized in that : The one or more programs include instructions, which, when executed by a computing device, cause the computing device to perform any of the methods according to claims 1 to 4.

Images