Method for detecting fatty acid value of corn by utilizing near-infrared technology

Abstract

The invention belongs to the technical field of feed quality detection, and particularly relates to a method for detecting the fatty acid value of corn by using a near-infrared technology. The method for detecting the fatty acid value of the corn by utilizing the near-infrared technology comprises the following steps: S1, collecting spectral information of a corn sample and measuring a chemical value of the fatty acid value: collecting the corn sample, pre-treating the corn sample, performing near-infrared spectrum scanning on the treated corn sample, and collecting near-infrared spectrum information of the corn sample; detecting a fatty acid value result, namely a chemical value, of the corn sample by adopting a chemical method; the method for pretreating the corn sample comprises the following steps: removing impurities, crushing the corn sample subjected to impurity removal, and passing through a screen with the aperture of 0.3-0.5 mm; s2, spectrum processing of the corn sample and establishment of a calibration model: preprocessing near infrared spectrum information of the corn sample, and establishing the calibration model between a near infrared spectrum and a corn fatty acid value by using a chemometrics method; s3, rapid detection of a to-be-detected corn sample: taking the to-be-detected corn sample, carrying out near infrared spectrum scanning, collecting near infrared spectrum information of the to-be-detected corn sample, and importing the near infrared spectrum information of the to-be-detected corn sample into the calibration model established in the step S2 to obtain a fatty acid value result of the to-be-detected corn sample. According to the method, the detection efficiency can be remarkably improved, real-time monitoring is realized, and the quality of raw materials is guaranteed.

Description

Technical Field

[0001] The invention belongs to the technical field of feed quality detection, and in particular relates to a method for detecting corn fatty acid value by utilizing near infrared technology. Background Art

[0002] Corn is one of the important food crops in the world today. As we all know, corn is the "king of energy feed". The reason why it is widely used in poultry and livestock farming is that its available energy is very high. It is reported that every 2-3 kg of corn can convert into 1 kg of meat. At present, the standard for determining energy feed is that the digestible energy per kilogram of feed dry matter is above 10MJ. Those above 12.5MJ are high-energy feeds. The average digestible energy of corn is 14.06MJ / kg, and the highest can reach 15.06MJ / kg, so the available energy of corn is the highest among cereal seeds. In addition, corn has the characteristics of good palatability and no restrictions on use. 65% to 70% of corn in the world is used as feed. However, during the storage of corn, especially when the moisture content and temperature are high, mold is easy to multiply in large quantities, and the fat hydrolysis rate is accelerated, resulting in a significant increase in the content of free fatty acids. Therefore, the determination of fatty acid value can reflect the intrinsic quality of corn to a certain extent. It is a very important quality control indicator for determining the storage quality of corn.

[0003] GB / T 20570-2015 "Maize Storage Quality Determination Rules" provides a manual titration method for testing the fatty acid value of corn. The operation is cumbersome, time-consuming and labor-intensive, and requires a variety of chemical reagents. Its promotion and application in the field of real-time quality monitoring has obvious limitations and brings inconvenience to feed companies in the quality management of raw materials. Therefore, when corn arrives, how to effectively and quickly evaluate the storage quality of the batch of corn is a key issue that feed companies need to pay attention to. Summary of the invention

[0004] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a method for detecting the fatty acid value of corn using near infrared technology. The method of the present invention can significantly improve the detection efficiency, realize real-time monitoring, and ensure the quality of raw materials.

[0005] The technical solution of the present invention is:

[0006] A method for detecting the fatty acid value of corn using near infrared technology comprises the following steps:

[0007] S1. Spectral information collection of corn samples and chemical value determination of fatty acid values

[0008] Collect corn samples, perform pre-treatment on the corn samples, perform near-infrared spectrum scanning on the treated corn samples, and collect near-infrared spectrum information of the corn samples; and simultaneously use chemical methods to detect the fatty acid value results of the corn samples, i.e., the chemical value;

[0009] The steps of pre-processing the corn sample are as follows: removing impurities, crushing the corn sample after impurity removal, and passing it through a sieve with a pore size of 0.3-0.5 mm;

[0010] S2, Spectral processing of corn samples and establishment of calibration model

[0011] The near infrared spectrum information of corn samples was preprocessed, and a calibration model between near infrared spectrum and corn fatty acid value was established by chemometric method.

[0012] S3, Rapid detection of corn samples to be tested

[0013] Take the corn sample to be tested and perform near infrared spectrum scanning to collect near infrared spectrum information of the corn sample to be tested, import the near infrared spectrum information of the corn sample to be tested into the calibration model established in step S2, and obtain the fatty acid value result of the corn sample to be tested.

[0014] Furthermore, in the step S1, before the scanning operation, the sample is loaded and compacted using a tabletting cover with a fixed weight of 75±5g, and the sample thickness is controlled to be 2-4mm.

[0015] Furthermore, in the step S1, when performing near-infrared spectral scanning and collecting near-infrared spectral information of the corn sample, the light source lamp is a tungsten lamp, a PbS detector is used, diffuse reflection technology is adopted, the spectral wavelength range is 1100-2500nm, the number of scans is 32, and the resolution is 0.5nm; the scanning environment: temperature 25.0±1.0℃, relative humidity 45±10%.

[0016] Furthermore, in step S1, the fatty acid value of the corn sample is detected by chemical titration according to the method in GB / T 20570-2015 "Corn Storage Quality Determination Rules".

[0017] The accuracy of chemical value determination can largely determine the accuracy of near-infrared results. Therefore, the chemical value determination method of the calibration set samples must be authoritative and standard. The present invention uses chemical titration to detect the fatty acid value results of corn samples according to the method in GB / T 20570-2015 "Rules for Determining Corn Storage Quality", which can ensure the accuracy of the results to the greatest extent.

[0018] Furthermore, in the preprocessing step S2, the preprocessing method is a combination of a smoothing processing method, a derivative processing method and a multivariate scattering correction method.

[0019] Furthermore, the smoothing method is a four-point smoothing method.

[0020] Furthermore, the derivation processing method is a first-order derivation processing method.

[0021] Furthermore, the multivariate scatter correction method is a combination of a standard normal variable transformation method and a detrending correction method.

[0022] Furthermore, the chemometric method in step S2 is partial least squares method.

[0023] Furthermore, the step S2 also includes evaluating the prediction performance of the established calibration model, wherein the evaluation parameters mainly include the calibration set correlation coefficient and standard deviation, the cross-validation correlation coefficient and cross-validation standard deviation, and the validation set correlation coefficient and standard deviation; screening out a calibration model in which the calibration set correlation coefficient and the validation set correlation coefficient are close to 1, and the standard deviation is smaller and closer to the standard deviation; finally, calculating the relative analysis error through the validation set, if the relative analysis error is greater than 3, it means that the calibration prediction effect is good, the prediction accuracy is high, and the established calibration model can be used for actual detection; if 2.5<relative analysis error<3, it means that it is feasible to use the established calibration model to quantitatively analyze the corn fatty acid value, but its prediction accuracy needs to be further improved; if the relative analysis error is less than 2.5, it means that the corn fatty acid value is difficult to detect and analyze using the established calibration model.

[0024] The main factors affecting the rationality of establishing the calibration model are: ① The accuracy of chemical value determination can largely determine the accuracy of near-infrared results. Therefore, the chemical value determination method of the calibration set samples must be authoritative and standard to ensure the accuracy of the results; ② The preprocessing method of near-infrared spectrum and calibration calculation technology have a great influence on modeling.

[0025] The present invention has been obtained through a large number of creative experiments, and the four-point smoothing processing method, the first-order derivative processing method and the multivariate scattering correction method (the standard normal variable transformation method and the detrending correction method) and the partial least squares method can be used to obtain a calibration model with good calibration prediction effect and high prediction accuracy, which can be used for actual detection. The method of the present invention can significantly improve the detection efficiency, realize real-time monitoring, and ensure the quality of raw materials.

[0026] Compared with the prior art, the present invention has the following advantages:

[0027] (1) Compared with the process steps of the standard method, the method of detecting the fatty acid value of corn using near-infrared technology in the present invention has no chemical treatment, no pollution, rapid detection, simple operation, and is suitable for wide promotion and use.

[0028] (2) The method of detecting the fatty acid value of corn using near-infrared technology in the present invention significantly improves the detection efficiency and is suitable for on-site detection and monitoring. The national standard method detects the fatty acid value of corn. The whole process from sample preparation to result is at least 2 hours. The method of the present invention only needs to remove impurities such as corn cobs and corn husks first, and use a grinder to grind them to pass through a sieve with a hole diameter of 0.3-0.5 mm. During the grinding process, it is important that the inside of the grinder cannot heat up. It needs to stop for about 12 seconds to dissipate heat and then continue for 2 seconds until the sample fineness reaches the requirement. The scanning is completed within 1 hour after the sample is prepared to prevent the sample from oxidizing and rancidifying. The steps of weighing and other detection operations can be eliminated, and the process from grinding to result is completed within 10 minutes. The detection efficiency is significantly improved. Because the method is simple in steps and takes a short time, it can be used for on-site detection and monitoring of feed factories, thereby effectively ensuring the storage quality of corn.

[0029] (3) The accuracy and stability of the results of the determination of the fatty acid value of corn in the national standard method depend to a large extent on the steps of the detection process and the determination of the titration end point, because fatty acids are weak acids, and the mutation is not obvious when the titration approaches the end point, which easily causes the test results to be too high or too low. In addition, the individual differences of the testers make it more difficult to grasp the titration end point. The method of using the present invention to detect the fatty acid value of corn using near-infrared technology is simple in process, has few influencing factors, and has guaranteed test results.

[0030] (4) The method for detecting the fatty acid value of corn using near-infrared technology of the present invention has simple and clear operation steps, and is convenient and practical for novices to use without extensive training and study. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The drawings constituting a part of the present application are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. Among them:

[0032] Figure 1 is the original near-infrared diffuse reflectance spectrum of the corn sample in Example 1 of the present invention;

[0033] Figure 2 It is a corn spectrum diagram after smoothing, first-order derivative, and multivariate scattering correction processing in Example 1 of the present invention;

[0034] Figure 3 It is a scatter plot of the relationship between the chemical values ​​of the fatty acid values ​​of the calibration set and the predicted values ​​of the fatty acid values ​​of the calibration model in the internal validation of the calibration model;

[0035] Figure 4 It is a scatter plot of the relationship between the chemical values ​​of fatty acid values ​​and the predicted values ​​of fatty acid values ​​in the validation set in the external validation of the calibration model. DETAILED DESCRIPTION

[0036] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention belong to the scope of protection of the present invention.

[0037] Unless otherwise specified, the test materials and reagents used in the following examples can be obtained from commercial sources.

[0038] Example 1: A method for detecting the fatty acid value of corn using near infrared technology

[0039] The method for detecting the fatty acid value of corn using near infrared technology comprises the following steps:

[0040] S1. Spectral information collection of corn samples and chemical value determination of fatty acid values

[0041] 309 corn samples were collected, which were from different origins, varieties and years. They were hard corn, concave corn and semi-concave corn commonly used in the feed industry on the market. They included both old and new grains, and tried to ensure that the regions and years covered by the collected samples had the largest variation range.

[0042] Sample pretreatment: first remove impurities such as corn cobs and corn husks, and use a grinder to grind the corn sample to pass through a sieve with a hole diameter of 0.425mm to ensure the uniformity of the sample and more stable and accurate test results. During the grinding process, be careful not to heat up the inside of the grinder. It takes about 12 seconds to stop and dissipate heat for 2 seconds before continuing until the sample fineness meets the requirements. The sample should be measured within 1 hour after preparation to prevent oxidation and rancidity of the sample.

[0043] The collected corn samples were scanned by near infrared spectrum at room temperature. Before the scanning operation, the samples were compacted with a tablet cover of fixed weight 75g. The sample thickness was controlled at 3mm. The near infrared spectrum information of the corn samples was collected. When collecting the near infrared spectrum information of the corn samples, the light source lamp was a tungsten lamp, a PbS detector was used, and diffuse reflection technology was adopted. The spectrum wavelength range was 1100-2500nm, the number of scans was 32, the sampling interval was 0.5nm, and the resolution was 0.5nm. The scanning environment requirements were: temperature 25.0℃, relative humidity 45%. The original near infrared diffuse reflection spectrum of the corn sample in Example 1 of the present invention is shown in FIG. Figure 1 As shown. Figure 1 It can be seen that the absorption spectra of corn samples have consistent changing trends, but do not completely overlap, which shows both the differences between different samples and the consistency of large sample groups.

[0044] According to the method in GB / T 20570-2015 "Maize Storage Quality Determination Rules", the fatty acid value of the corn sample is tested by chemical titration method, that is, the chemical value, which is used as the measured data of the calibration set;

[0045] S2, Spectral processing of corn samples and establishment of calibration model

[0046] The near infrared spectrum information of corn samples was preprocessed, and a calibration model between near infrared spectrum and corn fatty acid value was established by chemometric method.

[0047] The spectral information is preprocessed by using the four-point smoothing method, the first-order derivative processing method (1, 4, 4, 1) and the multivariate scattering correction method (standard normal variable transformation method and detrending correction method). That is, (1, 4, 4, 1) (first-order derivative processing, the derivative processing spectrum interval is 4nm, the smoothing processing interval is 4nm, and one smoothing processing is performed) and the multivariate scattering correction method (standard normal variable transformation method and detrending correction method) are used to preprocess the spectral information. The corn spectrum after smoothing, first-order derivative, and multivariate scattering correction is shown in Figure 1. Figure 2 As shown. Figure 2 It can be clearly seen that the fineness of the spectrum is improved, the instability and drift of the baseline are also reduced, and the spectral characteristics of the sample can be more accurately reflected. This shows that the use of the above method to pre-process the spectral information can reduce the influence of noise in the spectrum, improve the fineness of the spectrum, eliminate baseline drift, strengthen the spectral band characteristics, etc., and reduce the influence of factors such as particle size and uniformity on the spectrum, so as to more accurately reflect the spectral characteristics of the sample.

[0048] The calibration model was established by cross-validation. First, the overall trends of the original spectra and the corn spectra after smoothing, first-order derivative, and multivariate scattering correction were compared. According to the Kennard-Stone computer recognition method, samples with large spectral differences were selected as calibration set samples, and the remaining samples were validation set samples. The content distribution of the calibration set samples was relatively uniform, so all the collected corn samples were divided into a calibration set (a total of 256) and an external validation set (a total of 53), which were used for the establishment of the calibration model and the external validation of the model, respectively.

[0049] After a large number of creative experiments, the calibration equation parameters calculated by different processing methods are shown in Table 1. From the determination coefficient and standard deviation of the calibration model, the regression methods of partial least squares (PLS) and modified partial least squares (MPLS) are significantly better than the regression method of principal component analysis (PCR); the first-order derivative processing effect is significantly better than the second-order derivative processing, and the first-order derivative combined with four-point smoothing processing (1, 4, 4, 1) method is slightly better than the first-order derivative combined with eight-point smoothing processing (1, 8, 8, 1) method; among the scattering correction methods, the standard normalization + descattering processing method finally obtains better results than other methods.

[0050] Finally, combined with various parameter indicators, the partial least squares method, standard normalization + descattering processing, and (1, 4, 4, 1) were used to establish the calibration model, and the calibration equation was obtained. The calibration set correlation coefficient RSQ was 0.8898, the standard deviation SEC was 4.3074, the interactive validation correlation coefficient 1-VR was 0.8213, the interactive validation standard deviation SECV was 5.4812, the relative analysis error RPD was greater than 3, and the calibration result was good.

[0051] Table 1 Parameters of calibration equation for corn fatty acid value

[0052]

[0053]

[0054]

[0055] S3, Rapid detection of corn samples to be tested

[0056] Take the calibration set samples (a total of 256) and the external validation set samples (a total of 53) screened in step S2, perform near-infrared spectral scanning, collect the near-infrared spectral information of the samples, and import the near-infrared spectral information of the samples into the calibration model established in step S2, and obtain the predicted values ​​of the fatty acid content of the calibration set samples and the predicted values ​​of the fatty acid content of the external validation set samples.

[0057] Effect evaluation test

[0058] (1) Internal verification of the calibration model

[0059] Internal verification was performed on the samples involved in the establishment of the calibration model, and a scatter plot of the relationship between the chemical value of the fatty acid value of the calibration set and the predicted value of the fatty acid value of the calibration model was obtained in the internal verification of the calibration model ( Figure 3 ).Depend on Figure 3It can be concluded that the correlation coefficient between the chemical value and the predicted value of the fatty acid value of the calibration set is 0.881, the average chemical value of the fatty acid value of the calibration set is 48.241 mg KOH / g, and the average predicted value is 48.298 mg KOH / g. The results of the chemical value and the predicted value are very close; there is a good linear relationship between the predicted value and the chemical value of the calibration set.

[0060] (2) External validation of the calibration model

[0061] After the calibration model is built, in addition to using the correlation coefficient RSQ of its own calibration set and the cross-validation correlation coefficient (1-VR) parameter indicators for measurement, external validation is also required to evaluate the reliability of the built model and prove the effectiveness of the model in practical applications.

[0062] In this experiment, 53 samples that did not participate in the calibration process were selected as the validation set to verify the prediction accuracy and reliability of the calibration model of the near-infrared spectroscopy analysis method. The scatter plot of the relationship between the chemical value of the fatty acid value and the predicted value of the fatty acid value of the validation set in the external validation of the calibration model was obtained by the univariate linear regression method ( Figure 4 ).Depend on Figure 4 It can be concluded that the determination coefficient of the validation set is 0.911, the standard deviation SEP is 5.02, and the relative analysis error RPD is greater than 3, indicating that the calibration model has good stability and prediction accuracy, and the prediction effect is good, that is, the near-infrared spectroscopy analysis method established in the present invention can be used to quantitatively determine the fatty acid value of corn.

[0063] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution of the present invention, which should be included in the scope of the claims of the present invention.

Claims

1. A method for detecting the fatty acid value of corn using near infrared technology, It is characterized in that The following steps are involved: S1. Spectral information collection of corn samples and chemical value determination of fatty acid values Collect corn samples, perform pre-treatment on the corn samples, perform near-infrared spectrum scanning on the treated corn samples, and collect near-infrared spectrum information of the corn samples; and simultaneously use chemical methods to detect the fatty acid value results of the corn samples, i.e., the chemical value; The steps of pre-processing the corn sample are as follows: removing impurities, crushing the corn sample after impurity removal, and passing it through a sieve with a pore size of 0.3-0.5 mm; S2, Spectral processing of corn samples and establishment of calibration model The near infrared spectrum information of corn samples was preprocessed, and a calibration model between near infrared spectrum and corn fatty acid value was established by chemometric method. S3, Rapid detection of corn samples to be tested Take the corn sample to be tested and perform near infrared spectrum scanning to collect near infrared spectrum information of the corn sample to be tested, import the near infrared spectrum information of the corn sample to be tested into the calibration model established in step S2, and obtain the fatty acid value result of the corn sample to be tested.

2. The method for detecting the fatty acid value of corn using near infrared technology as claimed in claim 1, It is characterized in that In the step S1, near-infrared spectral scanning is performed to collect near-infrared spectral information of the corn sample. The light source lamp is a tungsten lamp, a PbS detector is used, and diffuse reflection technology is adopted. The spectral wavelength range is 1100-2500nm, the number of scans is 32, and the resolution is 0.5nm; the scanning environment: temperature 25.0±1.0℃, relative humidity 45±10%.

3. The method for detecting the fatty acid value of corn using near infrared technology as claimed in claim 1, It is characterized in that In the step S1, the fatty acid value of the corn sample is detected by chemical titration according to the method in GB / T 20570-2015 "Corn Storage Quality Determination Rules".

4. The method for detecting the fatty acid value of corn using near infrared technology as claimed in claim 1, It is characterized in that In the preprocessing step S2, the preprocessing method is a combination of a smoothing processing method, a derivative processing method and a multivariate scatter correction method.

5. The method for detecting the fatty acid value of corn using near infrared technology as claimed in claim 4, It is characterized in that The smoothing method is a four-point smoothing method.

6. The method for detecting the fatty acid value of corn using near infrared technology as claimed in claim 4, It is characterized in that The derivation processing method is a first-order derivation processing method.

7. The method for detecting the fatty acid value of corn using near infrared technology as claimed in claim 4, It is characterized in that The multivariate scatter correction method is a combination of a standard normal variable transformation method and a detrending correction method.

8. The method for detecting the fatty acid value of corn using near infrared technology as claimed in claim 4, It is characterized in that The chemometric method in step S2 is partial least squares method.

9. The method for detecting the fatty acid value of corn using near infrared technology as claimed in claim 1, It is characterized in that The step S2 also includes evaluating the prediction performance of the established calibration model, wherein the evaluation parameters mainly include the calibration set correlation coefficient and standard deviation, the cross-validation correlation coefficient and cross-validation standard deviation, and the validation set correlation coefficient and standard deviation; screening out a calibration model in which the calibration set correlation coefficient and the validation set correlation coefficient are close to 1, and the standard deviation is smaller and closer to the standard deviation; finally, calculating the relative analysis error through the validation set, if the relative analysis error is greater than 3, it means that the calibration prediction effect is good, the prediction accuracy is high, and the established calibration model can be used for actual detection; if 2.5<relative analysis error<3, it means that it is feasible to use the established calibration model to quantitatively analyze the corn fatty acid value, but its prediction accuracy needs to be further improved; if the relative analysis error is less than 2.5, it means that the corn fatty acid value is difficult to detect and analyze using the established calibration model.

Images