Methods and apparatus for analyzing milk components based on ultraviolet / visible spectroscopy

Abstract

This invention discloses a method and apparatus for analyzing milk components based on ultraviolet / visible spectroscopy, belonging to the field of agricultural product quality testing. The technical problem solved is to provide a low-cost, portable method and apparatus for analyzing milk components. The invention includes a display, a housing, cuvettes, a light source, a collimating lens, a cuvette holder, a spectral detection module, a power supply module, and a computational control module. Ultraviolet / visible light emitted by the light source is transmitted through the collimating lens to a cuvette containing milk diluted 100 times with deionized water and interacts with the milk. The spectral detection module receives the transmitted light signal, converts it into a digital signal, and transmits it to the computational control module. The computational control module calculates the fat, protein, lactose, and total solids content of the milk and provides feedback to the user through the display. The power supply module supplies power to the device. This invention can be used to detect the fat, protein, lactose, and total solids content in milk.

Description

Technical Field

[0001] This invention relates to the field of agricultural product quality testing, and in particular to a method and apparatus for analyzing milk components based on ultraviolet / visible spectroscopy. Background Technology

[0002] Milk is rich in nutrients and is one of the most important sources of nutrition for humans. Its commercial and nutritional value is determined by its fat, protein, lactose, and total solids content. Therefore, monitoring milk composition has become a crucial part of the daily management of modern dairy plants. Traditional methods for detecting milk composition, such as Soxhlet extraction for fat content, Kjeldahl method for protein content, high-performance liquid chromatography for lactose content, and drying method for total solids content, all suffer from high costs and large instrument sizes. Therefore, it is essential to develop a low-cost, portable method and device for milk composition analysis.

[0003] Chinese Patent Publication No. CN101769866A, published on July 7, 2010, entitled "A Device and Method for Detecting Milk Components," discloses a device and method for detecting milk components. The method is characterized by: "a model-building step and a milk detection step. The model-building step is used to establish and store a near-infrared spectral classification and correction model, and to classify the near-infrared spectra of different types of milk using ultrasonic self-excited frequencies and establish near-infrared spectral classification and correction models corresponding to the different types of milk. The milk detection step obtains the component content of the milk sample based on the ultrasonic self-excited frequency and near-infrared spectral information of the milk sample, corresponding to the near-infrared spectral classification and correction model." The "milk component detection device includes a model-building module and a milk detection module, characterized in that the model-building module is used to establish and store a near-infrared spectral classification and correction model, and to classify the near-infrared spectra of different types of milk using ultrasonic self-excited frequencies." The invention classifies near-infrared spectra and establishes near-infrared spectral classification correction models corresponding to different types of milk. The milk detection module is used to obtain the component content of the milk sample according to the ultrasonic self-excitation frequency and near-infrared spectral information of the milk sample, corresponding to the near-infrared spectral classification correction model. The detection wavelength range is 900-1700nm. This patent "achieves rapid, accurate, automatic and efficient detection of conventional milk components". However, due to the limited detection wavelength range and the need to simultaneously acquire ultrasonic and spectral signals, the "milk component detection device and method" described in this patent has the problems of high cost and non-portability of the instrument. From the perspective of detection principle, the basic principle of the near-infrared spectroscopy technology used in this invention is the frequency doubling absorption of near-infrared light by various components in milk, while the basic principle of the ultraviolet / visible spectroscopy used in this invention is that the substance to be tested undergoes atomic energy level transitions after being irradiated by ultraviolet / visible light. The two are fundamentally different.

[0004] Chinese Patent Publication No. CN107389599A, published on November 24, 2017, entitled "A Near-Infrared Diffuse Reflectance Spectrometer for Milk Component Detection," discloses a near-infrared diffuse reflectance spectrometer for milk component detection. Its features include: "the instrument comprises a light source system, a spectroscopic system, an integrating sphere sample cell, a detector, a data acquisition unit, an embedded system, and a power supply device," and "the light source in the light source system uses a tungsten bromide lamp, which has high intensity, stable performance, and a spectrum covering the entire near-infrared spectral region." "The detector uses..." The patent, which uses an uncooled single-point extended InGaAs detector, "predicts the protein, fat, and lactose content in milk using a multiple linear regression method." However, the invention is costly and bulky due to the use of an integrating sphere and an InGaAs detector in the detection device. From the perspective of detection principle, the basic principle of the near-infrared diffuse reflection technology used in the patent is the frequency doubling absorption of near-infrared light by various components in milk, while the basic principle of the ultraviolet / visible spectroscopy used in this patent is that the substance to be tested undergoes atomic energy level transitions after being irradiated by ultraviolet / visible light. The two are fundamentally different.

[0005] In summary, existing technologies suffer from high costs and lack of portability. Summary of the Invention

[0006] The technical problem to be solved is to provide a low-cost, portable method and apparatus for analyzing milk components.

[0007] This invention is achieved through the following technical solution:

[0008] A milk component analysis device based on ultraviolet / visible spectroscopy is characterized by comprising a display, a housing, cuvettes, a light source, a collimating lens, a cuvette holder, a spectral detection module, a power supply module, and a computational control module.

[0009] The display, light source, cuvette holder, spectral detection module, power supply module, and computing control module are all fixed to the outer casing for easy portability;

[0010] The display, the spectrum detection module, and the power supply module are connected to the computing control module via cables.

[0011] The light source is connected to the cuvette holder by a collimating lens, and the collimating lens is at the same horizontal height as the light source;

[0012] The sensor of the spectral detection module is a silicon-based multi-channel array detector, and its spectral acquisition wavelength range is 183-650 nm in the ultraviolet / visible band.

[0013] The cuvette holder and the spectral detection module are connected via optical fiber;

[0014] The light source is connected to the power supply module by a cable;

[0015] The display is mounted on the top of the housing;

[0016] The outer casing is provided with heat dissipation and ventilation holes;

[0017] The milk component analysis device based on ultraviolet / visible spectroscopy is characterized in that its length, width, and height are no greater than 30 cm, 20 cm, and 15 cm, respectively.

[0018] The method for analyzing milk components based on ultraviolet / visible spectroscopy is characterized in that ultraviolet / visible light emitted by a light source is transmitted through a collimating lens to a cuvette containing milk diluted 100 times with deionized water and interacts with the milk. The spectral detection module receives the transmitted light signal, converts it into a digital signal, and then transmits it to the calculation and control module. The calculation and control module calculates the fat, protein, lactose, and total solids content of the milk and feeds it back to the user through a display. The power supply module supplies power to the milk component analysis device based on ultraviolet / visible spectroscopy.

[0019] The method for analyzing milk components using ultraviolet / visible spectroscopy includes the following steps:

[0020] (1) Sample preparation: Take 1 mL of milk sample and pour it into a beaker containing 100 mL of deionized water. Stir well to obtain diluted milk sample.

[0021] (2) Set the reference spectrum: preheat the milk component analysis device based on ultraviolet / visible spectroscopy for 20 minutes, turn off the light source (4), and collect and save the dark spectrum. T d Turn on the light source, place the cuvette (3) containing deionized water into the cuvette holder (6), and collect and save the reference spectrum. T w ;

[0022] (3) Collect the UV / Vis absorbance spectrum of the milk sample to be tested: Add the diluted milk sample to the cuvette (3), place the cuvette (3) in the cuvette holder (6), and collect and calculate the transmission spectrum of the milk sample to be tested. T s The UV / Vis absorbance spectrum of the milk sample to be tested was calculated.

[0023] The ultraviolet / visible absorbance spectrum of the milk A The calculation method is as follows: A =log( T w - T d ) / ( Ts - T d );

[0024] The ultraviolet / visible absorbance spectrum of the milk A The wavelength range is 183-650 nm;

[0025] (4) Analyze the milk composition of the milk sample to be tested: calculate the fat, protein, lactose and total solids content of the milk using partial least squares regression algorithm;

[0026] The partial least squares regression algorithm is calculated using the following formula:

[0027] Y =B· A +c

[0028] In the formula Y This is the algorithm's output value, a four-dimensional matrix. Each dimension represents the fat, protein, lactose, and total solids content of the milk sample being tested, respectively. B is the coefficient matrix. A is the UV / Vis absorbance spectral matrix of the milk sample to be tested, where c is a constant term;

[0029] The method for determining the coefficient matrix B and the constant term c in the algorithm includes the following steps:

[0030] a. Sample preparation: Take more than 100 milk samples with a wide range of milk component content distribution. Take 1 mL of each sample and pour it into a beaker containing 100 mL of deionized water. Stir thoroughly to obtain diluted milk samples.

[0031] b. Data acquisition: The ultraviolet / visible absorbance spectrum of each milk sample was acquired using the ultraviolet / visible spectroscopy-based milk component analysis device, and the true values ​​of fat, protein, lactose and total solids content of each milk sample were measured using standard methods.

[0032] c. Determine model parameters B and c: Using the UV / Vis absorbance spectrum of the milk sample and the true values ​​of fat, protein, lactose and total solids content measured by standard methods as inputs, calculate the coefficient matrix B and constant term c using the partial least squares regression algorithm;

[0033] (5) Display the calculation results.

[0034] The detection principle of this invention is:

[0035] When various substances in milk are irradiated with ultraviolet / visible light, atomic energy level transitions and scattering phenomena occur. This interaction between ultraviolet / visible light and the substances in milk at the microscopic level is reflected in the macroscopic ultraviolet / visible absorbance spectrum. Therefore, by measuring the ultraviolet / visible absorbance spectrum of a milk sample, the content of its various components can be calculated. The milk component analysis method and device based on ultraviolet / visible spectroscopy described in this invention can automatically measure the ultraviolet / visible absorbance spectrum of a milk sample and calculate the content of fat, protein, lactose, and total solids in the milk.

[0036] Compared with the prior art, the advantages of the present invention are as follows:

[0037] a. The sensor of the spectral detection module of this invention uses a silicon-based multi-channel array detector, which is low in cost;

[0038] b. Most of the components of this invention are integrated inside a portable milk composition analysis device, which is small in size and easy to carry.

[0039] Figure Labels

[0040] 1. Display; 2. Housing; 3. Cuvette; 4. Light source; 5. Collimating lens; 6. Cuvette holder; 7. Spectral detection module; 8. Power supply module; 9. Calculation and control module. Attached Figure Description

[0041] Figure 1 This is a structural diagram of the present invention.

[0042] Figure 2 This is the main interface of the milk composition analysis software.

[0043] Figure 3 This is the main interface of the spectrum acquisition software.

[0044] Figure 4 These are UV / Vis absorbance spectra of milk samples with different total solids contents.

[0045] Figure 5 It is a graph showing the relationship between the measured and actual total solids content. Detailed Implementation

[0046] The present invention will be further described below with reference to a preferred embodiment and accompanying drawings:

[0047] like Figure 1The diagram shows the structural composition of a milk component analysis device based on ultraviolet / visible spectroscopy. The device includes a display 1, a housing 2, cuvettes 3, a light source 4, a collimating lens 5, a cuvette holder 6, a spectral detection module 7, a power supply module 8, and a computational control module 9. The display 1, light source 4, cuvette holder 6, spectral detection module 7, power supply module 8, and computational control module 9 are all fixed to the housing 2 for easy portability. The display 1, spectral detection module 7, and power supply module 8 are connected to the computational control module 9 via cables. The light source 4 is connected to the cuvette holder 6 via the collimating lens 5, which is at the same horizontal height as the light source 4. The sensor of the spectral detection module 7 uses a silicon-based multi-channel array detector, with a spectral acquisition wavelength range of 183-650 nm in the ultraviolet / visible band. The cuvette holder 6 and spectral detection module 7 are connected via optical fiber. The light source 4 is connected to the power supply module 8 via a cable. The display 1 is mounted on the top of the housing 2. The housing 2 has ventilation holes for heat dissipation. The length, width, and height of the milk component analysis device based on ultraviolet / visible spectroscopy are no greater than 30 cm, 20 cm, and 15 cm, respectively. cm.

[0048] The display features a 3.5-inch LCD touchscreen.

[0049] The outer casing is made of plastic.

[0050] The cuvette is made of quartz and has an optical path of 2 mm.

[0051] The light source is a DD2.5TZ type deuterium lamp.

[0052] The power supply module uses the WXDY01302.5 type deuterium lamp dedicated power supply.

[0053] The computing control module uses a Raspberry Pi 3B+.

[0054] Figure 2 This is the main interface of the milk composition analysis software, which can automatically measure the fat, protein, lactose, and total solids content of the milk sample to be tested.

[0055] Figure 3 This is the main interface of the spectral acquisition software. This software controls the milk component analysis device based on ultraviolet / visible spectroscopy to acquire the ultraviolet / visible absorbance spectrum of milk samples, which is used to determine the coefficient matrix B and constant term c in the milk component analysis algorithm.

[0056] The method for analyzing milk components based on ultraviolet / visible spectroscopy includes the following steps:

[0057] (1) Sample preparation: Take 1 mL of milk sample and pour it into a beaker containing 100 mL of deionized water. Stir well to obtain diluted milk sample.

[0058] (2) Setting the reference spectrum: Preheat the milk component analysis device based on ultraviolet / visible spectroscopy for 20 minutes, turn off the light source 4, and turn on the reference spectrum 5. Figure 2 In the medium-spectral acquisition software, click "Updata DarkSpec" to acquire and save the dark spectrum. T d Turn on the light source 4, place the cuvette 3 containing deionized water into the cuvette holder 6, and click "Updata RefSpec" to collect and save the reference spectrum. T w ;

[0059] (3) Collect the UV / Vis absorbance spectrum of the milk sample to be tested: Add the diluted milk sample to the cuvette 3, place the cuvette 3 in the cuvette holder 6, and click the “Analysis” button. Figure 2 The milk composition analysis software shown automatically collects and calculates the spectrum of the milk sample to be tested. T s Then, the UV / Vis absorbance spectrum of the milk sample to be tested is calculated;

[0060] Milk UV / Vis absorbance spectrum A The calculation method is as follows: A =log( T w - T d ) / ( T s - T d );

[0061] Milk UV / Vis absorbance spectrum A The wavelength range is 183-650 nm;

[0062] (4) Analyze the milk composition of the milk sample to be tested: After clicking the “Analysis” button, Figure 2 The milk composition analysis software shown automatically calculates the fat, protein, lactose, and total solids content of milk using partial least squares regression algorithm.

[0063] The partial least squares regression algorithm is calculated using the following formula:

[0064] Y =B· A +c

[0065] In the formula Y This is the algorithm's output value, a four-dimensional matrix. Each dimension represents the fat, protein, lactose, and total solids content of the milk sample being tested, respectively. B is the coefficient matrix. A is the UV / Vis absorbance spectral matrix of the milk sample to be tested, where c is a constant term;

[0066] The method for determining the coefficient matrix B and the constant term c in the algorithm includes the following steps:

[0067] a. Sample preparation: Take 200 milk samples with a wide range of milk component content distribution, take 1 mL of each sample, pour it into a beaker containing 100 mL of deionized water, stir thoroughly to obtain diluted milk samples;

[0068] b. Data Acquisition: Open Figure 3 After setting the parameters in the spectral acquisition software shown, click "Get Spec" and "Save Spec" in sequence to acquire the UV / Vis absorbance spectrum of each milk sample. Use the MilkScan FT1 milk component analyzer to measure the true values ​​of fat, protein, lactose and total solids content of each milk sample.

[0069] c. Determine model parameters B and c: Using the UV / Vis absorbance spectrum of the milk sample and the true values ​​of fat, protein, lactose, and total solids content measured by MilkScan FT1 as input values, calculate the coefficient matrix B and the constant term c using the partial least squares regression algorithm. Then, import the calculated coefficient matrix B and constant term c into... Figure 2 The milk composition analysis software shown is equipped with the ability to analyze the content of fat, protein, lactose and total solids in cow's milk.

[0070] (5) Display the calculation results: such as Figure 2 As shown, the detected levels of fat, protein, lactose, and total solids in the milk are displayed on screen 1.

[0071] Figure 4 The images show the UV / Vis spectra of milk samples with different total solids content measured using the UV / Vis spectroscopy-based milk component analysis device described in this embodiment. In the 240-560nm range, the total solids content of the milk samples increases with the increase of their absorbance spectrum.

[0072] Figure 5 This is a graph showing the relationship between the measured and true values ​​of total solids content measured using the ultraviolet / visible spectroscopy-based milk component analysis device described in this embodiment. The data points are evenly distributed around the 45° line on the coordinate axis, and the root mean square error (RMSE) between the two is 0.29, indicating that the ultraviolet / visible spectroscopy-based milk component analysis device described in this embodiment has good accuracy in determining the total solids content in milk.

[0073] Furthermore, the root mean square errors for measuring fat, protein, and lactose in milk in this embodiment are 0.20%, 0.07%, and 0.06%, respectively.

[0074] The above embodiments are merely illustrative examples of the present invention and do not constitute a limitation on the scope of protection of the present invention. Any designs that are the same as or similar to the present invention are within the scope of protection of the present invention.

Claims

1. A method for analyzing milk components based on ultraviolet / visible spectroscopy, characterized in that, The ultraviolet / visible light emitted by the light source is transmitted through a collimating lens to a cuvette containing milk diluted 100 times with deionized water and interacts with the milk. The spectral detection module receives the transmitted light signal, converts it into a digital signal, and then transmits it to the calculation and control module. The calculation and control module calculates the fat, protein, lactose, and total solids content of the milk and feeds it back to the user through a display. The power module supplies power to the milk component analysis device based on ultraviolet / visible spectroscopy. The method for analyzing milk components based on ultraviolet / visible spectroscopy includes the following steps: (1) Sample preparation: Take 1 mL of milk sample and pour it into a beaker containing 100 mL of deionized water. Stir well to obtain diluted milk sample. (2) Set the reference spectrum: preheat the milk component analysis device based on ultraviolet / visible spectroscopy for 20 minutes, turn off the light source (4), and collect and save the dark spectrum. T d Turn on the light source, place the cuvette (3) containing deionized water into the cuvette holder (6), collect and save the reference spectrum. T w ; (3) Collect the UV / Vis absorbance spectrum of the milk sample to be tested: Add the diluted milk sample to the cuvette (3), place the cuvette (3) in the cuvette holder (6), and collect and calculate the transmission spectrum of the milk sample to be tested. T s The UV / Vis absorbance spectrum of the milk sample to be tested was calculated. A ; The ultraviolet / visible absorbance spectrum of the milk A The calculation method is as follows: A =log( T w - T d ) / ( T s - T d ); The ultraviolet / visible absorbance spectrum of the milk A The wavelength range is 183-650 nm; (4) Analyze the milk composition of the milk sample to be tested: calculate the fat, protein, lactose and total solids content of the milk using partial least squares regression algorithm; The partial least squares regression algorithm is calculated using the following formula: Y =B· A +c In the formula Y This is the algorithm's output value, a four-dimensional matrix. Each dimension represents the fat, protein, lactose, and total solids content of the milk sample being tested, respectively. B is the coefficient matrix. A is the UV / Vis absorbance spectral matrix of the milk sample to be tested, where c is a constant term; The method for determining the coefficient matrix B and the constant term c in the algorithm includes the following steps: a. Sample preparation: Take more than 100 milk samples with a wide range of milk component content distribution. Take 1 mL of each sample and pour it into a beaker containing 100 mL of deionized water. Stir thoroughly to obtain diluted milk samples. b. Data acquisition: The ultraviolet / visible absorbance spectrum of each milk sample was acquired using the ultraviolet / visible spectroscopy-based milk component analysis device, and the true values ​​of fat, protein, lactose and total solids content of each milk sample were measured using standard methods. c. Determine model parameters B and c: Using the UV / Vis absorbance spectrum of the milk sample and the true values ​​of fat, protein, lactose and total solids content measured by standard methods as inputs, calculate the coefficient matrix B and constant term c using the partial least squares regression algorithm; (5) Display the calculation results.

2. A milk component analysis device based on ultraviolet / visible spectroscopy, characterized in that: The method for implementing claim 1 includes a display (1), a housing (2), a cuvette (3), a light source (4), a collimating lens (5), a cuvette holder (6), a spectral detection module (7), a power supply module (8), and a computational control module (9). The display (1), light source (4), cuvette holder (6), spectral detection module (7), power supply module (8), and operation control module (9) are all fixed on the outer shell (2) for easy carrying; The display (1), the spectrum detection module (7), and the power supply module (8) are connected to the operation control module (9) via cables; The light source (4) is connected to the cuvette holder (6) by a collimating lens (5), and the collimating lens is at the same horizontal height as the light source (4); The sensor of the spectral detection module (7) is a silicon-based multi-channel array detector, and its spectral acquisition wavelength range is 183-650 nm in the ultraviolet / visible band. The cuvette holder (6) and the spectral detection module (7) are connected by optical fiber; The light source (4) is connected to the power supply module (8) by a cable; The display (1) is mounted on the top of the housing (2); The outer shell (2) has heat dissipation and ventilation holes.

3. The milk component analysis device based on ultraviolet / visible spectroscopy according to claim 2, characterized in that: The length, width, and height shall not exceed 30 cm, 20 cm, and 15 cm, respectively.

Images