CN103163047A - Method for detecting colony total amount and viscosity of unpackaged milk - Google Patents

Abstract

The invention discloses a method for detecting the total number of bacterial colonies and viscosity of milk under the condition of no encapsulation. Its steps are as follows: 1) Use the electrode array of the detector to react with milk under non-encapsulated conditions of different storage times to obtain a response signal; 2) Perform the detection of the total number of colonies and viscosity of the tested samples; 3) Test the detected electrodes The response signal of the array is used to extract the sample eigenvalue; 4) The partial least squares regression and support vector machine prediction model are established with the sample eigenvalue as an independent variable, and the total number of colonies and viscosity values of milk under non-encapsulated conditions of different storage times are quantified predictive analytics. The invention provides a simple and feasible detection method aiming at the change of total bacterial colony and viscosity of milk under the condition of no encapsulation, and the method has relatively high sensitivity, reliability and repeatability.

Description

A method for detecting the total number of colonies and viscosity of milk under the condition of no encapsulation

technical field

The invention relates to a detection method, in particular to a detection method for the total number of colonies and viscosity of milk under the condition of no packaging.

Background technique

Milk is a very nutritious food, which is rich in lactose, amino acids, vitamins, inorganic salts and trace elements. In addition, the high water activity and neutral pH make milk a natural medium for the growth of various microorganisms.

With the accelerated pace of life, people usually buy milk in large packages (>1L), and these large packages of milk are usually consumed within 1-2 weeks. As a result, these opened packages of milk can become contaminated with airborne bacteria when stored in the refrigerator. Moreover, some heat-resistant bacteria that cannot be killed during milk production will also regain activity during storage. These bacteria ultimately lead to changes in the physical, chemical and microbiological properties of the milk. In past studies, chemical, microbiological and sensory evaluation methods were often used to detect changes in the total bacterial count and viscosity of milk. However, these techniques require rigorous training and complex preprocessing. Among the current analysis techniques, gas chromatography-mass spectrometry, high-performance liquid chromatography and capillary electrophoresis are often used to detect changes in the total number of colonies and viscosity of milk, but these instruments are expensive, complicated and time-consuming.

Contents of the invention

The purpose of the invention is to overcome the deficiencies of the prior art and provide a method for detecting the total number of colonies and viscosity of milk under the condition of no packaging.

The total number of milk colonies and viscosity detector used in the present invention obtains the overall information of the liquid sample through the electrode array, and combines the chemometrics method to perform qualitative and quantitative analysis on the physical and chemical characteristics of the sample, thus eliminating the need for chemical analysis At the same time, the sample does not need to be pre-treated, and rapid detection can be realized, and it has the characteristics of high sensitivity, reliability, and repeatability.

The steps of the method for detecting the total number of colonies and viscosity of milk under non-encapsulation conditions are as follows:

1) After sampling the milk with a beaker, insert the electrode array of the detector into the milk sample, and use the electrode array of the detector to react with milk under unencapsulated conditions of different storage times to obtain a response signal; the electrode array reacts with the milk to obtain a response signal indicator What is more important is the current signal generated when the milk sample undergoes oxidation and reduction reactions on the electrode surface under the action of the external rectangular multi-frequency pulse and trapezoidal multi-frequency pulse. Since the detector only receives the voltage signal, the current signal is converted by the current-to-voltage circuit. For the voltage signal; the detector used includes a data acquisition part, a signal processing part and a pattern recognition part: the data acquisition part has a swing arm that can drive the electrode array to move up and down and a sample injector that can place a sample beaker, and the signal processing part Part is a potentiostat, its core component is a data acquisition card, and the pattern recognition part is mainly a data analysis model; the electrode array used is one or more of gold, silver, platinum, palladium, tungsten, copper, and titanium;

2) For the tested samples, the plate count method is used to determine the total number of colonies, and the rheometer is used to measure the viscosity value;

3) Combine the eigenvalues based on the rectangular multi-frequency AM pulse and the trapezoidal multi-frequency AM pulse to obtain the sample eigenvalue; the eigenvalue refers to the area between the electrode response signal curve and the X-axis;

4) Establish partial least squares regression and support vector machine forecasting models with sample eigenvalues as independent variables, and at the same time use the cross-validation method to determine the forecasting parameters gam and sig ² of the support vector machine. Quantitative predictive analysis of the total number of colonies and viscosity values.

Beneficial effects of the present invention: the present invention can accurately detect the total number of colonies and viscosity of milk under non-encapsulation conditions with ordinary electrode arrays, and compared with traditional instruments, the detection cost is reduced, the detection steps are simplified, and the detection efficiency is improved. Use qualitative and quantitative pattern recognition systems to process the fingerprint information of milk under non-encapsulated conditions with different storage times, and accurately convert the fingerprint information of milk under non-encapsulated conditions with different storage times into results that are consistent with the detection results of conventional monitoring methods . The electrode array can detect the information of milk under non-encapsulation conditions with different storage times, and bring these fingerprint information into the mathematical model, and predict the total number of colonies and viscosity of milk under non-encapsulation conditions according to the calculated values of the model.

Description of drawings

Fig. 1 (a) Rectangular multi-frequency pulse sweep voltage adopted by the detector based on electrode array.

Fig. 1 (b) The trapezoidal multi-frequency pulse scanning voltage adopted by the detector based on the electrode array.

Fig. 2 (a) Response signal curve of milk sample to rectangular multi-frequency pulse scanning voltage.

Fig. 2 (b) The response signal curve of the milk sample to the trapezoidal multi-frequency pulse scanning voltage.

Fig. 3 Rheological properties of milk samples at 0 hours.

Fig. 4 (a) The eigenvalue selection of the response signal curve of the rectangular multi-frequency pulse sweep voltage.

Fig. 4 (b) The eigenvalue selection of the response signal curve of the trapezoidal multi-frequency pulse sweep voltage.

Figure 5 (a) The example diagram of the present invention is based on the prediction results of the partial least squares model of the viscosity values of seven milk samples with different storage times.

Figure 5 (b) The present invention is based on the prediction results of the support vector machine model for testing the viscosity values of seven milk samples with different storage times.

Figure 6 (a) The present invention is based on the prediction results of the partial least squares model of the total number of colonies of seven milk samples tested for different storage times.

Fig. 6 (b) The example diagram of the present invention is based on the prediction results of the partial least squares model of the total number of colonies of milk samples tested for seven different storage times.

Detailed ways

The present invention will be further described below in conjunction with the accompanying drawings and embodiments. The detector based on the electrode array used in the present invention includes a data acquisition part, a signal processing part and a pattern recognition part. The data acquisition part has a swing arm that can drive the electrode array to move up and down, and can place a sample injector equipped with a sample beaker. The signal processing part is a potentiostat, and its core component is a data acquisition card. The electrode array used is one or more of gold, silver, platinum, palladium, tungsten, copper and titanium. The electrode array is connected with the data acquisition card through the current/voltage conversion circuit. The data acquisition card generates a pulse signal. Under the action of the pulse signal, the electrode array reacts with the milk to obtain the fingerprint information of the total number of milk colonies and viscosity. The fingerprint information refers to the voltage obtained when the milk test sample undergoes redox reaction on the surface of the electrode. Signal.

An electrode array-based detector assay is performed on the sample to be detected. Measure the sample into a beaker and let it stand for 15 minutes. After the sample is stable, place the electrode array in the sample. Under the action of the pulse signal, the electrode array reacts with the milk to obtain the fingerprint of the total number of colonies and viscosity of the milk under the condition of no encapsulation Information, the fingerprint information is converted into a digital signal by the acquisition card and input to the computer. The instrument is equipped with a constant temperature water bath device, which can keep the sample at a constant temperature (±0.1°C) during the test.

Use the computer to extract the eigenvalues and pattern recognition of the obtained data, and use qualitative and quantitative methods such as: principal component analysis, cluster analysis, least squares regression and support vector machine. Through these pattern recognition systems, the fingerprint information of milk samples and the mathematical model of the total number of milk colonies and viscosity under non-encapsulation conditions are established: the principal component analysis and cluster analysis pattern recognition models are established with the sample characteristic values as independent variables for non-encapsulation of different storage times Qualitative difference analysis of milk under different conditions; Partial least squares regression and support vector machine prediction models were established with sample eigenvalues as independent variables to quantitatively predict and analyze the total number of colonies and viscosity values of milk under non-encapsulated conditions for different storage times.

Example

Now introduce the implementation process of the present invention in detail in conjunction with examples. Example is to utilize the present invention to detect and predict the total number of sample colonies and the viscosity of the milk sample under the non-encapsulation condition of different storage times, wherein principal component analysis, cluster analysis, partial least squares regression and support vector machine model are all by software MATLAB R2008a (MathWorks, USA) completed. The test samples were milk samples of seven different storage times, and the milk was sampled according to 0, 12, 24, 36, 48, 60 and 72 hours after opening the package. The detection process for milk can be as follows:

Put the milk on the day of delivery in a 4°C environment for 3 hours and then open the package. Samples were taken at 0 (immediately after opening the package), 12, 24, 36, 48, 60 and 72 hours after opening the package. During the milk sampling process, 650mL of milk samples were taken for each storage time, divided into 26 beakers, and 25ml of samples were contained in each beaker. Place the beakers containing the samples in the injector and drain a beaker of deionized water after each cup of sample.

After the sample was left to stand for 15 minutes, it reached stability, and the swing arm was lowered to drive the electrode array into the sample. At the same time, the data acquisition card generates rectangular multi-frequency pulses and trapezoidal multi-frequency pulses (Figure 1). Under the sequential action of the two pulses, the sample reacts with the electrode S1 in the electrode array and obtains the corresponding wine age fingerprint information. The fingerprint information It is converted into digital by the data acquisition card and input to the computer, and the acquisition stops after 1.11s. Under the action of the relay array, the electrodes S2-S4 react with the sample in turn under the action of the pulse signal. After the reaction between S4 and the sample, the pendulum is lifted Grind the electrode with fine sandpaper with abrasive powder for 5 seconds, then turn the injector, turn the beaker containing deionized water to the original sample position, then lower the swing arm to insert the electrode into the deionized water The electrode is cleaned for 15 seconds in order to measure the next sample. Repeat the above steps for multiple measurements. As shown in Figure 2, in the example of the present invention, the fingerprint information obtained by the electrode from the milk sample under the action of different pulses, the abscissa is the sampling time, and the ordinate is the voltage value. The oxidation-reduction reaction generates a current signal, and the current signal becomes a voltage signal after passing through a current/voltage conversion circuit.

Collect samples after milk fingerprint information collection, and measure the microbial content of milk by plate counting method. After 1ml of milk samples are diluted 1 times, 10 times and 100 times respectively, there are three repetitions for each dilution, using agar Put it as a culture medium in an environment at 37°C for 48 hours to count the number of bacterial communities; use a rheometer to detect the milk viscosity value, choose a rheometer with shear rate control mode, and set the range of shear rate to 2- 200S ^-1 , select the spindle rotor as the propeller, set the program of the rheometer, the detection of each sample is divided into three processes: the first process is the sample mixing process, the time is set to 320 seconds, the rotor The speed change is 2-120S ^-1 ; the second process is the sample stabilization process, the time is set to 120 seconds, and the speed of the rotor is 0; the third process is the sample detection process, the time is set to 10 minutes, The speed of the rotor varies from 2-200S ^-1 . The sample temperature was controlled at 25 °C by a constant temperature water bath device. As shown in Fig. 3, the average value of the viscosity values when the shear rate is 150S ^-1 -200S ^-1 is selected as the viscosity value of the milk sample.

As shown in Figure 4, the computer is used to select and extract the characteristic value of the detector response signal, and the area between the electrode response signal curve and the X axis is selected as the characteristic value. The eigenvalues of the samples are combined as the sample eigenvalues; each sample corresponds to 4 (4 electrodes) × 2 (2 kinds of scanning voltages) = 8 eigenvalues.

The partial least squares regression and support vector machine prediction models were established with sample eigenvalues as independent variables, and the quantitative prediction analysis was carried out on the total number of colonies and viscosity of milk under non-encapsulated conditions for different storage times. The prediction parameters of the support vector machine determined by the cross-validation method to the total number of colonies and the viscosity value are respectively: gam=418.1165, sig ² =0.854 and gam=455.1312, sig ² =4.6137.

Its predictive ability is shown in Figure 5 and Figure 6. The correlation coefficients between the predicted value and the actual value of the total number of colonies and viscosity of milk under the condition of no packaging by the partial least squares regression model are R ² = 0.9444 and R ² respectively = 0.9894; the correlation coefficients between the predicted value and the actual value of the total number of colonies and viscosity of milk under the condition of no packaging by the support vector machine were R ² = 0.9864 and R ² = 0.9994, respectively.

Claims (2)

1. the detection method without milk total plate count and viscosity under the encapsulation condition, is characterized in that, its step is as follows:

1) electrod-array of detector is inserted into the different resting periods without in the milk sample under the encapsulation condition, utilize electrod-array to the response signal of milk; Described electrod-array refers to the response signal of milk and is adding under the scanning voltage effect, the electric signal that milk sample produces when electrode surface generation redox reaction; The described scanning voltage that adds is respectively rectangle multifrequency amplitude-modulated pulse and trapezoidal multifrequency amplitude-modulated pulse;

2) milk sample that detected is carried out total plate count and viscosity number detection; Total plate count detects and adopts colony counting method, viscosity number to detect the employing rheometer measurement;

3) will obtain the sample characteristic value after the eigenwert merging based on rectangle multifrequency amplitude-modulated pulse and trapezoidal multifrequency amplitude-modulated pulse; The electrod-array that described eigenwert refers to is to the area between the response signal of milk and coordinate axis X; Described coordinate axis X refers to that electrod-array is to the response signal collection point number of milk;

4) set up partial least squares regression and SVM prediction model take the sample characteristic value as independent variable, total plate count and the viscosity number without milk under the encapsulation condition of different resting periods carried out the quantitative forecast analysis.

2. a kind of detection method without milk total plate count and viscosity under the encapsulation condition according to claim 1, is characterized in that described step 2) in, the variation range of the shear rate of the rotor of flow graph is set as 2-200S ^-1, sample frequency is 12 Min ^-1, the sampling time is 10 minutes, and to get shear rate be 150-200S ^-1The time the average of viscosity number as the viscosity number of milk sample.

Images