Coffee bean producing area tracing method

Abstract

The invention discloses a coffee bean origin tracing method. The method comprises the following steps: obtaining coffee bean samples from different producing areas, and preparing a coffee solution corresponding to each coffee bean sample; preparing NaOH solutions with different concentrations, uniformly mixing the NaOH solution with each concentration with the coffee solution, analyzing, and screening to obtain the optimal concentration of the NaOH solution; uniformly mixing the coffee solution corresponding to each coffee bean sample with the NaOH solution with the optimal concentration, and extracting a corresponding characteristic signal set; inputting the characteristic signal set and the origin label of the coffee bean sample into a machine learning model for training to obtain a coffee bean origin traceability model; preparing a coffee solution corresponding to a to-be-detected coffee bean sample, uniformly mixing the coffee solution with the NaOH solution with the optimal concentration, extracting a corresponding characteristic signal set, inputting the characteristic signal set into the coffee bean production place traceability model, and outputting the production place of the to-be-detected coffee bean sample by the coffee bean production place traceability model. According to the invention, the coffee bean producing area can be traced quickly and accurately.

Description

Technical Field

[0001] This invention relates to the field of food traceability technology, and in particular to a method for tracing the origin of coffee beans. Background Technology

[0002] The origin of coffee beans is a key factor influencing their quality and value. Differences in environmental conditions such as climate, soil, and altitude between different origins lead to significant differences in flavor and composition. Currently, methods for tracing the origin of coffee beans mainly include sensory evaluation, spectroscopic analysis, chromatographic analysis, and mass spectrometry. Sensory evaluation relies on the experience of professional tasters, is highly subjective, has poor repeatability, and is difficult to standardize for traceability. While spectroscopic analysis, chromatographic analysis, and mass spectrometry offer high detection accuracy, they suffer from drawbacks such as bulky equipment, complex operation, long testing cycles, and high costs, failing to meet the needs for low-cost, rapid, and on-site traceability. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention provides a method for tracing the origin of coffee beans, which can quickly and accurately trace the origin of coffee beans, is simple to operate, and is inexpensive.

[0004] To solve the above problems, the present invention adopts the following technical solution: The present invention provides a method for tracing the origin of coffee beans, comprising the following steps: S1: Obtain coffee bean samples from different origins, prepare m samples of coffee beans from each origin, and prepare a coffee solution corresponding to each coffee bean sample; S2: Prepare NaOH solutions of different concentrations, mix each NaOH solution with coffee solution evenly, analyze the mixture, and screen to obtain the optimal concentration of NaOH solution. S3: Mix the coffee solution corresponding to each coffee bean sample with the optimal concentration of NaOH solution evenly, and extract the corresponding feature signal set; S4: Input the feature signal set and origin label of coffee bean sample into the machine learning model for training to obtain the coffee bean origin traceability model; S5: Prepare a coffee solution corresponding to the coffee bean sample to be tested, and mix it evenly with a NaOH solution of the optimal concentration. Extract the corresponding feature signal set, input the feature signal set into the coffee bean origin traceability model, and the coffee bean origin traceability model outputs the origin of the coffee bean sample to be tested.

[0005] Preferably, step S2 includes the following steps: S21: Prepare NaOH solutions of different concentrations, with the concentration range of NaOH solutions being 0.001~1 mol / L; S22: After uniformly mixing NaOH solution of each concentration with coffee solution, analyze the mixture and extract the characteristic value corresponding to each NaOH solution concentration. S23: The concentration with the largest characteristic value is taken as the optimal concentration of the NaOH solution.

[0006] Preferably, the method for uniformly mixing the NaOH solution of concentration d with the coffee solution in step S22 and then analyzing the mixture to extract the characteristic value corresponding to the NaOH solution of concentration d, where 0.001≤d≤1, includes the following steps: M1: Mix 1 ml of distilled water with 19 ml of NaOH solution of concentration d to obtain the corresponding mixed solution. Use a three-electrode sensor to collect the cyclic voltammetry curve of the mixed solution and use the cyclic voltammetry curve as the comparison curve. M2: 1 ml of each coffee solution was uniformly mixed with 19 ml of NaOH solution of concentration d to obtain the corresponding mixed solution. Cyclic voltammetry curves of each mixed solution were collected using a three-electrode sensor. Subtract each cyclic voltammetry curve from the comparison curve to obtain the corresponding net response current curve, and obtain the restored peak current value of each net response current curve. Calculate the average value of the reduction peak current , , Where n is the number of coffee bean producing regions. Let be the reduction peak current value corresponding to the j-th coffee bean sample from the i-th origin, where 1≤i≤n, 1≤j≤m; M3: 1 mL of 0.01 mol / L sucrose solution, glucose solution, and fructose solution were mixed with 19 mL of NaOH solution of concentration d to obtain the corresponding mixed solutions. Cyclic voltammetry curves of each mixed solution were collected using a three-electrode sensor. Subtract the control curve from each cyclic voltammetry curve to obtain the corresponding net response current curve, and obtain the reduction peak current value of each net response current curve. The reduction peak current value corresponding to the sucrose solution is denoted as... The reduction peak current value corresponding to the glucose solution is denoted as The reduction peak current value corresponding to the fructose solution is denoted as ; M4: Calculate the characteristic value corresponding to a NaOH solution with concentration d. The calculation formula is as follows: , , , , in, These are the eigenvalues.

[0007] , , The signal-to-noise ratios of the detection results for coffee with sucrose, glucose, and fructose solutions are shown below. Since the sugars in coffee, such as sucrose, glucose, and fructose, are electrochemically active in strongly alkaline solutions and can interfere with the origin characteristic signals, this invention uses the above method to find the optimal NaOH solution concentration that best highlights the coffee characteristic signals. This effectively suppresses the interference of these sugar components while maximizing the differences in origin characteristics, significantly improving the accuracy of traceability.

[0008] Preferably, the method for uniformly mixing the coffee solution corresponding to the coffee bean sample with a NaOH solution of optimal concentration and extracting the corresponding feature signal set includes the following steps: N1: Mix the coffee solution corresponding to 1 ml of coffee bean sample with 19 ml of NaOH solution of optimal concentration to obtain the corresponding mixed solution; N2: The mixed solution was characterized using a three-electrode sensor and differential pulse voltammetry to obtain the peak potential. Peak current Peak area , peak potential Peak current Peak area The set of characteristic signals that make up the coffee bean sample.

[0009] Preferably, the machine learning model is a support vector machine.

[0010] As a preferred method, the method for preparing the coffee solution corresponding to the coffee bean sample includes the following steps: grinding the coffee bean sample and passing it through an 80-100 mesh sieve, dissolving it in distilled water at 95°C at a solid-liquid ratio of 1:10 (g / mL), stirring evenly, and then allowing it to stand and filter to obtain the coffee solution.

[0011] Preferably, the three-electrode sensor includes a copper film working electrode, a saturated KCl calomel reference electrode, and a platinum counter electrode.

[0012] Preferably, the parameters of the cyclic voltammetry are: potential range -0.9 to -0.2V, scan rate of 50mV / s, and number of scans of 2.

[0013] Preferably, the parameters of the differential pulse voltammetry method are: potential range 0 to -0.5V, voltage step size 4mV, pulse amplitude 0.05V, pulse duration 0.05s, sampling interval 0.0167s, and pulse period 0.5s.

[0014] The beneficial effects of this invention are: (1) It can quickly and accurately trace the origin of coffee beans, and the operation is simple and the cost is low. (2) By optimizing the concentration of NaOH solution, the electrochemical response interference of sucrose, glucose and fructose in coffee is effectively suppressed, highlighting the differences in characteristic signals of coffee from different origins and significantly improving the accuracy of traceability. Attached Figure Description

[0015] Figure 1 This is a flowchart of an embodiment; Figure 2 These are the net response current curves of coffee solution under different concentrations of NaOH solution; Figure 3 The figures show the net response current curves for coffee solution, sucrose solution, glucose solution, and fructose solution in a 0.01 mol / L NaOH solution. Detailed Implementation

[0016] The technical solution of the present invention will be further described in detail below through embodiments and in conjunction with the accompanying drawings.

[0017] Example: This example illustrates a method for tracing the origin of coffee beans, such as... Figure 1 As shown, it includes the following steps: S1: Obtain coffee bean samples from different origins. Prepare m samples of coffee beans from each origin, where m > 1. Prepare a coffee solution corresponding to each coffee bean sample. S2: Prepare NaOH solutions of different concentrations, mix each NaOH solution with coffee solution evenly, analyze the mixture, and screen to obtain the optimal concentration of NaOH solution. S3: Mix the coffee solution corresponding to each coffee bean sample with the optimal concentration of NaOH solution evenly, and extract the corresponding feature signal set; S4: Input the feature signal set and origin label of coffee bean sample into the machine learning model for training to obtain the coffee bean origin traceability model; S5: Prepare a coffee solution corresponding to the coffee bean sample to be tested, and mix it evenly with a NaOH solution of the optimal concentration. Extract the corresponding feature signal set, input the feature signal set into the coffee bean origin traceability model, and the coffee bean origin traceability model outputs the origin of the coffee bean sample to be tested.

[0018] Step S2 includes the following steps: S21: Prepare NaOH solutions of different concentrations, with the concentration range of NaOH solutions being 0.001~1 mol / L; S22: After uniformly mixing NaOH solution of each concentration with coffee solution, analyze the mixture and extract the characteristic value corresponding to each NaOH solution concentration. S23: The concentration with the largest characteristic value is taken as the optimal concentration of the NaOH solution.

[0019] Step S22 involves uniformly mixing a NaOH solution of concentration d with a coffee solution and then analyzing the mixture to extract the characteristic value corresponding to the NaOH solution of concentration d, where 0.001 ≤ d ≤ 1. This includes the following steps: M1: Mix 1 ml of distilled water with 19 ml of NaOH solution of concentration d to obtain the corresponding mixed solution. Place the three-electrode sensor in the mixed solution and collect the cyclic voltammetry curve of the mixed solution using cyclic voltammetry. Use the cyclic voltammetry curve as the comparison curve. M2: Mix 1 ml of each coffee solution with 19 ml of NaOH solution of concentration d to obtain the corresponding mixed solution. Place the three-electrode sensor in the mixed solution and use cyclic voltammetry to collect the cyclic voltammetric curve of each mixed solution. Subtract the comparison curve from each cyclic voltammetry curve (i.e., subtract the comparison curve from each cyclic voltammetry curve) to obtain the corresponding net response current curve, and obtain the restored peak current value of each net response current curve. Calculate the average value of the reduction peak current , , Where n is the number of coffee bean producing regions. Let be the reduction peak current value corresponding to the j-th coffee bean sample from the i-th origin, where 1≤i≤n, 1≤j≤m; For example: 1 ml of coffee solution corresponding to the j-th coffee bean sample from the i-th origin is uniformly mixed with 19 ml of NaOH solution of concentration d to obtain the corresponding mixed solution. A cyclic voltammetry curve of this mixed solution is acquired using a three-electrode sensor. The control curve is subtracted from this cyclic voltammetry curve to obtain the corresponding net response current curve. The reduction peak current value of this net response current curve is then obtained. , It is the reduction peak current value corresponding to the j-th coffee bean sample from the i-th origin; M3: 1 mL of 0.01 mol / L sucrose solution, glucose solution, and fructose solution were mixed with 19 mL of NaOH solution of concentration d to obtain the corresponding mixed solutions. Cyclic voltammetry curves of each mixed solution were collected using a three-electrode sensor. Subtract the control curve from each cyclic voltammetry curve (i.e., subtract the control curve from each cyclic voltammetry curve) to obtain the corresponding net response current curve, and obtain the reduction peak current value of each net response current curve. The reduction peak current value corresponding to the sucrose solution is denoted as... The reduction peak current value corresponding to the glucose solution is denoted as The reduction peak current value corresponding to the fructose solution is denoted as ; M4: Calculate the characteristic value corresponding to a NaOH solution with concentration d. The calculation formula is as follows: , , , , in, These are the eigenvalues.

[0020] , , The signal-to-noise ratios of the detection results for coffee with sucrose, glucose, and fructose solutions are shown below. Since the sugars in coffee, such as sucrose, glucose, and fructose, are electrochemically active in strongly alkaline solutions and can interfere with the origin characteristic signals, this invention uses the above method to find the optimal NaOH solution concentration that best highlights the coffee characteristic signals. This effectively suppresses the interference of these sugar components while maximizing the differences in origin characteristics, significantly improving the accuracy of traceability.

[0021] The method for uniformly mixing the coffee solution corresponding to the coffee bean sample with a NaOH solution of optimal concentration and extracting the corresponding feature signal set includes the following steps: N1: Mix the coffee solution corresponding to 1 ml of coffee bean sample with 19 ml of NaOH solution of optimal concentration to obtain the corresponding mixed solution; N2: The mixed solution was characterized using a three-electrode sensor and differential pulse voltammetry to obtain the peak potential. Peak current Peak area , peak potential Peak current Peak area The set of characteristic signals that make up the coffee bean sample.

[0022] Preferably, the peak potential can also be... Peak current Peak area The corresponding standardized feature signals are obtained by performing standardization processing on each signal. Standardized feature signals Standardized feature signals Standardize the feature signal Standardized feature signals Standardized feature signals The set of characteristic signals that make up the coffee bean sample.

[0023] The method for preparing a coffee solution corresponding to a coffee bean sample includes the following steps: after grinding the coffee bean sample, pass it through an 80-100 mesh sieve, dissolve it in distilled water at 95℃ at a solid-liquid ratio of 1:10 (g / mL), stir evenly, let it stand and filter to obtain a coffee solution.

[0024] In this scheme, a coffee solution corresponding to each coffee bean sample is first prepared, and then multiple NaOH solutions of different concentrations are prepared, with the concentration range of NaOH solutions being 0.001~1 mol / L.

[0025] For each concentration of NaOH solution, the following procedure was performed: The cyclic voltammetry curves collected after mixing with distilled water were used as control curves (i.e., the blank control system corresponding to this concentration of NaOH solution). By subtracting the control curve from the cyclic voltammetry curves collected after mixing with each coffee solution, the net response current curve for each coffee bean sample at that concentration of NaOH solution was obtained. The corresponding reduction peak current value was then acquired, and the average reduction peak current was calculated. ; By subtracting the control curve from the cyclic voltammetry curve collected after mixing with sucrose solution, the net response current curve of sucrose solution at this concentration of NaOH solution was obtained, and the corresponding reduction peak current value was acquired. ; By subtracting the control curve from the cyclic voltammetry curve obtained after mixing the glucose solution with the glucose solution, the net response current curve of the glucose solution at this concentration of NaOH solution was obtained, and the corresponding reduction peak current value was acquired. ; By subtracting the control curve from the cyclic voltammetry curve collected after mixing with fructose solution, the net response current curve of fructose solution at this concentration of NaOH solution was obtained, and the corresponding reduction peak current value was acquired. ; Finally, the characteristic value corresponding to this concentration of NaOH solution was calculated.

[0026] Taking Yunnan Baoshan coffee beans as an example, the net response current curves of the coffee solutions corresponding to Yunnan Baoshan coffee bean samples under different concentrations of NaOH solutions are shown below. Figure 2 As shown, Figure 2There are seven net response current curves, corresponding to seven NaOH solution concentrations: 0.001 mol / L, 0.005 mol / L, 0.01 mol / L, 0.05 mol / L, 0.1 mol / L, 0.5 mol / L, and 1.0 mol / L. The net response current curves for coffee, sucrose, glucose, and fructose solutions corresponding to the Yunnan Baoshan coffee bean sample in a 0.01 mol / L NaOH solution are shown below. Figure 3 As shown.

[0027] After calculating the characteristic value corresponding to each concentration of NaOH solution, the concentration with the highest characteristic value is taken as the optimal concentration of NaOH solution. Next, the coffee solution corresponding to each coffee bean sample is uniformly mixed with the optimal concentration of NaOH solution, and the corresponding feature signal set is extracted. This feature signal set and the origin label of the coffee bean sample are used as training samples to train the machine learning model, resulting in a coffee bean origin traceability model. Finally, the coffee solution corresponding to the coffee bean sample to be tested is prepared, uniformly mixed with the optimal concentration of NaOH solution, and the corresponding feature signal set is extracted. This feature signal set is input into the coffee bean origin traceability model, and the model outputs the origin of the coffee bean sample to be tested.

[0028] This invention enables rapid and accurate traceability of coffee bean origins, is simple to operate, and is inexpensive. By optimizing the concentration of the NaOH solution, it effectively suppresses the electrochemical interference of sucrose, glucose, and fructose in coffee, highlighting the characteristic signal differences of coffee from different origins and significantly improving traceability accuracy.

[0029] The machine learning model is a support vector machine. The three-electrode sensor includes a copper film working electrode, a saturated KCl calomel reference electrode, and a platinum counter electrode. The thickness of the copper film working electrode is 0.02 mm. The parameters for cyclic voltammetry are: potential range -0.9 to -0.2 V, scan rate 50 mV / s, and 2 scan cycles. The parameters for differential pulse voltammetry are: potential range 0 to -0.5 V, voltage step size 4 mV, pulse amplitude 0.05 V, pulse duration 0.05 s, sampling interval 0.0167 s, and pulse period 0.5 s.

[0030] The method employs a three-electrode electrochemical sensor combined with cyclic voltammetry and differential pulse voltammetry, which is simple to operate and fast to detect. After the coffee bean origin traceability model is trained, the traceability cycle for a single sample is ≤1 minute, which significantly shortens the detection time compared to chromatography and mass spectrometry methods.

Claims

1. A method for tracing the origin of coffee beans, characterized in that, Includes the following steps: S1: Obtain coffee bean samples from different origins, prepare m samples of coffee beans from each origin, and prepare a coffee solution corresponding to each coffee bean sample; S2: Prepare NaOH solutions of different concentrations, mix each NaOH solution with coffee solution evenly, analyze the mixture, and screen to obtain the optimal concentration of NaOH solution. S3: Mix the coffee solution corresponding to each coffee bean sample with the optimal concentration of NaOH solution evenly, and extract the corresponding feature signal set; S4: Input the feature signal set and origin label of coffee bean sample into the machine learning model for training to obtain the coffee bean origin traceability model; S5: Prepare a coffee solution corresponding to the coffee bean sample to be tested, and mix it evenly with a NaOH solution of the optimal concentration. Extract the corresponding feature signal set, input the feature signal set into the coffee bean origin traceability model, and the coffee bean origin traceability model outputs the origin of the coffee bean sample to be tested.

2. The method for tracing the origin of coffee beans according to claim 1, characterized in that, Step S2 includes the following steps: S21: Prepare NaOH solutions of different concentrations, with the concentration range of NaOH solutions being 0.001~1 mol / L; S22: After uniformly mixing NaOH solution of each concentration with coffee solution, analyze the mixture and extract the characteristic value corresponding to each NaOH solution concentration. S23: The concentration with the largest characteristic value is taken as the optimal concentration of the NaOH solution.

3. The method for tracing the origin of coffee beans according to claim 2, characterized in that, The method for uniformly mixing NaOH solution of concentration d with coffee solution in step S22 and then analyzing the mixture to extract the characteristic value corresponding to NaOH solution of concentration d, where 0.001≤d≤1, includes the following steps: M1: Mix 1 ml of distilled water with 19 ml of NaOH solution of concentration d to obtain the corresponding mixed solution. Use a three-electrode sensor to collect the cyclic voltammetry curve of the mixed solution and use the cyclic voltammetry curve as the comparison curve. M2: 1 ml of each coffee solution was uniformly mixed with 19 ml of NaOH solution of concentration d to obtain the corresponding mixed solution. Cyclic voltammetry curves of each mixed solution were collected using a three-electrode sensor. Subtract each cyclic voltammetry curve from the comparison curve to obtain the corresponding net response current curve, and obtain the restored peak current value of each net response current curve. Calculate the average value of the reduction peak current , ， Where n is the number of coffee bean producing regions. Let be the reduction peak current value corresponding to the j-th coffee bean sample from the i-th origin, where 1≤i≤n, 1≤j≤m; M3: 1 mL of 0.01 mol / L sucrose solution, glucose solution, and fructose solution were mixed with 19 mL of NaOH solution of concentration d to obtain the corresponding mixed solutions. Cyclic voltammetry curves of each mixed solution were collected using a three-electrode sensor. Subtract the control curve from each cyclic voltammetry curve to obtain the corresponding net response current curve, and obtain the reduction peak current value of each net response current curve. The reduction peak current value corresponding to the sucrose solution is denoted as... The reduction peak current value corresponding to the glucose solution is denoted as The reduction peak current value corresponding to the fructose solution is denoted as ; M4: Calculate the characteristic value corresponding to a NaOH solution with concentration d. The calculation formula is as follows: ， ， ， ， in, These are the eigenvalues.

4. The method for tracing the origin of coffee beans according to claim 1, characterized in that, The method for uniformly mixing the coffee solution corresponding to the coffee bean sample with a NaOH solution of optimal concentration and extracting the corresponding feature signal set includes the following steps: N1: Mix the coffee solution corresponding to 1 ml of coffee bean sample with 19 ml of NaOH solution of optimal concentration to obtain the corresponding mixed solution; N2: The mixed solution was characterized using a three-electrode sensor and differential pulse voltammetry to obtain the peak potential. Peak current Peak area , peak potential Peak current Peak area The set of characteristic signals that make up the coffee bean sample.

5. The method for tracing the origin of coffee beans according to claim 1, characterized in that, The machine learning model is a support vector machine.

6. The method for tracing the origin of coffee beans according to claim 1, characterized in that, The method for preparing a coffee solution corresponding to a coffee bean sample includes the following steps: after grinding the coffee bean sample, pass it through an 80-100 mesh sieve, dissolve it in distilled water at 95℃ at a solid-liquid ratio of 1:10 (g / mL), stir evenly, let it stand and filter to obtain a coffee solution.

7. A method for tracing the origin of coffee beans according to claim 3 or 4, characterized in that, The three-electrode sensor includes a copper film working electrode, a saturated KCl calomel reference electrode, and a platinum counter electrode.

8. The method for tracing the origin of coffee beans according to claim 3, characterized in that, The parameters for the cyclic voltammetry method are: potential range -0.9 to -0.2V, scan rate of 50mV / s, and number of scans of 2.

9. The method for tracing the origin of coffee beans according to claim 4, characterized in that, The parameters of the differential pulse voltammetry method are: potential range 0 to -0.5V, voltage step size 4mV, pulse amplitude 0.05V, pulse duration 0.05s, sampling interval 0.0167s, and pulse period 0.5s.

Images