Soybean raw material tofu product quality prediction method and device and electronic equipment

Abstract

The invention provides a bean curd product quality prediction method and device of soybean raw materials and electronic equipment, and relates to the technical field of artificial intelligence. The method comprises the following steps: obtaining quality index data of a soybean raw material to be detected; inputting the quality index data of the to-be-detected soybean raw material into the tofu product quality prediction model to obtain a tofu product quality prediction result of the to-be-detected soybean raw material output by the tofu product quality prediction model; wherein the tofu product quality prediction model is obtained by training based on training samples and corresponding label data, the training samples are determined based on quality index data of the soybean raw materials, and the label data are determined based on quality index data of tofu products prepared by processing the soybean raw materials. According to the method, the accuracy of bean curd product quality prediction can be improved, so that soybean raw materials can be effectively screened, and the stability of the quality of processed bean curd products is ensured.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular, to a method, device and electronic device for predicting the quality of tofu products from soybean raw materials. Background Art

[0002] Soybeans, as raw materials for tofu processing, have an important impact on the quality of tofu products. China has extremely rich soybean germplasm resources. The mixed planting, harvesting and storage modes result in unstable raw material quality among different batches, which will affect the quality of processed tofu. Therefore, how to predict the quality of tofu processed from soybean raw materials has become an urgent problem to be solved in this field.

[0003] In the prior art, an analysis method based on a linear basis is usually selected to explore the correlation between soybean raw materials and tofu quality. Due to the diversity of soybean raw material quality indicators, the impact on the quality of tofu products is very complex, and the prediction accuracy of tofu quality is relatively low, which cannot well solve the problem of soybean raw material screening. Summary of the Invention

[0004] Embodiments of the present invention provide a method, device and electronic device for predicting the quality of tofu products from soybean raw materials, which are used to improve the accuracy of predicting the quality of tofu products from soybean raw materials.

[0005] In a first aspect, embodiments of the present invention provide a method for predicting the quality of tofu products from soybean raw materials, which is characterized by including: obtaining quality index data of the to-be-tested soybean raw materials; inputting the quality index data of the to-be-tested soybean raw materials into a tofu product quality prediction model to obtain a tofu product quality prediction result of the to-be-tested soybean raw materials output by the tofu product quality prediction model; wherein, the tofu product quality prediction model is trained based on training samples and their corresponding label data, the training samples are determined based on the quality index data of soybean raw materials, and the label data are determined based on the quality index data of tofu products prepared from soybean raw materials.

[0006] In some embodiments, after obtaining the tofu product quality prediction result of the to-be-tested soybean raw materials output by the tofu product quality prediction model, it further includes: determining whether the to-be-tested soybean raw materials are suitable for tofu processing based on the tofu product quality prediction result.

[0007] As a possible implementation manner, determining whether the to-be-tested soybean raw materials are suitable for tofu processing based on the tofu product quality prediction result includes: determining an applicable score of the to-be-tested soybean raw materials based on the tofu product quality prediction result; and determining whether the to-be-tested soybean raw materials are suitable for tofu processing based on the applicable score.

[0008] In some embodiments, the quality index data of the soy raw material to be measured includes at least one of the following: soy sensory quality index data, soy physical and chemical nutritional quality index data, soy functional quality index data, soy processing quality index data, and soy safety quality index data.

[0009] In some embodiments, the quality index data of the tofu product includes at least one of the following: tofu taste quality index data, tofu texture quality index data, tofu economic benefit index data, tofu physical and chemical nutritional quality index data, and tofu processing quality index data.

[0010] As an example, the tofu product quality prediction model is a gradient boosting model.

[0011] In a second aspect, an embodiment of the present invention provides a device for predicting the quality of a tofu product from a soy raw material, including: an acquisition module for acquiring the quality index data of the soy raw material to be measured; a prediction module for inputting the quality index data of the soy raw material to be measured into a tofu product quality prediction model to obtain a prediction result of the quality of the tofu product of the soy raw material output by the tofu product quality prediction model; wherein, the tofu product quality prediction model is trained based on training samples and their corresponding label data, the training samples are determined based on the quality index data of the soy raw material, and the label data is determined based on the quality index data of the tofu product prepared from the soy raw material.

[0012] In a third aspect, an embodiment of the present invention provides an electronic device, including a processor and a memory storing a computer program, and the processor implements the method for predicting the quality of a tofu product from a soy raw material according to the first aspect when executing the program.

[0013] In a fourth aspect, an embodiment of the present invention provides a non-transitory computer-readable storage medium, on which a computer program is stored, and the computer program implements the method for predicting the quality of a tofu product from a soy raw material according to the first aspect when executed by a processor.

[0014] In a fifth aspect, an embodiment of the present invention provides a computer program product, including a computer program, and the computer program implements the method for predicting the quality of a tofu product from a soy raw material according to the first aspect when executed by a processor.

[0015] A method, apparatus, and electronic device for predicting the quality of tofu products from soybean raw materials according to an embodiment of the present invention. Obtain quality index data of the to-be-tested soybean raw materials; input the quality index data of the to-be-tested soybean raw materials into a tofu product quality prediction model to obtain a tofu product quality prediction result output by the tofu product quality prediction model. Among them, the tofu product quality prediction model is trained based on training samples and their corresponding label data. The training samples are determined based on the quality index data of soybean raw materials, and the label data is determined based on the quality index data of tofu products prepared from soybean raw materials. By learning the mapping relationship between the quality index data of soybean raw materials and the quality index data of tofu products prepared therefrom through the tofu product quality prediction model, the present invention can improve the accuracy of tofu product quality prediction, thereby effectively screening soybean raw materials and ensuring the stability of the quality of processed tofu. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] To more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0017] Figure 1 It is a schematic flowchart of a method for predicting the quality of tofu products from soybean raw materials provided by an embodiment of the present invention; Figure 2 It is a schematic structural diagram of a device for predicting the quality of tofu products from soybean raw materials provided by an embodiment of the present invention; Figure 3 It is a schematic structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] To make the objectives, technical solutions, and advantages of the present invention clearer, the following clearly and completely describes the technical solutions in the present invention with reference to the accompanying drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments in the present invention belong to the scope of protection of the present invention.

[0019] It should be noted that soybeans, as raw materials for tofu processing, have an important impact on the quality of tofu. China has extremely rich soybean germplasm resources. The mixed planting, harvesting, and storage modes result in unstable raw material quality among different batches, which will affect the quality of processed tofu. Therefore, how to predict the quality of tofu products processed from soybean raw materials has become an urgent problem in this field.

[0020] Since tofu formation is a complex result of multi-factor coupling, there is a large uncertainty and non-linearity between raw materials and products. Currently, in the research on exploring the correlation between soybean raw materials and tofu quality, analysis methods based on linear basis are usually selected, which not only make it difficult to explain the complex correlation between the two, but also the selected soybeans for special tofu are limited to the quality evaluation and comparison among limited soybean varieties in the research. Although some studies introduce mathematical and statistical methods to predict the processing suitability of soybeans, the measurement indicators and processing conditions are different, resulting in low model accuracy and small application scope, and cannot well solve the problem of screening raw materials for special tofu soybeans.

[0021] To solve the above problems, an embodiment of the present invention provides a method, device and electronic device for predicting the quality of tofu products from soybean raw materials.

[0022] Figure 1 The flowchart of a method for predicting the quality of tofu products from soybean raw materials provided by an embodiment of the present invention is shown in Figure 1 As shown, the method may include the following steps.

[0023] Step 101, obtain the quality index data of the to-be-tested soybean raw material.

[0024] The to-be-tested soybean raw material refers to the soybean raw material for which the quality of tofu products needs to be predicted, and the to-be-tested soybean raw material can be soybeans of any variety.

[0025] In some embodiments, the quality index data of the to-be-tested soybean raw material may include at least one of the following: soybean sensory quality index data, soybean physical and chemical nutritional quality index data, soybean functional quality index data, soybean processing quality index data, soybean safety quality index data. For example, the quality index data of the to-be-tested soybean raw material may include: soybean sensory quality index data, soybean physical and chemical nutritional quality index data, soybean functional quality index data, soybean processing quality index data and soybean safety quality index data.

[0026] In some embodiments, the soybean sensory quality index data may include index data such as 100-seed weight, color and luster, etc. The soybean physical and chemical nutritional quality index data may include index data such as moisture content, protein content, fat content, amino acid content, mineral element content such as calcium, iron, phosphorus, magnesium, fatty acid content, etc. The soybean functional quality index data may include index data such as isoflavone content, saponin content, etc. The soybean processing quality index data may include index data such as protein subunit composition, lipoxygenase content, water-soluble protein content, protein solubility ratio, etc. The soybean safety quality index data may include index data such as phytic acid content, etc.

[0027] As an example, the 100-grain weight of the soybean raw material to be tested can be measured based on the provisions of the national standard GB / T5519-2018 "Determination of 1,000-grain weight of cereals and legumes". The color of the soybean raw material to be tested can be the color parameter measured by an electronic eye after 30 g of soybeans are weighed after manual impurities are removed and diseased, impure and unfull grains are removed.

[0028] As an example, the moisture content of the soybean raw material to be tested can be measured based on the provisions of GB5009.3-2016 "National Food Safety Standard Determination of Water in Food"; the protein content of the soybean raw material to be tested can be measured based on the automatic Kjeldahl nitrogen analyzer method in GB5009.5-2016 "National Food Safety Standard Determination of Protein in Food", and the conversion coefficient of nitrogen to protein is 5.71; the fat content of the soybean raw material to be tested can be measured based on the Soxhlet extraction method in GB5009.6-2016 "National Food Safety Standard Determination of Fat in Food"; the amino acid content of the soybean raw material to be tested can be measured based on the provisions of GB5009.124-2016 "National Food Safety Standard Determination of Amino Acids in Food"; the mineral content of calcium, iron, phosphorus, magnesium, etc. in the soybean raw material to be tested can be measured based on the inductively coupled plasma emission spectrometry in GB 5009.268-2016 "National Food Safety Standard Determination of Multiple Elements in Food".

[0029] As an example, the fatty acid content of the soy raw material to be tested can be measured in the following way: Grind the soy raw material to be tested into powder, weigh 30 mg of the powder of the soy raw material to be tested that has passed through a 40-mesh sieve into a 5-mL centrifuge tube, and add 1 mL of n-hexane solution; vortex the centrifuge tube for 1 min and then heat and extract it in a water bath at 50 °C for 20 min, shaking it well at intervals; after it cools to room temperature, add 1 mL of 0.5 mol / L sodium methoxide solution to the centrifuge tube and vortex for 10 min to fully methylate it; let it stand until it layers, filter it through a 0.22-μm needle filter and transfer it to an injection vial for storage at 4 °C, and detect it by a gas chromatography-mass spectrometry (GC-MS) instrument. Among them, the relevant parameters of the GC-MS instrument are described as follows: Shimadzu GCMS-QP2020 NX series single quadrupole gas chromatography-mass spectrometry instrument; detector: flame ionization detector (FID); chromatographic column: Agilent J&W DB-5ms ultra-inert chromatographic column (30 m × 0.25 mm × 0.25 μm); chromatographic column temperature: 40 °C; equilibration time: 3 min; column oven temperature program: hold at 40 °C for 2 min, increase the temperature to 210 °C at a rate of 4 °C / min, and increase the temperature to 300 °C at a rate of 10 °C / min and hold for 5 min; carrier gas: helium; total helium flow rate: 50.0 mL / min; chromatographic column flow rate: 1.32 mL / min; linear velocity: 41.5 cm / s; purge flow rate: 3 mL / min; inlet temperature: 200 °C; injection volume: 1 μL; injection mode: 10:1; ion source temperature: 220 °C; interface temperature: 200 °C; acquisition mode: full scan; mass-to-charge ratio range: 50 - 400 m / z.

[0030] As an example, the method for determining the isoflavone content of the soy raw material to be tested can be as follows: Weigh 1 g of the powder of the soy raw material to be tested that has passed through a 40-mesh sieve into a 15 mL centrifuge tube, and add 10 mL of 70% ethanol solution; vortex the centrifuge tube for 1 min and then perform ultrasonic extraction at 60 °C for 40 min; after ultrasonic treatment, place the centrifuge tube in a centrifuge and centrifuge at 3000 r / min for 15 min; transfer the supernatant in the test tube to a 10 mL volumetric flask and make up the volume to 10 mL with 70% ethanol solution; pipette 2 mL of the solution, filter it through a 0.22 μm organic phase syringe filter, transfer it to an injection vial, and store it at 4 °C for testing; perform detection using a liquid chromatograph. Among them, the relevant parameters of the liquid chromatograph are described as follows: Liquid chromatograph (Agilent 1260 Infinity Ⅱ, Agilent Corporation, USA); Chromatographic column: Agilent ZORBAX SB C18 5 μm 4.6×250 mm; Mobile phase: A: 0.1% aqueous acetic acid solution, B: acetonitrile; Flow rate: 1 mL / min; Detector: ultraviolet detector, wavelength 260 nm; Chromatographic column temperature: 40 °C; Injection volume: 10 μL; Gradient elution program: Initial gradient is 90% A (B 10%); 0 - 5 minutes, 80% A (20% B); 5 - 10 minutes, 80% A (20% B); 10 - 15 minutes, 75% A (25% B); 15 - 18 minutes, 70% A (30% B); 18 - 22 minutes, 65% A (35% B); 22 - 25 minutes, 55% A (45% B); 25 - 27 minutes, 55% A (45% B); 27 - 28 minutes, 90% A (10% B); 28 - 32 minutes, 90% A (10% B).

[0031] As an example, the method for determining the saponin content of the soybean raw material to be tested can be as follows: Freeze-dry the soybean raw material grains to be tested and grind them into powder with a ball mill; accurately weigh 1 g of the powder of the soybean raw material to be tested passing through a 40-mesh sieve into a 15-mL centrifuge tube, and add 10 mL of 70% ethanol solution (containing 0.01% acetic acid); vortex the centrifuge tube for 1 min and then perform ultrasonic extraction for 10 min, and extract in the dark at room temperature for 24 hours; place the centrifuge tube in a centrifuge and centrifuge at 3000 r / min for 10 min; transfer the supernatant in the test tube to a 10-mL volumetric flask, and make up the volume to 10 mL with 70% ethanol solution (containing 0.01% acetic acid). Finally, pipette 2 mL of the solution, filter it through a 0.22-μm organic phase syringe filter, transfer it to an injection vial, and store it at 4 °C for testing; perform detection with a liquid chromatograph. Among them, the relevant parameters of the liquid chromatograph are described as follows: Liquid chromatograph (Agilent 1260 InfinityⅡ, Agilent Technologies, USA); chromatographic column: Agilent ZORBAX SB C18 5 μm 4.6×250 mm; mobile phase: A: 0.05% aqueous acetic acid solution, B: acetonitrile; flow rate: 1 mL / min; detector: ultraviolet detector, wavelength 205 nm; column temperature: 30 °C; injection volume: 20 μL; gradient elution program: 0 - 10 minutes, 90% A (10% B); 10 - 20 minutes, 30% A (70% B); 20 - 35 minutes, 60% A (40% B); 35 - 37 minutes, 60% A (40% B); 37 - 38 minutes, 90% A (10% B); 38 - 40 minutes, 90% A (10% B).

[0032] As an example, the protein subunit composition and deoxygenase content of the soybean raw material to be tested can be determined in the following way: Weigh 5 mg of the powder of the soybean raw material to be tested that has passed through a 40-mesh sieve into a 2-mL centrifuge tube, and add 0.5 ml of 1.5M Tris-HCI buffer (pH = 8.8); vortex the centrifuge tube for 10 min and then centrifuge it at 12,000 r / min for 5 min in a centrifuge; aspirate 200 μL of the supernatant and transfer it to a new centrifuge tube, and add 50 μL of 5x protein loading buffer (containing DTT); place the new centrifuge tube in a boiling water bath at 100 °C for 5 min, cool it to room temperature, and then centrifuge it at 12,000 rpm for 5 min; finally, take 2 - 5 μL of the supernatant for sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). The instrument conditions are described as follows: Use constant voltage electrophoresis at 80 V in the stacking gel stage (about 30 min), and adjust it to constant voltage electrophoresis at 110 V in the separating gel stage (about 1.5 h). After the bromophenol blue indicator band moves to the bottom of the separating gel, the power supply can be turned off; Gel staining and decolorization: Stain the intact gel after stripping with Coomassie brilliant blue staining solution for 1 h; After the staining is completed, put it into Coomassie brilliant blue decolorizing solution for decolorization, and change the decolorizing solution every 4 hours until the gel background color is transparent.

[0033] As an example, the water-soluble protein content of the soybean raw material to be tested can be determined based on the provisions of NY / T 1205 - 2006 "Determination of Water-Soluble Protein Content in Soybeans". The protein dissolution ratio of the soybean raw material to be tested can be the ratio of the mass concentration of water-soluble protein to the mass concentration of total protein.

[0034] Step 102, input the quality index data of the soybean raw material to be tested into the tofu product quality prediction model to obtain the tofu product quality prediction result of the soybean raw material to be tested output by the tofu product quality prediction model.

[0035] Among them, the tofu product quality prediction model is trained based on the training samples and their corresponding label data. The training samples are determined based on the quality index data of the soybean raw material, and the label data are determined based on the quality index data of the tofu product processed from the soybean raw material.

[0036] In some embodiments, the tofu product quality prediction model can be a neural network model, a machine learning model, etc. The applicant of the present invention tested different machine learning algorithms during the model training stage, namely Linear Regression, Lasso Regression, Ridge Regression, Decision Tree, K-Nearest Neighbors, Support Vector Machine, Random Forest, Gradient Boosting, Adaptive Boosting, Extremely Randomized Trees, and Artificial Neural Network. Finally, it was found that the Gradient Boosting model had the best performance. Preferably, the tofu product quality prediction model in the embodiments of the present invention can be constructed based on the Gradient Boosting model.

[0037] In some embodiments, the tofu product quality prediction model has learned the mapping relationship between the quality index data of the soybean raw materials and the quality index data of the tofu products prepared from the soybean raw materials. After inputting the quality index data of the to-be-tested soybean raw materials into the tofu product quality prediction model, the quality prediction result of the tofu products of the to-be-tested soybean raw materials is obtained through feature extraction and mapping. Among them, the quality prediction result of the tofu products of the to-be-tested soybean raw materials refers to the quality index prediction result of the tofu products prepared by using the to-be-tested soybean raw materials for tofu processing.

[0038] It should be noted that the quality index data of the tofu products included in the quality prediction result of the tofu products of the to-be-tested soybean raw materials can be determined based on actual needs. For example, it can include the color, flavor, soymilk yield, tofu yield, etc. of the tofu products, or it can also include the moisture content, protein content, etc. of the tofu products.

[0039] In some embodiments, the training process of the tofu product quality prediction model can include: obtaining the quality index data of different types of soybean raw materials; performing tofu processing based on each type of soybean raw material to obtain the quality index data of the tofu products prepared from each type of soybean raw material; for each type of soybean raw material, determining the quality index data of this type of soybean raw material as the training sample, and determining the quality index data of the tofu products prepared from this type of soybean raw material as the label data corresponding to this training sample; inputting the training sample into the initial tofu product quality prediction model for model training, and finally obtaining the trained tofu product quality prediction model.

[0040] It should be noted that the quality index data of different types of soybean raw materials are consistent with the index items included in the quality index data of the soybean raw material to be tested, and the measurement methods are the same as those of the corresponding indicators of the soybean raw material to be tested. The index items included in the quality prediction result of the tofu product of the soybean raw material to be tested are consistent with the tofu quality index items included in the label data, and the measurement methods are also the same.

[0041] As a possible implementation method, the quality index data of the tofu product includes at least one of the following: tofu taste quality index data, tofu texture quality index data, tofu economic benefit index data, tofu physical and chemical nutrition quality index data, and tofu processing quality index data. For example, the quality index data of the tofu product includes: tofu taste quality index data, tofu texture quality index data, tofu economic benefit index data, tofu physical and chemical nutrition quality index data, and tofu processing quality index data.

[0042] Among them, the tofu taste quality index data can include index data such as the color, flavor, and volatile substance content of the tofu product. The flavor can be the concentration of different aroma types, and the volatile substances can include hexanal, n-hexanol, trans-2-hexenal, 1-octen-3-ol, n-nonanal, etc. The tofu texture quality index data can include the hardness, viscosity, resilience, cohesiveness, elasticity, gumminess, and chewiness of the tofu product. The tofu economic benefit index data can include index data such as the soybean milk yield and the tofu yield. The tofu physical and chemical nutrition quality index data can include index data such as the moisture content and protein content of the tofu product. The tofu processing quality index data can include index data such as the water holding capacity and protein recovery rate of the tofu product.

[0043] As an example, the color of the tofu product in the tofu taste quality index data can be the color value data measured by an electronic eye device; the flavor of the tofu product can be determined by the following method: weigh 5 g of the tofu product, place it in a clean 25 mL headspace vial, tighten the headspace vial cap, and let it stand at room temperature for 2 h before testing on the machine; use an electronic nose probe to suck the air at the top of the headspace vial and analyze and determine its volatile flavor; the electronic nose contains 10 different metal oxide sensors, and each sensor can detect different aroma types.

[0044] As an example, the content of volatile substances in the data of tofu taste quality indicators can be determined in the following way: Grind the tofu evenly with a glass rod and accurately weigh 5 g of the tofu product into a headspace extraction bottle; Use a microsyringe to add 2 μL of the internal standard (2-methyl-3-heptanone, 0.1 mg / ml) solution and seal it with a headspace bottle cap; Insert the aged solid-phase microextraction syringe needle into the sealed headspace bottle and push out the SPME extraction head. The sample is extracted in the headspace at a constant temperature of 60 °C for 1 h; Finally, immediately introduce the SPME extraction head through the injection port for the next separation and identification; Before each extraction of the sample, age the SPME extraction head at 250 °C for 5 min to reduce the memory effect. Among them, the instructions for using the instrument are as follows: Shimadzu GCMS-QP2020 NX series single quadrupole gas chromatography-mass spectrometry; Flame ionization detector (FID); Agilent J&W DB-5ms ultra-inert chromatographic column (30 m × 0.25 mm × 0.25 μm); Column temperature: 40 °C; Equilibration time: 3 min; Column oven temperature program: Hold at 40 °C for 2 min, increase the temperature to 210 °C at a rate of 4 °C / min, and increase the temperature to 300 °C at a rate of 10 °C / min and hold for 5 min; Carrier gas: Helium; Total helium flow rate: 50.0 mL / min; Column flow rate: 1.32 mL / min; Linear velocity: 41.5 cm / s; Purge flow rate: 3 mL / min; Injection port temperature: 200 °C; Injection mode: Splitless; Ion source temperature: 220 °C; Interface temperature: 200 °C; Acquisition method: Full scan; Mass-to-charge ratio range: 50 - 400 m / z.

[0045] As an example, the hardness, adhesiveness, resilience, cohesiveness, elasticity, gumminess, chewiness, etc. of the tofu product in the data of tofu texture quality indicators can be measured by a texture analyzer. For example, equilibrate the tofu product to room temperature. In the middle area of the tofu product, cut the tofu product into small cubes of 2×2×2 cm 3 and use the 36R cylindrical probe equipped with the texture analyzer to measure the texture characteristics of the tofu. The texture analyzer parameters are set as follows: The compression ratio of compressing the tofu twice is 40%, the pressure is 5 g, the time interval is 5 s, the pre-test speed is 1 mm / s, the in-test speed is 1 mm / s, and the post-test speed is 1 mm / s.

[0046] As an example, the soymilk yield in the data of tofu economic benefit indicators can be the ratio of the mass of the raw soymilk passed through a 120-mesh sieve to the mass of the soybean raw materials corresponding to this raw soymilk. The tofu yield can be the ratio of the mass of the fresh tofu product cooled to room temperature to the mass of the corresponding soybean raw materials.

[0047] As an example, the moisture content in the physicochemical and nutritional quality index data of tofu can be determined based on GB 5009.3-2016 "National Food Safety Standard - Determination of Moisture in Foods". The protein content can be determined based on GB 5009.5-2016 "National Food Safety Standard - Determination of Protein in Foods", and the conversion coefficient of nitrogen to protein is 6.25.

[0048] As an example, the water holding capacity of tofu in the tofu processing quality index data can be determined as follows: Weigh 20 g of tofu product (denoted as W0) into a 50 mL centrifuge tube filled with sufficient absorbent cotton; Place the centrifuge tube in a centrifuge and centrifuge at 5000 r / min for 20 min; Dry the surface moisture of the tofu product with filter paper and weigh it (denoted as W1); The ratio of W1 to W0 is determined as the water holding capacity of tofu. The protein recovery rate can be calculated by the following formula (1).

[0049] (1); Where, is the protein recovery rate; is the tofu yield; is the protein content of the tofu product; is the soybean protein content.

[0050] As an example, the tofu product quality prediction model is a Gradient Boosting model. During the training process, the batch gradient descent method is adopted, and when updating each parameter, all the samples used are used to update the gradient.

[0051] Since the purpose of predicting the quality of tofu products from soybean raw materials is to screen soybean raw materials to determine which soybean raw materials are suitable for tofu processing, therefore, after obtaining the tofu product quality prediction results of the to-be-tested soybean raw materials output by the tofu product quality prediction model, it also includes: Based on the tofu product quality prediction results, determining whether the to-be-tested soybean raw materials are suitable for tofu processing.

[0052] As a possible implementation method, the implementation process of determining whether the to-be-tested soybean raw materials are suitable for tofu processing based on the tofu product quality prediction results can include: Presetting the threshold range for each index, comparing the tofu product quality prediction results with the preset threshold range for each index. If the proportion of the number of indexes within the threshold range is greater than the index number threshold, it is determined that the to-be-tested soybean raw materials are suitable for tofu processing.

[0053] As another possible implementation manner, the implementation process of determining whether the to-be-tested soybean raw material is suitable for tofu processing based on the tofu product quality prediction result may include: determining the suitability score of the to-be-tested soybean raw material based on the tofu product quality prediction result; and determining whether the to-be-tested soybean raw material is suitable for tofu processing based on the suitability score. Among them, determining the suitability score of the to-be-tested soybean raw material based on the tofu product quality prediction result may include: performing weighted summation on the predicted values of each index in the tofu product quality prediction result, and determining the weighted summation result as the suitability score of the to-be-tested soybean raw material. If the suitability score of the to-be-tested soybean raw material is greater than or equal to the score threshold, it is determined that the to-be-tested soybean raw material is suitable for tofu processing.

[0054] According to the tofu product quality prediction method for soybean raw materials of the embodiments of the present invention, quality index data of the to-be-tested soybean raw material is obtained; the quality index data of the to-be-tested soybean raw material is input into the tofu product quality prediction model to obtain the tofu product quality prediction result of the to-be-tested soybean raw material output by the tofu product quality prediction model; among them, the tofu product quality prediction model is trained based on training samples and their corresponding label data, the training samples are determined based on the quality index data of the soybean raw material, and the label data is determined based on the quality index data of the tofu product prepared from the soybean raw material. By learning the mapping relationship between the quality index data of the soybean raw material and the quality index data of the tofu product prepared therefrom through the tofu product quality prediction model, the present invention can improve the accuracy of tofu product quality prediction, thereby effectively screening soybean raw materials and ensuring the stability of the quality of the processed tofu.

[0055] To implement the above embodiments, the present invention also provides a device for predicting the quality of tofu products of soybean raw materials.

[0056] Figure 2 It is a schematic structural diagram of the device for predicting the quality of tofu products of soybean raw materials provided by the embodiments of the present invention. As Figure 2 shown, the device may include an acquisition module 210 and a prediction module 220. The acquisition module 210 is configured to acquire the quality index data of the to-be-tested soybean raw material; the prediction module 220 is configured to input the quality index data of the to-be-tested soybean raw material into the tofu product quality prediction model to obtain the tofu product quality prediction result of the to-be-tested soybean raw material output by the tofu product quality prediction model; among them, the tofu product quality prediction model is trained based on training samples and their corresponding label data, the training samples are determined based on the quality index data of the soybean raw material, and the label data is determined based on the quality index data of the tofu product prepared from the soybean raw material.

[0057] In some embodiments, the device may further include a determination module 230. The determination module 230 is configured to determine whether the to-be-tested soybean raw material is suitable for tofu processing based on the tofu product quality prediction result after obtaining the tofu product quality prediction result of the to-be-tested soybean raw material output by the tofu product quality prediction model.

[0058] As a possible implementation manner, the determination module 230 is specifically configured to: determine an applicable score of the to-be-tested soybean raw material based on the tofu product quality prediction result; and determine whether the to-be-tested soybean raw material is suitable for tofu processing based on the applicable score.

[0059] In some embodiments, the quality index data of the to-be-tested soybean raw material includes at least one of the following: soybean sensory quality index data, soybean physical and chemical nutritional quality index data, soybean functional quality index data, soybean processing quality index data, and soybean safety quality index data.

[0060] In some embodiments, the quality index data of the tofu product includes at least one of the following: tofu taste quality index data, tofu texture quality index data, tofu economic benefit index data, tofu physical and chemical nutritional quality index data, and tofu processing quality index data.

[0061] Preferably, the tofu product quality prediction model is a gradient boosting model.

[0062] According to the tofu product quality prediction device for soybean raw materials of the embodiments of the present invention, by obtaining the quality index data of the to-be-tested soybean raw material; inputting the quality index data of the to-be-tested soybean raw material into the tofu product quality prediction model to obtain the tofu product quality prediction result of the to-be-tested soybean raw material output by the tofu product quality prediction model; wherein, the tofu product quality prediction model is trained based on training samples and their corresponding label data, the training samples are determined based on the quality index data of the soybean raw material, and the label data are determined based on the quality index data of the tofu product prepared from the soybean raw material. The present invention can improve the accuracy of tofu product quality prediction by learning the mapping relationship between the quality index data of the soybean raw material and the quality index data of the tofu product prepared therefrom, thereby effectively screening soybean raw materials and ensuring the stability of the quality of the processed tofu.

[0063] It should be noted that the above explanations in the embodiments of the tofu product quality prediction method for soybean raw materials are equally applicable to the tofu product quality prediction device for soybean raw materials, and will not be elaborated here.

[0064] Figure 3 An example of the physical structure diagram of an electronic device is shown in Figure 3As shown, the electronic device may include: a processor 310, a communication interface 320, a memory 330, and a communication bus 340. Among them, the processor 310, the communication interface 320, and the memory 330 complete communication with each other through the communication bus 340. The processor 310 can call the computer program in the memory 330 to implement the steps of the method for predicting the quality of tofu products from soybean raw materials provided in the above embodiments.

[0065] For example, the method includes: obtaining quality index data of the to-be-tested soybean raw material; inputting the quality index data of the to-be-tested soybean raw material into a tofu product quality prediction model to obtain a tofu product quality prediction result of the to-be-tested soybean raw material output by the tofu product quality prediction model; wherein, the tofu product quality prediction model is trained based on training samples and their corresponding label data, the training samples are determined based on the quality index data of soybean raw materials, and the label data is determined based on the quality index data of tofu products processed from soybean raw materials.

[0066] In addition, when the logical instructions in the above-mentioned memory 330 are implemented in the form of software functional units and sold or used as an independent product, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0067] On the other hand, the embodiments of the present invention also provide a computer program product. The computer program product includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the steps of the method for predicting the quality of tofu products from soybean raw materials provided in the above embodiments.

[0068] On the other hand, the embodiments of the present invention also provide a non-transitory computer-readable storage medium. The non-transitory computer-readable storage medium stores a computer program. The computer program is used to cause a processor to execute the steps of the method for predicting the quality of tofu products from soybean raw materials provided in the above embodiments.

[0069] The non-transitory computer-readable storage medium may be any available medium or data storage device accessible by a computer, including but not limited to magnetic memories (such as floppy disks, hard disks, magnetic tapes, magneto-optical disks (MO), etc.), optical memories (such as CDs, DVDs, BDs, HVDs, etc.), and semiconductor memories (such as ROMs, EPROMs, EEPROMs, non-volatile memories (NAND FLASH), solid-state drives (SSD)), etc.

[0070] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. A person of ordinary skill in the art can understand and implement it without creative effort.

[0071] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disks, optical disks, etc., including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0072] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting the quality of tofu products made from soybean raw materials, characterized in that: include: Obtaining quality index data of soybean raw materials to be tested; Inputting the quality index data of the soybean raw material to be tested into a tofu product quality prediction model to obtain a tofu product quality prediction result of the soybean raw material to be tested output by the tofu product quality prediction model; Among them, the tofu product quality prediction model is trained based on training samples and their corresponding label data, the training samples are determined based on the quality index data of soybean raw materials, and the label data are determined based on the quality index data of tofu products processed from soybean raw materials.

2. The method according to claim 1, characterized in that After obtaining the tofu product quality prediction result of the tofu product raw material to be tested output by the tofu product quality prediction model, the method further includes: Based on the tofu product quality prediction result, it is determined whether the soybean raw material to be tested is suitable for tofu processing.

3. The method according to claim 2, characterized in that The step of determining whether the soybean raw material to be tested is suitable for tofu processing based on the tofu product quality prediction result comprises: Determining the applicable score of the soybean raw material to be tested based on the tofu product quality prediction result; Based on the applicability score, it is determined whether the soybean raw material to be tested is suitable for tofu processing.

4. The method according to claim 1, characterized in that The quality index data of the soybean raw material to be tested includes at least one of the following: soybean sensory quality index data, soybean physical and chemical nutrition quality index data, soybean functional quality index data, soybean processing quality index data, and soybean safety quality index data.

5. The method according to claim 1, characterized in that The quality index data of the tofu product includes at least one of the following: tofu taste quality index data, tofu texture quality index data, tofu economic benefit index data, tofu physical and chemical nutrition quality index data, and tofu processing quality index data.

6. The method according to any one of claims 1 to 5, characterized in that The tofu product quality prediction model is a gradient boosting model.

7. A tofu product quality prediction device for soybean raw materials, characterized in that: include: An acquisition module is used to obtain quality index data of the soybean raw material to be tested; A prediction module, used for inputting the quality index data of the soybean raw material to be tested into a tofu product quality prediction model, and obtaining a tofu product quality prediction result of the soybean raw material to be tested output by the tofu product quality prediction model; Among them, the tofu product quality prediction model is trained based on training samples and their corresponding label data, the training samples are determined based on the quality index data of soybean raw materials, and the label data are determined based on the quality index data of tofu products processed from soybean raw materials.

8. An electronic device comprising a processor and a memory storing a computer program, characterized in that: When the processor executes the computer program, the method for predicting the quality of a tofu product using a soybean raw material as claimed in any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for predicting the quality of a tofu product of a soybean raw material according to any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for predicting the quality of a tofu product of a soybean raw material according to any one of claims 1 to 6 is implemented.

Images