Method, equipment, medium and product for nondestructive testing of quality of wolfberry puree
Through low-field nuclear magnetic resonance technology and multivariate linear regression model, the problem of non-destructive testing of wolfberry puree quality was solved, and rapid and accurate multi-index simultaneous analysis was achieved, thereby improving detection efficiency and accuracy.
Patent Information
- Application Number
- CN202510922959.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-04
- Publication Date
- 2025-09-12
AI Technical Summary
Existing technologies make it difficult to quickly and non-destructively detect the quality of wolfberry puree, especially indicators such as moisture distribution, sugar content and soluble solids. In addition, there is a problem in the market where counterfeit 100% fresh fruit puree is difficult to distinguish.
A low-field nuclear magnetic resonance analyzer was used to collect the transverse relaxation time of wolfberry puree using CPMG pulse sequence. The characteristic peak area ratio was fitted by inverse Laplace transform, combined with weighted geometric mean and multivariate linear regression model, to achieve non-destructive testing of wolfberry puree.
It has achieved rapid and non-destructive testing of the quality of wolfberry puree, and can simultaneously analyze the total sugar content, total acid content and total moisture content online. The testing time is shortened from 2 hours of traditional methods to 5 minutes, and the accuracy is within 0.5%.
Smart Images
Figure CN120629241A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of food testing, and in particular to a method, equipment, medium and product for non-destructive testing of wolfberry puree quality. Background Art
[0002] As a fruit and vegetable product with high nutritional value, the quality of wolfberry puree (such as water distribution, sugar content, soluble solids, etc.) directly affects the product stability and taste.
[0003] There are products on the market that are made into wolfberry pulp by rehydrating dried fruits, pretending to be 100% fresh fruit pulp. Some products have a water dilution rate of up to 100%, but there is no significant difference in appearance. It is difficult for food supervision departments and ordinary consumers to distinguish, and there is a lack of means to quickly judge.
[0004] Traditional detection methods (such as drying method, high performance liquid chromatography, etc.) have problems such as being time-consuming, damaging samples, and being unable to monitor in real time.
[0005] Existing low-field nuclear magnetic resonance technology is mostly used for meat or dairy product analysis. There is no research on the correlation between the moisture status and quality of wolfberry puree system and a standardized testing process. Summary of the Invention
[0006] The purpose of this application is to provide a method, equipment, medium and product for non-destructive testing of wolfberry puree quality, so as to improve the detection efficiency and realize non-destructive testing.
[0007] To achieve the above objectives, this application provides the following solutions:
[0008] In a first aspect, the present application provides a method for non-destructive detection of wolfberry puree quality, comprising:
[0009] The transverse relaxation time of each component of the wolfberry puree to be tested was collected using a low-field nuclear magnetic resonance analyzer and a CPMG pulse sequence.
[0010] Using an inverse Laplace transform, the transverse relaxation time is fitted to obtain a first characteristic peak, a second characteristic peak, and a third characteristic peak, and the first peak area ratio, the second peak area ratio, and the third peak area ratio are determined; the relaxation time of the first characteristic peak is 1 ms-10 ms; the relaxation time of the second characteristic peak is 10 ms-100 ms; and the relaxation time of the third characteristic peak is 100 ms-1000 ms;
[0011] When the transverse relaxation time is less than the set relaxation time and the main peak signal ratio is greater than the set ratio, the wolfberry puree to be tested is 100% pure wolfberry puree; the main peak signal ratio is the maximum value among the first peak area ratio, the second peak area ratio and the third peak area ratio;
[0012] Performing a weighted geometric mean on the transverse relaxation time of each component of the wolfberry puree to be tested to obtain a weighted geometric mean relaxation time;
[0013] According to the weighted geometric mean relaxation time, the first peak area ratio, the second peak area ratio and the third peak area ratio, the quality prediction model is used to determine the total sugar content, total acid content and total moisture content of the wolfberry pulp to be tested; the quality prediction model includes a total sugar content prediction model, a total acid content prediction model and a total moisture content prediction model; the total sugar content prediction model is a multiple linear regression equation about the first peak area ratio and the weighted geometric mean relaxation time; the total acid content prediction model is a multiple linear regression equation about the second peak area ratio and the weighted geometric mean relaxation time; the total moisture content prediction model is a multiple linear regression equation about the third peak area ratio and the weighted geometric mean relaxation time.
[0014] Optionally, the set relaxation time is 330 ms; and the set ratio is 99%.
[0015] Optionally, a weighted geometric mean is performed based on the transverse relaxation time of each component of the wolfberry puree to be tested to obtain the weighted geometric mean relaxation time, specifically comprising:
[0016] Using the formula Determine the weighted geometric mean relaxation time; where T 2,W is the weighted geometric mean relaxation time; W i is the normalized weight corresponding to the i-th component; T 2，i is the transverse relaxation time of the i-th component; n is the number of components in the wolfberry puree to be tested.
[0017] Optionally, the quality prediction model is:
[0018] Total sugar content = k1*A1+k2*T 2,w +C;
[0019] Total acid content = k1*A2+k2*T 2,w +C;
[0020] Total moisture content = k1*A3+k2*T 2,w +C;
[0021] Wherein, A1 is the first peak area ratio; A2 is the second peak area ratio; A3 is the third peak area ratio; T 2,w is the weighted geometric mean relaxation time; k1, k2, and C are the coefficients of the quality prediction model, and k1, k2, and C are obtained by optimizing the quality prediction model through the partial least squares method.
[0022] Optionally, the process of constructing the quality prediction model specifically includes:
[0023] Establishing a training data set and an initial quality prediction model; the training data set includes the transverse relaxation time, the first peak area ratio, the second peak area ratio, the third peak area ratio, and the corresponding total sugar content, total acid content, and total moisture content of multiple groups of 100% pure wolfberry purees from different origins;
[0024] Preprocessing the training data set to obtain a preprocessed training data set;
[0025] Using a random forest algorithm or an XGBoost algorithm to filter the preprocessed training data set to obtain a filtered training data set;
[0026] Based on the screened training data set, the initial quality prediction model was optimized using partial least squares method and 10-fold cross validation to make the correlation coefficient R 2 >0.9, the optimal coefficient is obtained;
[0027] The quality prediction model is established based on the optimal coefficient.
[0028] Optionally, preprocessing the training data set to obtain a preprocessed training data set specifically includes:
[0029] performing standardization processing on the training data set to obtain a standardized training data set;
[0030] Performing outlier detection on the standardized training data set using a median absolute deviation method or a box plot method, replacing outliers using a median or interpolation method to obtain an outlier-processed training data set;
[0031] The training data set after outlier processing is processed using multicollinearity check to obtain a preprocessed training data set.
[0032] Optionally, based on the screened training data set, the initial quality prediction model is optimized using partial least squares method and 10-fold cross validation, and then further comprising:
[0033] Determine the correlation coefficient R 2 Is it greater than 0.9?
[0034] If so, the optimal coefficient is obtained;
[0035] If not, the initial quality prediction model is optimized using kernel partial least squares, or the training data set is enhanced using random forest algorithm, gradient boosting decision tree or data enhancement algorithm, and then the initial quality prediction model is optimized using partial least squares and 10-fold cross validation until the correlation coefficient R 2 >0.9.
[0036] In a second aspect, the present application provides a computer device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement any of the above-described methods for non-destructive detection of wolfberry puree quality.
[0037] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-described methods for non-destructive detection of wolfberry puree quality.
[0038] In a fourth aspect, the present application provides a computer program product, comprising a computer program, which, when executed by a processor, implements any of the above-mentioned methods for non-destructive detection of wolfberry puree quality.
[0039] According to the specific embodiments provided in this application, this application has the following technical effects:
[0040] The present application provides a method, equipment, medium and product for non-destructive testing of wolfberry pulp quality, which utilizes a low-field nuclear magnetic resonance analyzer and a CPMG pulse sequence to collect the transverse relaxation time of each component of the wolfberry pulp to be tested; utilizes inverse Laplace transform to fit the transverse relaxation time to obtain the first characteristic peak, the second characteristic peak and the third characteristic peak, and determines the first peak area ratio, the second peak area ratio and the third peak area ratio; when the transverse relaxation time is less than the set relaxation time and the main peak signal ratio is greater than the set ratio, the wolfberry pulp to be tested is 100% pure wolfberry pulp; according to the transverse relaxation time of each component of the wolfberry pulp to be tested, a weighted geometric average is performed to obtain the weighted geometric mean relaxation time; based on According to the weighted geometric mean relaxation time, the first peak area ratio, the second peak area ratio and the third peak area ratio, the quality prediction model is used to determine the total sugar content, total acid content and total moisture content of the wolfberry puree to be tested; the quality prediction model includes a total sugar content prediction model, a total acid content prediction model and a total moisture content prediction model; the total sugar content prediction model is a multiple linear regression equation about the first peak area ratio and the weighted geometric mean relaxation time; the total acid content prediction model is a multiple linear regression equation about the second peak area ratio and the weighted geometric mean relaxation time; the total moisture content prediction model is a multiple linear regression equation about the third peak area ratio and the weighted geometric mean relaxation time. The nondestructive detection method for wolfberry puree quality of the present application can realize rapid, nondestructive and online detection of wolfberry puree quality, and realize multi-index synchronous analysis through the correlation model between LF-NMR relaxation characteristics and quality parameters. BRIEF DESCRIPTION OF THE DRAWINGS
[0041] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0042] Figure 1 A schematic diagram of a process for non-destructive testing of wolfberry puree quality provided in one embodiment of the present application;
[0043] Figure 2 This is the interface diagram of the relaxation time distribution curve of typical wolfberry puree T2;
[0044] Figure 3 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION
[0045] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.
[0046] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.
[0047] In an exemplary embodiment, Figure 1 As shown, a non-destructive detection method for wolfberry puree quality is provided, comprising the following steps:
[0048] S1: Using a low-field nuclear magnetic resonance analyzer and a CPMG pulse sequence, the transverse relaxation time of each component of the wolfberry puree to be tested is collected.
[0049] In this embodiment, wolfberry pulp is placed in a nuclear magnetic resonance glass tube with a diameter of 10mm-40mm, shaken evenly, and placed in a low-field nuclear magnetic resonance analyzer after ultrasonic defoaming. A low-field nuclear magnetic resonance analyzer with a magnetic field strength of 0.3T-0.5T is used, and the parameters are set: resonance frequency: 13MHz-21MHz; waiting time (TW): 2000ms-5000ms; echo time (TE): 0.2ms; number of echoes (NECH): 8000-10000. The CPMG (Carr-Purcell-Meiboom-Gill) pulse sequence is used to collect the transverse relaxation time (T2) signal in ms. The corresponding relationship between relaxation time and product composition is established through this parameter, that is, the initial quality prediction model. The CPMG sequence parameters are set to TE=0.1ms-0.5ms, NECH≥5000.
[0050] S2: Use the inverse Laplace transform to fit the transverse relaxation time to obtain the first characteristic peak, the second characteristic peak and the third characteristic peak, and determine the area ratio of the first peak, the area ratio of the second peak and the area ratio of the third peak; the relaxation time of the first characteristic peak is 1ms-10ms; the relaxation time of the second characteristic peak is 10ms-100ms; the relaxation time of the third characteristic peak is 100ms-1000ms.
[0051] In this embodiment, the T2 decay curve is fitted by inverse Laplace transform to obtain three characteristic peaks:
[0052] f(t)=D -1 {F(s)}.
[0053] Among them, F(s) is the frequency domain function, f(t) is the time domain function, and D -1 represents the inverse Laplace transform operation. F(s) = a / (s+a), f(t) = e^(-at), where a is a constant.
[0054] Based on this, the first characteristic peak, the second characteristic peak and the third characteristic peak are obtained.
[0055] T 21 (1ms-10ms): Bound water, corresponding to the first characteristic peak (characterizing total sugar).
[0056] T 22 (10ms-100ms): Water that is not easy to flow, corresponding to the second characteristic peak (characterizing total acid).
[0057] T 23 (100ms-1000ms): free water, corresponding to the third characteristic peak (characterizing the total water content).
[0058] Calculate the peak area ratio (the first peak area ratio A1, the second peak area ratio A2, the third peak area ratio A3) and the weighted geometric mean relaxation time (T 2，w ).
[0059] T2 (transverse relaxation time) describes the decay rate of the transverse magnetization intensity of a spin system and is widely used in Nuclear Magnetic Resonance (NMR) and Magnetic Resonance Imaging (MRI). In a multi-component system (such as biological tissues and porous materials), different components may have different T2 values. The weight W i Usually represents the normalized weight corresponding to each component.
[0060] S3: When the transverse relaxation time is less than the set relaxation time and the main peak signal ratio is greater than the set ratio, the wolfberry puree to be tested is 100% pure wolfberry puree; the main peak signal ratio is the maximum value among the first peak area ratio, the second peak area ratio and the third peak area ratio.
[0061] The set relaxation time is 330 ms; the set ratio is 99%.
[0062] S4: performing a weighted geometric mean according to the transverse relaxation time of each component of the wolfberry puree to be tested to obtain a weighted geometric mean relaxation time.
[0063] As an optional implementation, S4 specifically includes:
[0064] Using the formula Determine the weighted geometric mean relaxation time; where T2,W is the weighted geometric mean relaxation time; W i is the normalized weight corresponding to the i-th component; T 2，i is the transverse relaxation time of the i-th component; n is the number of components in the wolfberry puree to be tested.
[0065] If there are n T2 relaxation times T 2,1 , T 2,2 ,…,T 2,n , each corresponding normalized weight is Wi (i = 1, 2 ... n), satisfying w1 + w2 + ... + w n =1.
[0066] S5: According to the weighted geometric mean relaxation time, the first peak area ratio, the second peak area ratio and the third peak area ratio, the quality prediction model is used to determine the total sugar content, total acid content and total moisture content of the wolfberry pulp to be tested; the quality prediction model includes a total sugar content prediction model, a total acid content prediction model and a total moisture content prediction model; the total sugar content prediction model is a multiple linear regression equation about the first peak area ratio and the weighted geometric mean relaxation time; the total acid content prediction model is a multiple linear regression equation about the second peak area ratio and the weighted geometric mean relaxation time; the total moisture content prediction model is a multiple linear regression equation about the third peak area ratio and the weighted geometric mean relaxation time.
[0067] As an optional implementation, the quality prediction model is:
[0068] Total sugar content = k1*A1+k2*T 2,w +C.
[0069] Total acid content = k1*A2+k2*T 2,w +C.
[0070] Total moisture content = k1*A3+k2*T 2,w +C.
[0071] Wherein, A1 is the first peak area ratio; A2 is the second peak area ratio; A3 is the third peak area ratio; T 2,w is the weighted geometric mean relaxation time; k1, k2, and C are the coefficients of the quality prediction model, and k1, k2, and C are obtained by optimizing the quality prediction model through the partial least squares method.
[0072] As an optional implementation manner, the process of constructing the quality prediction model specifically includes:
[0073] (1) Establishing a training data set and an initial quality prediction model; the training data set includes the transverse relaxation time, first peak area ratio, second peak area ratio, third peak area ratio and the corresponding total sugar content, total acid content and total moisture content of multiple groups of 100% pure wolfberry purees from different origins.
[0074] Establish a multiple linear regression equation (initial quality prediction model) between T2 parameters and (total sugar, total acid, total water content), for example:
[0075] Total sugar content = k1*A1+k2*T 2,w +C.
[0076] Total acid content = k1*A2+k2*T 2,w +C.
[0077] Total moisture content = k1*A3+k2*T 2,w +C.
[0078] In this step, k1, k2, and C are unknowns.
[0079] The initial quality prediction model (total sugar content prediction model, total acid content prediction model, total moisture content prediction model) was optimized by partial least squares (PLS), and the correlation coefficient R 2 >0.90.
[0080] (2) Preprocessing the training data set to obtain a preprocessed training data set.
[0081] As an optional implementation manner, preprocessing the training data set to obtain a preprocessed training data set specifically includes:
[0082] The training data set is standardized to obtain a standardized training data set.
[0083] In this example, the independent variables X (i.e., transverse relaxation time, first peak area ratio, second peak area ratio, and third peak area ratio) were standardized (mean = 0, variance = 1) to prevent dimensional differences from affecting the weights. The dependent variables Y (i.e., total sugar content, total acid content, and total moisture content) were centered (if this is a regression task).
[0084] The outlier detection is performed on the standardized training data set using the median absolute deviation method or the box plot method, and the outliers are replaced using the median or interpolation method to obtain the training data set after outlier processing.
[0085] In this embodiment, MAD (Median Absolute Deviation) or box plot is used to detect outliers, and outliers are replaced by medians or interpolation methods.
[0086] The training data set after outlier processing is processed using multicollinearity check to obtain a preprocessed training data set.
[0087] In this embodiment, the variance inflation factor (VIF) between independent variables is calculated. If VIF>10, highly collinear independent variables are eliminated.
[0088] (3) Using a random forest algorithm or an XGBoost algorithm to filter the preprocessed training data set to obtain a filtered training data set.
[0089] In this example, the importance of independent variables was calculated using the random forest algorithm or the XGBoost algorithm, and the top-K important variables were retained (K was determined by cross-validation). Sparse PLS modeling was implemented using the sparsepls package (R) or custom Python code. The regularization parameter λ was optimized by grid search.
[0090] (4) Based on the screened training data set, the initial quality prediction model was optimized using partial least squares method and 10-fold cross validation to make the correlation coefficient R 2 >0.9, and the optimal coefficients, namely k1, k2, and C, are obtained.
[0091] After obtaining the filtered training data set, the partial least squares method is used for optimization, and then the parameters are tuned. The key parameters include the number of dependent variables (LatentVariables, LV) and the regularization parameter. The R 2 The maximum LV, the L1 regularization parameter to control sparsity, and 10-fold cross validation are used to define the search space.
[0092] As an optional implementation manner, based on the screened training data set, the initial quality prediction model is optimized using partial least squares method and 10-fold cross validation, and then further comprising:
[0093] Determine the correlation coefficient R 2 Is it greater than 0.9?
[0094] If so, the optimal coefficient is obtained.
[0095] If not, the initial quality prediction model is optimized using kernel partial least squares, or the training data set is enhanced using random forest algorithm, gradient boosting decision tree or data enhancement algorithm, and then the initial quality prediction model is optimized using partial least squares and 10-fold cross validation until the correlation coefficient R 2 >0.9.
[0096] In this embodiment, if R 2 If the standard is not met, the following methods are adopted:
[0097] 1. Kernel PLS (non-linear data).
[0098] 2. Ensemble learning:
[0099] Use random forest or GBDT (Gradient Boosting Decision Tree) to generate new features and then input them into PLS.
[0100] 3. Data augmentation:
[0101] Increase the sample size through SMOTE (Synthetic Minority Over-sampling Technique, a technology used to deal with class imbalance problems in classification tasks) or Generative Adversarial Networks (GAN).
[0102] (5) Based on the optimal coefficient, the quality prediction model is established.
[0103] During the above training process, 30 groups of 100% wolfberry puree samples from different origins were selected and subjected to simultaneous LF-NMR testing and national standard physical and chemical testing.
[0104] The model was built using 70% of the samples as a training set and 30% as a validation set. The results show:
[0105] The total prediction error is ≤0.5%, and the detection time of total sugar, total acid and total water content is shortened from 2 hours of traditional methods to 5 minutes.
[0106] This application establishes for the first time a quantitative relationship model (quality prediction model) between the LF-NMR relaxation characteristics of wolfberry puree and multiple quality indicators.
[0107] In this embodiment, the typical relaxation time distribution curve (T2 spectrum) of wolfberry puree T2 is as follows: Figure 2 As shown, it can be seen that T2 is less than 330ms and the main peak signal accounts for more than 99%.
[0108] The relaxation time of wolfberry puree after adding water in proportion is shown in Table 1. 1:1 means that 1 part of puree is added to 1 part of water, 2:1 means that 2 parts of puree are added to 1 part of water, and 1:0 means that the puree is not added with water. The total amount of the test sample remains unchanged. It can be seen from Table 1 that the relaxation time becomes longer after adding water.
[0109] Table 1 Corresponding results of relaxation time after adding water in proportion
[0110]
[0111] This application proposes to quickly determine whether the original pulp is 100% pure wolfberry pulp by using the transverse relaxation time (T2) and the main peak signal ratio (the ratio of the maximum peak area to the total area among A1, A2, and A3). When T2 is <330ms and the main peak signal ratio is >99%, it can be proved to be pure wolfberry pulp. On the contrary, if the T2 time is too long, it may be diluted with water. A low main peak signal ratio indicates the presence of insoluble precipitates, which may be dried fruit rehydrated pulp.
[0112] The nondestructive detection method for wolfberry puree quality of the present application can realize rapid, nondestructive and online detection of wolfberry puree quality, and realize multi-index synchronous analysis through the correlation model of LF-NMR relaxation characteristics and quality parameters.
[0113] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the above-mentioned method for non-destructive detection of wolfberry puree quality is implemented.
[0114] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, which, when executed by a processor, implements the above-mentioned method for non-destructive detection of wolfberry puree quality.
[0115] In an exemplary embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the above-mentioned method for non-destructive detection of wolfberry puree quality.
[0116] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 3As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O) and a communication interface. The processor, the memory and the input / output interface are connected via a system bus, and the communication interface is connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a method for non-destructive detection of wolfberry pulp quality is implemented.
[0117] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.
[0118] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.
[0119] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, database or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).
[0120] The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, distributed databases based on blockchains. The processors involved in the various embodiments provided herein may include, but are not limited to, general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic units, data processing logic units based on quantum computing, and the like.
[0121] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0122] This document uses specific examples to illustrate the principles and implementation methods of this application. The description of the above examples is only intended to help understand the method and core concept of this application. At the same time, for those skilled in the art, based on the concept of this application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.
Claims
1. A non-destructive testing method for wolfberry puree quality, characterized in that: include: The transverse relaxation time of each component of the wolfberry puree to be tested was collected using a low-field nuclear magnetic resonance analyzer and a CPMG pulse sequence. Using an inverse Laplace transform, the transverse relaxation time is fitted to obtain a first characteristic peak, a second characteristic peak, and a third characteristic peak, and the first peak area ratio, the second peak area ratio, and the third peak area ratio are determined; the relaxation time of the first characteristic peak is 1 ms-10 ms; the relaxation time of the second characteristic peak is 10 ms-100 ms; and the relaxation time of the third characteristic peak is 100 ms-1000 ms; When the transverse relaxation time is less than the set relaxation time and the main peak signal ratio is greater than the set ratio, the wolfberry puree to be tested is 100% pure wolfberry puree; the main peak signal ratio is the maximum value among the first peak area ratio, the second peak area ratio and the third peak area ratio; Performing a weighted geometric mean on the transverse relaxation time of each component of the wolfberry puree to be tested to obtain a weighted geometric mean relaxation time; According to the weighted geometric mean relaxation time, the first peak area ratio, the second peak area ratio and the third peak area ratio, the quality prediction model is used to determine the total sugar content, total acid content and total moisture content of the wolfberry pulp to be tested; the quality prediction model includes a total sugar content prediction model, a total acid content prediction model and a total moisture content prediction model; the total sugar content prediction model is a multiple linear regression equation about the first peak area ratio and the weighted geometric mean relaxation time; the total acid content prediction model is a multiple linear regression equation about the second peak area ratio and the weighted geometric mean relaxation time; the total moisture content prediction model is a multiple linear regression equation about the third peak area ratio and the weighted geometric mean relaxation time.
2. The method for nondestructive detection of wolfberry puree quality according to claim 1, characterized in that: The set relaxation time is 330 ms; the set ratio is 99%.
3. The nondestructive testing method for wolfberry puree quality according to claim 1, characterized in that: According to the transverse relaxation time of each component of the wolfberry puree to be tested, a weighted geometric average is performed to obtain a weighted geometric mean relaxation time, specifically comprising: Using the formula Determine the weighted geometric mean relaxation time; where T 2,W is the weighted geometric mean relaxation time; W i is the normalized weight corresponding to the i-th component; T 2，i is the transverse relaxation time of the i-th component; n is the number of components in the wolfberry puree to be tested.
4. The method for nondestructive testing of wolfberry puree quality according to claim 1, wherein: The quality prediction model is: Total sugar content = k1*A1+k2*T 2,w +C; Total acid content = k1*A2+k2*T 2,w +C; Total moisture content = k1*A3+k2*T 2,w +C; Wherein, A1 is the first peak area ratio; A2 is the second peak area ratio; A3 is the third peak area ratio; T 2,w is the weighted geometric mean relaxation time; k1, k2, and C are the coefficients of the quality prediction model, and k1, k2, and C are obtained by optimizing the quality prediction model through the partial least squares method.
5. The nondestructive testing method for wolfberry puree quality according to claim 1, characterized in that: The construction process of the quality prediction model specifically includes: Establishing a training data set and an initial quality prediction model; the training data set includes the transverse relaxation time, the first peak area ratio, the second peak area ratio, the third peak area ratio, and the corresponding total sugar content, total acid content, and total moisture content of multiple groups of 100% pure wolfberry purees from different origins; Preprocessing the training data set to obtain a preprocessed training data set; Using a random forest algorithm or an XGBoost algorithm to filter the preprocessed training data set to obtain a filtered training data set; Based on the screened training data set, the initial quality prediction model was optimized using partial least squares method and 10-fold cross validation to make the correlation coefficient R 2 >0.9, the optimal coefficient is obtained; The quality prediction model is established based on the optimal coefficient.
6. The method for nondestructive testing of wolfberry puree quality according to claim 5, characterized in that: Preprocessing the training data set to obtain a preprocessed training data set specifically includes: performing standardization processing on the training data set to obtain a standardized training data set; Performing outlier detection on the standardized training data set using a median absolute deviation method or a box plot method, replacing outliers using a median or interpolation method to obtain an outlier-processed training data set; The training data set after outlier processing is processed using multicollinearity check to obtain a preprocessed training data set.
7. The method for nondestructive testing of wolfberry puree quality according to claim 5, characterized in that: Based on the screened training data set, the initial quality prediction model is optimized using partial least squares and 10-fold cross validation, which also includes: Determine the correlation coefficient R 2 Is it greater than 0.9? If so, the optimal coefficient is obtained; If not, the initial quality prediction model is optimized using kernel partial least squares, or the training data set is enhanced using random forest algorithm, gradient boosting decision tree or data enhancement algorithm, and then the initial quality prediction model is optimized using partial least squares and 10-fold cross validation until the correlation coefficient R 2 >0.
9.
8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for nondestructive detection of wolfberry puree quality according to any one of claims 1 to 7.
9. A 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 nondestructive detection of wolfberry puree quality according to any one of claims 1 to 7 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 nondestructive detection of wolfberry puree quality according to any one of claims 1 to 7 is implemented.