Selenium nutrition prediction and process optimization method for selenium-rich pork processing
By processing data and fusing models on cloud servers and terminals, the processing parameters of selenium-enriched pork are automatically optimized, solving the problems of selenium loss and process instability, improving selenium retention rate, selenomethionine ratio and bioavailability, and achieving efficient and accurate nutritional prediction and processing control.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GANNAN ACAD OF SCI
- Filing Date
- 2026-01-15
- Publication Date
- 2026-07-21
AI Technical Summary
In the current process of processing selenium-enriched pork, selenium loss is severe, the total selenium retention rate and the proportion of selenomethionine are reduced, bioavailability is decreased, and the processing parameters rely on human experience, resulting in poor reproducibility and control accuracy.
By working collaboratively between cloud servers and terminals, multi-source process parameter data is classified, decomposed, and feature-fused. Combined with multiple prediction models and optimization functions, the optimal process parameters are automatically selected to improve selenium retention rate, selenomethionine ratio, and bioavailability.
It improves the nutritional value and edible quality of selenium-enriched pork processing, enhances the perception accuracy of key nutrients, has good scalability and self-learning ability, and achieves precise optimization of process parameters.
Smart Images
Figure CN121960872B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the interdisciplinary field of food processing and nutritional analysis, and relates to, but is not limited to, a method for predicting selenium nutrition and optimizing the process of processing selenium-enriched pork. Background Technology
[0002] Insufficient selenium intake is associated with weakened immune function, cardiovascular health problems, and increased risk of certain chronic diseases. Selenium-enriched pork, as an effective dietary source of selenium, is of significant value and helps improve residents' selenium nutritional status. However, significant selenium loss occurs during processing. Selenium compounds, such as organic selenium, are easily volatilized or undergo form transformation during high-temperature processing, such as frying and baking. For example, selenomethionine is converted into selenite, resulting in a decrease in total selenium retention, a reduction in the proportion of selenomethionine, and ultimately, a decrease in bioavailability.
[0003] In related technologies, improved thermal processing techniques, such as low-temperature vacuum slow cooking, can synergistically improve the total selenium retention rate, the proportion of selenomethionine, and bioavailability. However, the process parameters, such as temperature and time, rely excessively on the experience of the operators, resulting in poor process reproducibility and control accuracy, which ultimately affects the stability of product quality.
[0004] Therefore, optimizing process parameters to achieve a synergistic improvement in total selenium retention, selenomethionine content, and bioavailability in selenium-enriched pork has become an urgent problem to be solved. Summary of the Invention
[0005] In view of this, the present invention provides a method for selenium nutrition prediction and process optimization in the processing of selenium-enriched pork, which at least solves the problem that related technologies cannot obtain optimal process parameters to achieve a synergistic improvement in total selenium retention rate, selenomethionine ratio and bioavailability in selenium-enriched pork.
[0006] According to a first aspect of the present invention, a method for predicting selenium nutrition and optimizing the processing of selenium-enriched pork is provided, applied to a cloud server, comprising: The receiving terminal sends a first process parameter, which includes water activity, temperature, time, and pH value; the first process parameter is the measured value at different time points during multiple processing of the pretreated selenium-enriched pork sample using the corresponding processing method. The second process parameter is classified in time and space to obtain classified data. The classified data is then decomposed, feature-filtered, and feature-fused to obtain fused features. The second process parameter is the result of preprocessing the first process parameter. The fused features are input into multiple initial prediction models to obtain the initial prediction results of the selenium-enriched pork samples under the second process parameters. The initial prediction results include the initial total selenium retention rate, the initial selenomethionine ratio, and the initial bioavailability. The initial prediction results are input into the fusion model to obtain the target prediction results; the target prediction results include the target total selenium retention rate, the target selenomethionine ratio, and the target bioavailability. Substitute the target prediction result into the preset target optimization function to obtain the total optimization value corresponding to each of the second process parameters, and obtain the target optimal process parameters based on the total optimization value; The target optimal process parameters and the corresponding target prediction results are sent to the terminal.
[0007] According to a second aspect of the present invention, a method for predicting selenium nutrition and optimizing the process of selenium-enriched pork processing is provided, applied to a terminal, comprising: The initial selenium content and the proportion of organic selenium were obtained by testing the selenium-enriched pork samples, and the corresponding processing method of the selenium-enriched pork samples was obtained based on the initial selenium content and the proportion of organic selenium. The selenium-enriched pork sample was cut and soaked in a reduced glutathione solution under preset conditions, and the reduced glutathione solution was subjected to ultrasonic treatment to obtain the ultrasonically treated selenium-enriched pork sample. After ultrasonic treatment, the selenium-enriched pork sample was subjected to a low-pressure pulse and then stored in a refrigerator at a preset temperature to obtain the treated selenium-enriched pork sample. The selenium-enriched pork sample was processed multiple times using the corresponding processing method, and the first process parameters at different time points were obtained by different types of sensors during the processing. The first process parameters included the measured values of water activity, temperature, time and pH. The first process parameters are sent to the cloud server; and the target optimal process parameters and corresponding target prediction results generated by the cloud server based on the first process parameters are received.
[0008] According to a third aspect of the present invention, an electronic device is provided, comprising: a processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other via the communication bus; the memory is used to store at least one executable instruction, wherein the executable instruction causes the processor to perform an operation corresponding to the method described in the first or second aspect.
[0009] According to a fourth aspect of the present invention, a computer storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the method as described in the first or second aspect.
[0010] According to the solution provided in the embodiments of the present invention, a first process parameter sent by a receiving terminal is included, which includes water activity, temperature, time, and pH value. The first process parameter is the measured value at different time points during multiple processing of a pretreated selenium-enriched pork sample using a corresponding processing method. The second process parameter is classified in time and space to obtain classified data, and the classified data is decomposed, feature-filtered, and feature-fused to obtain fused features. The second process parameter is the result of preprocessing the first process parameter. The fused features are input into multiple initial prediction models to obtain the initial prediction results of the selenium-enriched pork sample under the second process parameter, which include the initial total selenium retention rate, the initial selenomethionine ratio, and the initial bioavailability. The initial prediction results are input into the fusion model to obtain the target prediction result, which includes the target total selenium retention rate, the target selenomethionine ratio, and the target bioavailability. The target prediction result is substituted into a preset target optimization function to obtain the total optimization value corresponding to each of the second process parameters, and the target optimal process parameter is obtained based on the total optimization value. The target optimal process parameter and the corresponding target prediction result are sent to the terminal. In this process, by receiving multi-source process parameter data uploaded by the terminal and classifying, decomposing, and fusing features in the temporal and spatial dimensions, the dynamic changes of selenium-enriched pork during processing can be comprehensively captured. This not only improves the data representation ability but also enhances the perception accuracy of key nutrients, such as the state of selenomethionine, laying the foundation for subsequent accurate predictions. The fused features are input into multiple initial prediction models, and the initial prediction results of each model are integrated through a fusion model, effectively overcoming the problems of large prediction bias and weak generalization ability of single models, and improving the accuracy and stability of nutritional indicator predictions. Furthermore, the prediction results are substituted into a preset target optimization function to quantitatively evaluate the comprehensive performance of each set of process parameters, automatically selecting the optimal target process parameters that maximize selenium retention and bioavailability, improving the efficiency and accuracy of obtaining the optimal target process parameters. The optimal target process parameters and their corresponding prediction results are fed back to the terminal to guide users or production equipment to execute the optimal processing plan. This not only improves the nutritional value and edible quality of selenium-enriched pork but also possesses good scalability and self-learning ability, providing an efficient and accurate technical path for the intelligent processing of functional foods. Attached Figure Description
[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, wherein: Figure 1 A flowchart illustrating a method for predicting selenium nutrition and optimizing the process of selenium-enriched pork processing, provided in an embodiment of the present invention. Figure 1 ; Figure 2 A flowchart illustrating a method for predicting selenium nutrition and optimizing the process of selenium-enriched pork processing, provided in an embodiment of the present invention. Figure 2 ; Figure 3 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention. Detailed Implementation
[0012] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. The following embodiments are used to illustrate the present invention, but are not intended to limit the scope of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0013] In the following description, references are made to “some embodiments,” which describe a subset of all possible embodiments. However, it is understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.
[0014] It should be noted that the terms "first, second, and third" used in the embodiments of the present invention are only used to distinguish similar objects and do not represent a specific ordering of objects. It is understood that "first, second, and third" can be interchanged in a specific order or sequence where permitted, so that the embodiments of the present invention described herein can be implemented in an order other than that illustrated or described herein.
[0015] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which these embodiments of the invention pertain. It should also be understood that terms such as those defined in general dictionaries should be understood to have the same meaning as in the context of the prior art and should not be interpreted in an idealized or overly formal sense unless specifically defined as herein.
[0016] Figure 1 A flowchart illustrating a method for predicting selenium nutrition and optimizing the process of selenium-enriched pork processing, provided in an embodiment of the present invention. Figure 1 The selenium nutrition prediction and process optimization method for processing selenium-enriched pork provided in this embodiment of the invention can be executed by an electronic device, such as a cloud server.
[0017] like Figure 1 As shown, the method for selenium nutrition prediction and process optimization in the processing of selenium-enriched pork applied to cloud servers includes: S101. Receive the first process parameters sent by the terminal. The first process parameters include water activity, temperature and pH value. The first process parameters are the measured values at different time points during multiple processing of the pretreated selenium-enriched pork sample using the corresponding processing method.
[0018] In embodiments of the present invention, the corresponding processing methods can be frying or steaming. The first process parameter is multi-source data collected at various time points during multiple processing cycles of the pretreated selenium-enriched pork using the corresponding processing method, including but not limited to water activity, temperature, time, and pH value. Water activity is the proportion of free water in the pretreated selenium-enriched pork sample; temperature is the thermal state of the processing environment or the interior of the meat; pH value is the acidity or alkalinity of the tissue fluid or processing solution of the pretreated selenium-enriched pork sample; and time is a point in time during the processing. This first process parameter is detected by the terminal and received by the cloud server.
[0019] The first process parameters include water activity, temperature, and pH value, but are not limited to these parameters.
[0020] S102. Classify the second process parameter in time and space to obtain classified data, and decompose, filter and fuse the classified data to obtain fused features; the second process parameter is the result of preprocessing the first process parameter.
[0021] In an embodiment of the present invention, the first process parameter is preprocessed to obtain the second process parameter, the second process parameter is classified in time and space to obtain classified data, and then the classified data is decomposed, feature filtered and feature fused to obtain fused features.
[0022] Time includes instantaneous moments and phases, and space includes molecules and cells.
[0023] S103. Input the fused features into multiple initial prediction models to obtain the initial prediction results of the corresponding selenium-enriched pork samples in the second process parameters. The initial prediction results include the initial total selenium retention rate, the initial selenomethionine ratio, and the initial bioavailability.
[0024] In embodiments of the present invention, the initial detection model, i.e., the base model, can be a random forest, neural network, or support vector machine, etc. The fused features are input into multiple initial prediction models. Each initial prediction model outputs a corresponding initial prediction result for different second process parameters. That is, each second process parameter corresponds to multiple sets of initial prediction results. The initial prediction results include the initial total selenium retention rate, the initial selenomethionine ratio, and the initial bioavailability. Among them, the total selenium retention rate is the percentage of total selenium retained in the processed selenium-enriched pork sample relative to the total selenium content before processing; the selenomethionine ratio is the percentage of selenium in the form of selenomethionine in the selenium-enriched pork sample relative to the total selenium content; and the bioavailability is the degree to which the selenium in the selenium-enriched pork sample is digested, absorbed, and utilized by the human body after consumption.
[0025] If the initial prediction model is a neural network, the number of hidden layers and neurons is no longer fixed, ranging from 2-4 layers and 32-128 neurons, respectively. When training the initial prediction model, historical information such as the selenoprotein thermal denaturation trajectory simulated by GROMACS software, experimental data from 50 sets of experimental parameters under 6 different process conditions, and absorption experimental data from human colon organoid microarrays are integrated. After data preprocessing and feature extraction, a comprehensive dataset for model training is constructed.
[0026] S104. Input the initial prediction results into the fusion model to obtain the target prediction results; the target prediction results include the target total selenium retention rate, the target selenomethionine ratio, and the target bioavailability.
[0027] In an embodiment of the present invention, the fusion model can be a meta-model (i.e., a combiner in ensemble learning). Multiple sets of initial prediction results corresponding to each second process parameter are input into the fusion model for analysis and fusion to obtain the target prediction results corresponding to each second process parameter. The target prediction results include the target total selenium retention rate, the target selenomethionine ratio, and the target bioavailability.
[0028] S105. Substitute the target prediction result into the preset target optimization function to obtain the total optimization value corresponding to each of the second process parameters, and obtain the target optimal process parameters based on the total optimization value.
[0029] In embodiments of the present invention, a multi-objective optimization function is established based on the target prediction results. This multi-objective optimization function takes into account both selenium and energy consumption indicators. The target prediction results corresponding to each of the second process parameters are substituted into the preset target optimization function to obtain a total optimized value for each of the second process parameters. Finally, the total optimized values are compared to obtain the optimal target process parameter among multiple second process parameters. The optimal target process parameter is the comprehensive optimal solution obtained after weighing multiple objectives such as total selenium retention rate, selenomethionine ratio, bioavailability, processing efficiency, and energy consumption. It aims to achieve synergistic optimization of nutritional quality, production efficiency, and energy consumption. Processing power and energy consumption can be directly obtained during processing. The target optimization function is as follows: ; In the above formula, , , , and For the corresponding weights, This represents the total optimized value.
[0030] S106. Send the target optimal process parameters and the corresponding target prediction results to the terminal.
[0031] In an embodiment of the present invention, the target optimal process parameters and the target prediction results corresponding to the target optimal process parameters are sent to the terminal and stored. Subsequently, when processing the current selenium-enriched pork, the target optimal process parameters corresponding to different types of selenium-enriched pork can be selected from the stored target optimal process parameters, and the selected optimal process parameters can be used to process the current selenium-enriched pork to achieve the optimal total selenium retention rate, selenomethionine ratio and bioavailability.
[0032] It is understood that, in the embodiments of the present invention, by receiving multi-source process parameter data uploaded by the receiving terminal, such as water activity, temperature, time, and pH value, and classifying, decomposing, and fusing features in the temporal and spatial dimensions, the dynamic changes of selenium-enriched pork during processing can be comprehensively captured. This not only improves the data characterization ability but also enhances the perception accuracy of key nutrients, such as the state of selenomethionine, laying the foundation for subsequent accurate prediction. The fused features are input into multiple initial prediction models, and the initial prediction results of each model are integrated through a fusion model, effectively overcoming the problems of large prediction bias and weak generalization ability of single models, and improving the accuracy and stability of nutritional indicator prediction. Furthermore, the prediction results are substituted into a preset target optimization function to quantitatively evaluate the comprehensive performance of each set of process parameters, automatically selecting the optimal target process parameters that maximize selenium retention and bioavailability, thus improving the efficiency and accuracy of obtaining the optimal target process parameters. The optimal process parameters and their corresponding prediction results are fed back to the terminal for storage. When processing selenium-enriched pork, the optimal process parameters corresponding to different types of selenium-enriched pork can be selected to improve processing efficiency, nutritional value and edible quality of selenium-enriched pork. It also has good scalability and self-learning ability, providing an efficient and precise technical path for the intelligent processing of functional foods.
[0033] In some embodiments of the present invention, the decomposition, feature filtering and feature fusion of the classified data in S102 to obtain the fused features can be achieved through S1021 to S1022, which will be described through the following steps.
[0034] S1021. Use multi-scale wavelet decomposition to decompose each type of data and obtain the features at each scale.
[0035] In some embodiments of the present invention, multi-scale wavelet decomposition is employed to decompose each type of data, resulting in multiple components, including a low-frequency trend component that reflects slow changes, such as a continuously slow heating process. Simultaneously, the multiple components also include high-frequency detail components that reflect instantaneous fluctuations, such as temperature oscillations. In this way, local features of the signal can be extracted from both time and frequency dimensions, effectively capturing key stages in the processing, such as initial heating, stabilization, and final cooling, ultimately yielding features at each scale.
[0036] S1022. Feature selection is performed by utilizing the correlation between features at different scales to obtain target features; and the target features are then fused to obtain fused features.
[0037] In some embodiments of the present invention, statistical quantities such as mean, variance, energy, and entropy are extracted from the features obtained at each scale through multi-scale wavelet decomposition as candidate features. Subsequently, the correlation between these features is calculated using methods such as Pearson correlation coefficient or mutual information. Features with high correlation are retained, while redundant or irrelevant low-correlation features are removed to obtain the target features. The target features are then fused, using methods such as simple vector concatenation, principal component analysis, and linear discriminant analysis, to finally obtain the fused features.
[0038] In some embodiments of the present invention, obtaining the target optimal process parameters based on the total optimization value in S105 can be achieved through S1051 to S1052, as described in the following steps.
[0039] S1051. Take the second process parameter corresponding to the maximum total optimization value in the total optimization value as the initial optimal process parameter. If the total carbon emissions calculated using the initial optimal process parameter meet the preset conditions, then take the initial optimal process parameter as the target optimal process parameter.
[0040] In some embodiments of the present invention, the total optimization values are sorted, the largest total optimization value is found, and the second process parameter corresponding to the largest total optimization value is used as the initial optimal process parameter. An environmental impact assessment is carried out, and the total carbon emissions are calculated by inputting the initial optimal process parameter using a carbon footprint calculation tool. If the calculated total carbon emissions meet the preset conditions, which can be a set threshold, then the initial optimal process parameter is used as the target optimal process parameter.
[0041] S1052. If the total carbon emissions calculated using the initial optimal process parameters do not meet the preset conditions, the second process parameter corresponding to the second largest total optimal value in the total optimal value will be used again as the initial optimal process parameter until the target optimal process parameter is found.
[0042] In some embodiments of the present invention, if the total carbon emissions calculated using the initial optimal process parameters do not meet the preset conditions, the second largest overall optimization value, which is second to the optimal overall optimization value, is found in the ranking results. The second process parameter corresponding to the second largest overall optimization value is used again as the initial optimal process parameter, and then the total carbon emissions are calculated. When the preset conditions are met, the second process parameter corresponding to the second largest overall optimization value is used as the target optimal process parameter. If the preset conditions are not met, the second process parameter corresponding to the third largest overall optimization value, which is second to the second largest overall optimization value, is used again as the initial optimal process parameter. This process is repeated until the target optimal process parameter is found.
[0043] In some embodiments of the present invention, S201 to S202 are included after S106, as described by the following steps.
[0044] S201. Receive the scoring data sent by the terminal and add the scoring data to the original training set to obtain a new training set; the scoring data is the feedback data of the user's consumption of the actual selenium-enriched pork after processing based on the stored target optimal process parameters and target prediction results.
[0045] In some embodiments of the present invention, selenium-enriched pork is processed based on suitable optimal process parameters and target prediction results from the stored target optimal process parameters and then supplied to consumers for consumption. After consumption, consumers give a rating based on dimensions such as taste, nutrition, and flavor, for example, 1 to 5 points. Then, the selected optimal process parameters, the corresponding processing method, the target prediction results, and the corresponding selenium-enriched pork type are combined to construct rating data, which is then sent to the cloud server. After receiving the rating data sent by the terminal, the cloud server treats it as a new labeled sample and merges it into the original training set to obtain a new training set.
[0046] S202. Based on the new training set, retrain multiple initial prediction models and fusion models to obtain multiple trained initial prediction models and trained fusion models.
[0047] In some embodiments of the present invention, multiple initial prediction models that have been trained using the original training set are retrained using a new training set, and the fusion model is retrained using the output results of the multiple initial prediction models, so as to obtain multiple trained initial prediction models and a trained fusion model.
[0048] Figure 2 This is a flowchart illustrating a method for predicting selenium nutrition and optimizing the process of processing selenium-enriched pork according to an embodiment of the present invention. This method can be executed by an electronic device, such as a terminal.
[0049] like Figure 2 As shown, the method for selenium nutrition prediction and process optimization applied to the processing of selenium-enriched pork at the terminal stage includes: S301. Test the selenium-enriched pork sample to obtain the initial selenium content and the proportion of organic selenium, and obtain the corresponding processing method for the selenium-enriched pork sample based on the initial selenium content and the proportion of organic selenium.
[0050] In an embodiment of the present invention, multiple groups of selenium-enriched pork samples were collected, digested with a mixture of nitric acid and perchloric acid, and then brought to a constant volume. The initial selenium content was determined according to the reference standard GB 5009.93-2017. Simultaneously, enzyme extraction combined with high-performance liquid chromatography-inductively coupled plasma mass spectrometry was used to determine the inorganic selenium content of each group of selenium-enriched pork samples. Furthermore, the proportion of organic selenium was calculated using the initial selenium content and the proportion of organic selenium. Based on the initial selenium content and the proportion of organic selenium in the selenium-enriched pork samples, combined with a prediction model, the corresponding processing method for each selenium-enriched pork sample could be determined. For example, if the initial selenium content was confirmed to be 0.52 mg / kg and the proportion of organic selenium was 96% (≥95%), the corresponding processing method was determined to be frying and baking.
[0051] S302. After the selenium-enriched pork sample is cut, it is soaked in a reduced glutathione solution under preset conditions, and the reduced glutathione solution is subjected to ultrasonic treatment to obtain ultrasonically treated selenium-enriched pork sample slices.
[0052] S303. After applying a low-pressure pulse to the ultrasonically treated selenium-enriched pork sample, it is placed in a refrigerator at a preset temperature for storage to obtain the treated selenium-enriched pork sample.
[0053] In an embodiment of the present invention, a composite stabilization treatment was carried out. The selenium-enriched pork sample was cut into pieces using an ultrasonic cleaner, a high-voltage pulsed electric field device, and a low-temperature freezer. The cut selenium-enriched pork sample was then immersed in a 0.1 mol / L reduced glutathione solution refrigerated at 4°C. After ultrasonic treatment at 20 kHz and 300 W for 10 minutes (while maintaining the solution at 4°C), it was placed in a pulsed electric field device to apply a low-voltage pulse of 200 V / cm. Immediately after treatment, it was placed in a liquid nitrogen freezer at -80°C and frozen for 30 minutes to obtain the treated selenium-enriched pork sample.
[0054] S304. The processed selenium-enriched pork sample is processed multiple times using the corresponding processing method, and the first process parameters at different time points are obtained through different types of sensors during the processing. The first process parameters include the measured values of water activity, temperature, time and pH value.
[0055] In an embodiment of the present invention, during the processing of the selenium-enriched pork sample using the corresponding processing method, a PT1000 temperature sensor is embedded in the heating plate, a capacitive water activity meter is placed inside the equipment, and an antimony electrode pH meter is inserted into the center of the meat slice. The first process parameters are obtained according to the above sensors at different time periods. The first process parameters include, but are not limited to, water activity, temperature, time, and pH value.
[0056] In response to equipment malfunctions such as temperature sensor failure or heating element damage, or in emergency situations involving selenium-enriched pork samples (e.g., initial selenium content below 10% of the standard value, abnormal pH levels), the terminal has multiple pre-set backup adjustment schemes to handle different situations. Specifically, after a sudden power outage and restoration, the system will automatically activate corresponding supplementary process parameters based on the duration of the power outage. If the power outage lasts no more than 5 minutes, the supplementary processing time will be 20% of the original time; if the power outage lasts between 5 and 10 minutes, the supplementary processing time will be adjusted to 40% of the original time, thus ensuring the continuity and stability of the processing. For example, if the temperature sensor suddenly malfunctions during frying, displaying -10℃, an emergency mechanism will be activated within 10 seconds, switching to a backup infrared temperature sensor to monitor the actual temperature and dynamically adjusting the process parameters accordingly, such as lowering the temperature to 160℃ and shortening the time by 2 minutes.
[0057] S305. Send the first process parameters to the cloud server; and receive the target optimal process parameters and corresponding target prediction results generated by the cloud server based on the first process parameters.
[0058] In an embodiment of the present invention, after the terminal sends the first process parameters to the cloud server, it receives the target optimal process parameters and the target prediction results corresponding to the target optimal process parameters generated by the cloud server based on the first process parameters, and stores them in combination with their respective corresponding processing methods, so that the actual processing of selenium-enriched pork can be carried out according to the stored data in the future.
[0059] In some embodiments of the present invention, S305 is followed by S401 to S402, which are described by the following steps.
[0060] S401. Receive the rating data sent by the user and send the rating data to the cloud server. The rating data is the feedback data of the user after consuming the actual selenium-enriched pork based on the stored target optimal process parameters and target prediction results.
[0061] In some embodiments of the present invention, when processing selenium-enriched pork, processing can be carried out based on the data stored in the terminal, selecting the corresponding processing method, the target optimal process parameters, and the target prediction results. After consuming the processed selenium-enriched pork, consumers give scores based on dimensions such as taste, nutritional perception, and meat quality, and the terminal receives the scores sent by the consumers.
[0062] Among these methods, consumer feedback on the taste of selenium-enriched pork products is collected through QR code scanning, with ≥1000 feedback data points collected weekly and a data validity rate of ≥90%. This data is used as auxiliary parameters for model iteration on the cloud server. When the consumer satisfaction score for the target optimal process parameters falls below 3 points, the model on the cloud server is retrained to optimize the model and improve the product's market adaptability. The model iteration cycle is ≤7 days.
[0063] S402: Receive the new target optimal process parameters and corresponding new target prediction results sent by the cloud server based on the score data adjustment.
[0064] In some embodiments of the present invention, the terminal constructs scoring data by combining user feedback ratings, selected optimal process parameters, corresponding processing methods, target prediction results, and corresponding selenium-enriched pork types. This scoring data is then sent to a cloud server. The cloud server merges this new scoring data with historical training data to form an updated dataset, which is used to retrain or fine-tune the original initial prediction model and fusion model. In this way, the model can learn which parameter combinations are more popular with users in real-world applications, thereby continuously improving prediction accuracy and personalized recommendation capabilities. After updating the model, the cloud server recalculates and distributes new optimal process parameters, such as adjusted temperature and time combinations, along with corresponding new target prediction results. The terminal receives these updated parameters and prediction results to guide the next processing recommendation.
[0065] Reference Figure 3 The diagram shows a structural schematic of an electronic device according to an embodiment of the present invention. The specific embodiments of the present invention do not limit the specific implementation of the electronic device.
[0066] like Figure 3 As shown, the electronic device may include: a processor 502, a communications interface 504, a memory 506, and a communications bus 508.
[0067] in: The processor 502, communication interface 504, and memory 506 communicate with each other via communication bus 508.
[0068] Communication interface 504 is used to communicate with other electronic devices or servers.
[0069] The processor 502 is used to execute program 510, specifically the relevant steps in the above method embodiments.
[0070] Specifically, program 510 may include program code that includes computer operation instructions.
[0071] Processor 502 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention. The smart device may include one or more processors of the same type, such as one or more CPUs; or it may include processors of different types, such as one or more CPUs and one or more ASICs.
[0072] Memory 506 is used to store program 510. Memory 506 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.
[0073] Specifically, program 510 can be used to cause processor 502 to perform the operations corresponding to the methods described in the above method embodiments.
[0074] The specific implementation of each step in program 510 can be found in the corresponding descriptions of the steps and units in the above method embodiments, and will not be repeated here. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the devices and modules described above can be referred to the corresponding process descriptions in the foregoing method embodiments, and will not be repeated here.
[0075] It should be noted that, depending on the implementation needs, the various components / steps described in the embodiments of the present invention can be broken down into more components / steps, or two or more components / steps or parts of the operation of components / steps can be combined into new components / steps to achieve the purpose of the embodiments of the present invention.
[0076] The methods described above according to embodiments of the present invention can be implemented in hardware, firmware, or as software or computer code that can be stored in a recording medium (such as a CD-ROM, RAM, floppy disk, hard disk, or magneto-optical disk), or as computer code originally stored on a remote recording medium or a non-transitory machine-readable medium and subsequently stored on a local recording medium, downloaded via a network. Thus, the methods described herein can be processed by software stored on a recording medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware (such as an ASIC or FPGA). It is understood that the computer, processor, microprocessor controller, or programmable hardware includes storage components (e.g., RAM, ROM, flash memory, etc.) capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods described herein. Furthermore, when a general-purpose computer accesses code used to implement the methods shown herein, the execution of the code transforms the general-purpose computer into a dedicated computer for executing the methods shown herein.
[0077] Those skilled in the art will recognize that the units and method steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of the embodiments of the present invention.
[0078] The above embodiments are only used to illustrate the embodiments of the present invention, and are not intended to limit the embodiments of the present invention. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the embodiments of the present invention. Therefore, all equivalent technical solutions also fall within the scope of the embodiments of the present invention, and the patent protection scope of the embodiments of the present invention should be defined by the claims.
Claims
1. A method for predicting selenium nutrition and optimizing the processing of selenium-enriched pork, characterized in that, Applications on cloud servers include: The receiving terminal sends a first process parameter, which includes water activity, temperature, time, and pH value; the first process parameter is the measured value at different time points during multiple processing of the pretreated selenium-enriched pork sample using the corresponding processing method. The second process parameter is classified in time and space to obtain classified data. The classified data is then decomposed, feature-filtered, and feature-fused to obtain fused features. The second process parameter is the result of preprocessing the first process parameter. The fused features are input into multiple initial prediction models to obtain the initial prediction results of the selenium-enriched pork samples under the second process parameters. The initial prediction results include the initial total selenium retention rate, the initial selenomethionine ratio, and the initial bioavailability. The initial prediction results are input into the fusion model to obtain the target prediction results; the target prediction results include the target total selenium retention rate, the target selenomethionine ratio, and the target bioavailability. Substitute the target prediction result into the preset target optimization function to obtain the total optimization value corresponding to each of the second process parameters, and obtain the target optimal process parameters based on the total optimization value; The target optimal process parameters and the corresponding target prediction results are sent to the terminal; The step of substituting the target prediction result into a preset target optimization function to obtain the total optimized value corresponding to each of the second process parameters includes: obtaining the processing power and energy consumption during the processing, and substituting the processing power, energy consumption, and the corresponding target prediction result into the target optimization function to obtain the total optimized value; wherein, the target optimization function is as follows: ； In the above formula, , , , and For the corresponding weights, This represents the total optimized value.
2. The method according to claim 1, characterized in that, The process of decomposing, filtering, and fusing the classified data to obtain fused features includes: Multi-scale wavelet decomposition is used to decompose each type of data to obtain features at each scale; Feature filtering is performed using the correlation between features at each scale to obtain target features; and the target features are then fused to obtain fused features.
3. The method according to claim 1, characterized in that, The process of obtaining the target optimal process parameters based on the total optimized value includes: The second process parameter corresponding to the maximum total optimization value among the total optimization values is taken as the initial optimal process parameter. If the total carbon emissions calculated using the initial optimal process parameter meet the preset conditions, then the initial optimal process parameter is taken as the target optimal process parameter. If the total carbon emissions calculated using the initial optimal process parameters do not meet the preset conditions, the second process parameter corresponding to the second largest total optimal value in the total optimal value will be used again as the initial optimal process parameter until the target optimal process parameter is found.
4. The method according to claim 1, characterized in that, After sending the target optimal process parameters and the corresponding target prediction results to the terminal, the method further includes: The system receives the scoring data sent by the terminal and adds the scoring data to the original training set to obtain a new training set. The scoring data is the feedback data from the user's consumption of the actual selenium-enriched pork after processing, based on the stored target optimal process parameters and target prediction results. The multiple initial prediction models and the fusion model are retrained based on the new training set to obtain the multiple trained initial prediction models and the trained fusion model.
5. A method for predicting selenium nutrition and optimizing the processing of selenium-enriched pork, characterized in that, Applied to terminals, including: The initial selenium content and the proportion of organic selenium were obtained by testing the selenium-enriched pork samples, and the corresponding processing method of the selenium-enriched pork samples was obtained based on the initial selenium content and the proportion of organic selenium. The selenium-enriched pork sample was cut and soaked in a reduced glutathione solution under preset conditions, and the reduced glutathione solution was subjected to ultrasonic treatment to obtain the ultrasonically treated selenium-enriched pork sample. After ultrasonic treatment, the selenium-enriched pork sample was subjected to a low-pressure pulse and then stored in a refrigerator at a preset temperature to obtain the treated selenium-enriched pork sample. The selenium-enriched pork sample was processed multiple times using the corresponding processing method, and the first process parameters at different time points were obtained by different types of sensors during the processing. The first process parameters included the measured values of water activity, temperature, time and pH. Send the first process parameters to the cloud server; and receive the target optimal process parameters and corresponding target prediction results generated by the cloud server based on the first process parameters. The system receives rating data sent by users and sends the rating data to the cloud server. The rating data is the feedback data from users after consuming the actual selenium-enriched pork processed based on the stored target optimal process parameters and target prediction results. Receive the new target optimal process parameters and corresponding new target prediction results, adjusted based on the scoring data, sent by the cloud server.