Optimization processing methods for distiller's grains formula data, computer equipment and storage media

Abstract

The present invention relates to an optimization processing method for distiller's grains formula data, a computer device and a storage medium, and relates to the field of data processing, including: responding to a distiller's grains formula request generated by a user end, which includes a target animal pregnancy identifier, a basic diet formula matrix and an initial distiller's grains formula matrix set. According to the pregnancy identifier, a corresponding nutritional value parser is matched from a model library of a server. Through the parser, the initial distiller's grains formula matrix set is traversed based on the basic diet formula matrix to obtain a nutritional value score value set. According to the score value, and in combination with the parser and the basic diet formula matrix, the initial formula matrix is ​​optimized to generate a recommended distiller's grains formula matrix sequence. Finally, the recommended formula sequence is sent to a visual interface of the user end for feedback display.

Description

Technical Field

[0001] The invention relates to the field of data processing, and in particular to an optimization processing method for distiller's grains formula data, a computer device and a storage medium. Background Art

[0002] In the process of animal husbandry, the nutritional needs of animals change with their growth stage and physiological state, especially during pregnancy, when the nutritional needs of animals become more refined. As an important component of feed, the nutritional value and applicability of distiller's grains in animal husbandry are increasingly receiving attention. In the existing formulation process, the production and addition of distiller's grains are mostly based on the experience of breeders, which often makes it difficult to quickly respond to the nutritional needs of animals at different stages, especially for the personalized needs of pregnant animals. There is a lack of effective formula optimization processing methods, and there are technical problems such as difficulty in formula optimization, low optimization efficiency, and poor adaptability. Summary of the invention

[0003] The present invention aims to solve the technical problems of great difficulty in formula optimization, low optimization efficiency and poor adaptability in the prior art, and provides an optimization processing method for distiller's grains formula data, a computer device and a storage medium.

[0004] The technical solution of the present invention to solve the above technical problems is as follows:

[0005] The present invention provides a method for optimizing distiller's grains formula data, comprising:

[0006] The method responds to a distiller's grains formula request generated by a user end, wherein the distiller's grains formula request includes a target animal pregnancy stage identifier, a basic diet formula matrix, and an initial distiller's grains formula matrix set.

[0007] According to the pregnancy period identifier of the target animal, a nutritional value parser is matched from the model library of the server.

[0008] The nutritional value parser is used to traverse the initial distiller's grains formula matrix set for parsing based on the basic diet formula matrix to obtain a nutritional value score value set.

[0009] According to the nutritional value score value set, in combination with the nutritional value parser and the basic diet formula matrix, the initial distiller's grains formula matrix set is optimized to obtain a recommended distiller's grains formula matrix sequence.

[0010] The recommended distiller's grains formula matrix sequence is sent to the visual interface of the user end for display feedback.

[0011] The beneficial effects of the present invention are as follows: by responding to a distiller's grains formula request generated by a user terminal, wherein the distiller's grains formula request includes a pregnancy identification of a target animal, a basic diet formula matrix, and an initial distiller's grains formula matrix set. According to the pregnancy identification of the target animal, a suitable nutritional value parser is matched from the model library of the server. Using the nutritional value parser, based on the basic diet formula matrix, the initial distiller's grains formula matrix set is traversed and parsed to generate a nutritional value score value set. According to the nutritional value score value set, the initial distiller's grains formula matrix set is optimized in combination with the nutritional value parser and the basic diet formula matrix to generate a recommended distiller's grains formula matrix sequence. The recommended distiller's grains formula matrix sequence is displayed and fed back through a visual interface of the user terminal. Thereby achieving the technical effect of reducing the difficulty of formula optimization and improving the efficiency and adaptability of formula development. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 A schematic flow chart of the method for optimizing the processing of distiller's grains formula data provided by the present invention;

[0013] Figure 2 A schematic diagram of a process for obtaining a set of nutritional value scoring values ​​in the method for optimizing distiller's grains formula data provided by the present invention;

[0014] Figure 3 A schematic diagram of the structure of an electronic device provided by the present invention;

[0015] Figure 4 A schematic diagram of the structure of a computer-readable storage medium provided by the present invention.

[0016] In the accompanying drawings, the components represented by the reference numerals are listed as follows:

[0017] Electronic device 500 , memory 510 , processor 520 , computer program 511 , computer readable storage medium 600 , computer program 611 . DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0019] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0020] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or explanation". Any embodiment described as "for example" in the present invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any technician in the field to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes will not be elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed in the present invention.

[0021] Embodiment 1:

[0022] like Figure 1 As shown, an embodiment of the present invention provides an optimization processing method for distiller's grains formula data.

[0023] Responding to a distiller's grains formula request generated by a user end, wherein the distiller's grains formula request includes a target animal pregnancy stage identifier, a basic diet formula matrix, and an initial distiller's grains formula matrix set;

[0024] Specifically, first, obtain the target user's distiller's grains formula request, which is generated by the target user's user end, and is based on the animal's nutritional needs, production stage and existing basic diet formula to generate a distiller's grains feed formula suitable for the target animal. Among them, the target animal's pregnancy identification is used to clarify the target animal's reproductive stage, especially which stage of pregnancy the animal is in. The animal's pregnancy stage affects its nutritional needs and thus affects the most suitable distiller's grains formula. The basic diet formula matrix is ​​the existing feed formula data, which usually contains different raw materials and their nutrients. The basic diet formula matrix is ​​the basic data for generating distiller's grains formulas and is the basis for adding distiller's grains. The initial distiller's grains formula matrix set is a historical or general distiller's grains formula data set, including a variety of distiller's grains formulas, which is used to provide reference formulas and nutritional parameters for subsequent distiller's grains formula optimization.

[0025] Exemplarily, the specific stages of pregnancy of the target animal include: early pregnancy, the animal needs more energy and protein to support the initial development of the embryo. Mid-pregnancy, the energy demand of mid-pregnant animals is relatively stable, but the intake of minerals and vitamins needs to be appropriately increased. Late pregnancy, the demand for nutrients such as protein, calcium, and phosphorus of late-pregnant animals further increases to support the rapid growth of the embryo.

[0026] The pregnancy markers, basic diets and historical distiller's grains formulas of the target animals obtained in the above steps can help generate personalized distiller's grains formulas to ensure that the formulas match the actual needs of the animals.

[0027] According to the pregnancy period identifier of the target animal, a nutritional value parser is matched from the model library of the server;

[0028] Specifically, different animals (such as cattle, pigs, and sheep) have significantly different requirements for protein, energy, minerals, vitamins, etc. in early, mid, and late pregnancy. Therefore, the pregnancy identification of target animals includes the specific type of animal (such as cattle, pigs, sheep, etc.) and its pregnancy stage (early, mid, and late pregnancy).

[0029] Specifically, the model library in the server stores multiple nutritional value analysis models. Each model is built according to the physiological state of different animals (such as breeding period, resting period, pregnancy, etc.) and their nutritional needs, and can analyze the nutritional components and corresponding health effects of different distiller's grains or feed formulas. In other words, the nutritional value parser is a nutritional value evaluation model in the model library, which is used to calculate and analyze the nutritional value of different distiller's grains formulas for target animals in a specific pregnancy period, so as to efficiently predict and optimize complex distiller's grains formulas, which helps to ensure that the distiller's grains formula meets the nutritional needs of target animals at a specific growth stage or pregnancy stage.

[0030] Optionally, the nutritional value parser is a regression analysis model or a machine learning model constructed by a data-driven approach based on existing nutritional value assessment results. The existing nutritional value assessment results are derived from historical distiller's grains formula experiments, nutrient composition analysis, animal growth data, etc. By performing regression analysis and machine learning model training on these data, a parser capable of predicting new distiller's grains formulas is generated.

[0031] By means of the nutritional value parser, based on the basic diet formula matrix, the initial distiller's grains formula matrix set is traversed for parsing to obtain a nutritional value score value set;

[0032] Specifically, the nutritional value parser analyzes the input distiller's grains formula and estimates the specific content of nutrients such as protein, fiber, energy, minerals, vitamins, etc. in the corresponding fermentation products in combination with the basic diet formula matrix. Then, the nutritional components of the formula are compared with the nutritional requirements of the target animal during pregnancy to determine whether the formula is excessive or insufficient, thereby determining the nutritional value of the distiller's grains formula for the target animal during pregnancy in the distiller's grains formula request, and achieving the technical effect of efficiently predicting and evaluating complex distiller's grains formulas.

[0033] In some embodiments, Figure 2 As shown, through the nutritional value parser, based on the basic diet formula matrix, the initial distiller's grains formula matrix set is traversed for parsing to obtain a nutritional value score value set, including:

[0034] The nutritional value parser includes a nutrient index retrieval layer and a nutritional value assessment layer; extracting a first initial distiller's grains formula matrix of the initial distiller's grains formula matrix set; indexing the first initial distiller's grains formula matrix through the nutrient index retrieval layer to obtain a first nutrient index matrix; at the same time, indexing the basic diet formula matrix through the nutrient index retrieval layer to obtain a second nutrient index matrix; receiving the first nutrient index matrix through the distiller's grains formula input channel of the nutritional value assessment layer, receiving the second nutrient index matrix through the diet formula input channel of the nutritional value assessment layer, and obtaining a first nutritional value score value; adding the first nutritional value score value to the nutritional value score value set.

[0035] Specifically, the nutritional value parser includes a nutrient index retrieval layer and a nutritional value assessment layer, wherein the nutrient index retrieval layer is used to extract and index the nutrient components of the input formula matrix (including distiller's grains formula and basic diet formula). Then, multiple key nutrient information is obtained, such as the content of dry matter, protein, energy, medium detergent fiber, amino acids (such as aspartic acid, threonine, serine, glutamic acid, proline, etc.), minerals, etc. The nutritional value assessment layer is used to comprehensively evaluate the nutritional value of the nutrient index matrix and generate a specific nutritional value score. In other words, this layer is responsible for analyzing, comparing, and calculating the nutrient components from distiller's grains formula and diet formula, and evaluating their effectiveness in meeting the nutritional needs of target animals.

[0036] Specifically, the first initial distiller's grains formula matrix of the initial distiller's grains formula matrix set is extracted, and the initial distiller's grains formula matrix set is a set of multiple distiller's grains formula matrices used for distiller's grains formula optimization in history, including different distiller's grains formulas and their corresponding nutritional information. The most suitable formula is screened by traversing multiple initial distiller's grains formulas in the initial distiller's grains formula matrix set. The first initial distiller's grains formula matrix is ​​any one formula in the initial distiller's grains formula matrix set, which is obtained by non-repetitive random extraction.

[0037] Specifically, the first initial distiller's grains formula matrix contains all the nutritional information of the ingredients of the first distiller's grains formula and the fermentation products corresponding to the first distiller's grains formula, which is the basic input for the nutritional value score. The first nutrient index matrix is ​​a numerical matrix of key nutrient indexes obtained by the nutrient index retrieval layer through index screening of the initial distiller's grains formula matrix according to the nutritional needs of the target animal at a specific growth stage or pregnancy stage. In other words, the first nutrient index matrix contains the index data of the fermentation products corresponding to the first distiller's grains formula that make the main contribution to the current pregnancy of the target animal. Through the above steps, the nutritional information in the first initial distiller's grains formula matrix is ​​streamlined, the amount of information that needs to be processed later is reduced, which helps to improve the efficiency of analysis and evaluation and reduce the interference of redundant data.

[0038] Specifically, based on the same method principle as above for indexing the first initial distiller's grains formula matrix, the nutrient index retrieval layer indexes the basic diet formula matrix to obtain a second nutrient index matrix, which is a numerical matrix of indicators of the nutrients that contribute mainly to the current pregnancy of the target animal from the basic diet formula. It should be understood that for the sake of brevity of the specification, no further explanation will be given here.

[0039] Specifically, the nutritional value assessment layer receives the first nutrient index matrix through the distiller's grains formula input channel, and integrates it with the second nutrient index matrix from the diet formula input channel, such as weighted fusion of the first nutrient index matrix and the second nutrient index matrix according to the mixing ratio of distiller's grains and diet in actual feeding. Next, the nutritional value assessment layer analyzes the matching degree between the fused nutrient index matrix and the current application requirements of the target animal during pregnancy, such as whether protein, energy, cellulose, etc. meet the nutritional requirements of the target animal, and calculates the relative contribution of each nutrient component. Comprehensive scoring is performed through weighted calculation or other models (such as regression analysis, machine learning model). Finally, the first nutritional value score value under this combination is generated.

[0040] Optionally, the first nutritional value score is added to the nutritional value score set, and then, the second nutrient index matrix is ​​maintained unchanged, and the initial distiller's grains formula matrix set is returned, the first initial distiller's grains formula matrix is ​​updated and extracted, and the above-mentioned nutritional value score results are repeated until the initial distiller's grains formula matrix set completes the nutritional value score, then the second nutrient index matrix is ​​updated, and the traversal extraction of the initial distiller's grains formula matrix set is repeated until the nutritional value score added to the nutritional value score set includes all the second nutrient index matrices and the first initial distiller's grains formula matrix combination. The nutritional value score set is used to record the score obtained by each initial distiller's grains formula after being combined with the basic diet formula. It is helpful to screen out the best distiller's grains formula for application in feed production.

[0041] The above method steps can screen out the best combination from multiple formulas by traversing multiple initial distiller's grains formulas and conducting comprehensive evaluations with the basic diet formulas, ensuring that the nutritional components of the distiller's grains formula meet the needs of the target animals. By analyzing the nutrient matching degree between the distiller's grains formula and the diet formula, it is ensured that the generated distiller's grains formula is not only nutritionally comprehensive, but also accurately matches the specific physiological stage needs of the animal (such as pregnancy). Through automated analysis and scoring calculation, the efficiency of distiller's grains formula generation is improved, and the trial and error cost of manual formula adjustment is reduced.

[0042] Furthermore, the first initial distiller's grains formula matrix is ​​indexed by the nutrient index retrieval layer to obtain a first nutrient index matrix, including:

[0043] The first initial distiller's grains formula matrix includes a distiller's grains formula component type list and a distiller's grains component type weight list; the distiller's grains formula component type list is matched through a weight tolerance threshold library pre-defined by a user of the server to obtain a weight tolerance threshold list, the distiller's grains component type weight list is calibrated to obtain a distiller's grains component type weight interval list; a first-level sample index is performed based on the distiller's grains component type weight interval list to obtain a first-level index sample set, wherein when the sample distiller's grains formula component type is the same as the distiller's grains component type and falls into the distiller's grains component type weight interval list, it is selected as the first-level index sample set. Introduce samples; based on the weight tolerance threshold list, traverse and calibrate the first-level index sample set to obtain a first-level distiller's grains component type weight interval list set; perform sample indexing based on the first-level distiller's grains component type weight interval list set until a k-level index sample set is obtained, wherein 10≥k≥2, and k is an integer; take the first initial distiller's grains formula matrix as the root node, the first-level index sample set as the first-level leaf node until the k-level index sample set is the k-level leaf node, and construct a k-level index sample tree; recursively fit the k-level index sample tree to obtain the first nutrient indicator matrix.

[0044] Specifically, the distiller's grains formula ingredient type list contains various ingredients used in distiller's grains formula, such as sorghum, wheat, corn, glutinous rice, barley and other grains. The distiller's grains ingredient type weight parts list corresponds to the specific weight parts of each ingredient type, indicating the proportion of each ingredient in the formula, such as 100g, 500g, etc.

[0045] Specifically, the fault tolerance threshold library is pre-defined by the user of the server and is used to set the allowable weight error range of each ingredient in the formula. For example, for a certain distiller's grains ingredient, the allowable weight error range is ±10g. The distiller's grains formula ingredient type list is matched with the ingredient type in the fault tolerance threshold library to find the weight tolerance range corresponding to each ingredient type, thereby generating a weight tolerance threshold list, which lists the weight parts allowed for each ingredient within the fault tolerance threshold range.

[0046] Specifically, based on the weight tolerance threshold list, the weight of the distiller's grains component type is calibrated to generate a weight interval, which is used to characterize the weight value interval corresponding to the allowable error range of each component. Then, the weight intervals of all components are integrated into a list to obtain a list of weight intervals of the distiller's grains component type representing the allowable weight range of each component. For example, if the weight of a certain component is 100g and the tolerance range is ±10g, the weight interval is [90g, 110g].

[0047] Specifically, first, search the sample library for samples whose distiller's grains formula ingredient type is the same as the ingredient type in the initial distiller's grains formula. Then, for samples that meet the ingredient type requirements, further screen out samples whose ingredient weights fall into the weight interval list. These samples are selected as first-level index samples. After obtaining the first-level index sample set, further screen them according to the weight tolerance threshold list to ensure that the weight of the samples meets the preset tolerance range. Finally, traverse the first-level index sample set, integrate the ingredient weight intervals of each sample into a new interval set for subsequent indexing, and the obtained first-level distiller's grains ingredient type weight interval list set represents the set of intervals to which the actual samples belong.

[0048] Further, multi-level indexing is performed according to the process of first-level sample indexing until a k-level index sample set is generated. The k-level index corresponds to the k-level interval size, that is, the samples are graded according to the interval length, and each level of indexing is screened and iterated according to the results of the previous level. Preferably, the value of k is an integer between 2 and 10, representing the number of index iterations.

[0049] Specifically, a k-level index sample tree is constructed, in which the first initial distiller's grains recipe matrix is ​​the root node for indexing other samples. The first-level index sample set under the K-level index is used as the first-level leaf node, representing the first-level sample that matches the initial recipe ingredients. Correspondingly, the sample index results of each layer in the k-level index sample set are used as corresponding leaf nodes until a k-level sample tree is generated.

[0050] The above method steps, through multi-level sample indexing and recursive fitting, can extract nutrient indicators that meet the requirements from historical recipe samples to ensure the accuracy of the lees recipe. The k-level index sample tree obtained based on the multi-level index can effectively divide the samples, which helps the final generated nutrient indicator matrix to meet the nutritional needs of specific animals.

[0051] Further, recursively fitting the k-level index sample tree to obtain the first nutrient index matrix includes:

[0052] Construct a recursive fitting rule:

[0053] According to the k-level index sample tree, an i+1-level subsample set of the i-level first sample of the i-level index sample set is obtained; the i+1-level subsample set is clustered by nutrient indicator type to obtain a j-th attribute nutrient indicator measurement value set; when the quantity ratio of the j-th attribute nutrient indicator measurement value set to the i+1-level subsample set is less than or equal to a quantity ratio threshold, the j-th attribute nutrient indicator is deleted; when the quantity ratio of the j-th attribute nutrient indicator measurement value set to the i+1-level subsample set is greater than the quantity ratio threshold, the j-th attribute nutrient indicator measurement value set is deleted. The method comprises the following steps: performing a mode analysis on the value set to obtain the characteristic value of the j-th attribute nutrient indicator; storing the characteristic value of the j-th attribute nutrient indicator into the i-th level first sample node in the k-level index sample tree, deleting the i+1-level sub-sample set nodes, and continuing to perform recursive fitting; based on the recursive fitting rule, aggregating the k-level index sample tree from the k-level leaf node to the root node to obtain a root node nutrient indicator type list and a root node nutrient indicator content list; constructing the first nutrient indicator matrix according to the root node nutrient indicator type list and the root node nutrient indicator content list.

[0054] Specifically, first, select the i-level index sample set from the k-level index sample tree, extract the first sample in the i-level index sample set, and further obtain the i+1-level sub-sample set of the sample. Then, perform nutrient index type clustering on all samples in the i+1-level sub-sample set. The clustering algorithm may include K-means, hierarchical clustering, etc. The multiple cluster clusters of the clustering results correspond to multi-attribute sample indicators. Next, randomly extract from the clustering results to obtain the j-th attribute nutrient index measurement value set, which contains all the indicator values ​​of the j-th attribute nutrient index in the sample, and the measurement values ​​in the j-th attribute nutrient index measurement value set are repeatable.

[0055] Furthermore, the data volume of the j-th attribute nutrient index measurement value set is judged, including calculating the quantity ratio of the j-th attribute nutrient index measurement value set to the i+1 level subsample set, and comparing the quantity ratio with a preset quantity ratio threshold.

[0056] Specifically, when the quantity ratio is less than or equal to the quantity ratio threshold, it means that the j-th attribute nutrient index is an indicator that appears unstably or has too little data, and the value of typical value analysis is low, so the j-th attribute nutrient index is deleted. When the quantity ratio is greater than the quantity ratio threshold, it means that the j-th attribute nutrient index appears frequently and the data volume is large, and the confidence of the result of typical value analysis is high, so the mode analysis of the j-th attribute nutrient index measurement value set is further performed to obtain the characteristic value of the j-th attribute nutrient index.

[0057] Specifically, after obtaining the characteristic value of the nutrient index of the jth attribute, in the k-level index sample tree, the characteristic value of the nutrient index of the jth attribute is stored in the first sample node of level i, and the i+1-level sub-sample set node is deleted. Then, repeat the above steps to recursively process the next sample in the i-level sample set until all samples are processed. Then, starting from the k-level leaf node, aggregate the nutrient indexes upward step by step, that is, at each level of node, merge the nutrient index characteristic values ​​of its child nodes, and summarize all nutrient index types and their corresponding contents at the root node to obtain the root node nutrient index type list and content list. Finally, according to the root node nutrient index type list and the root node nutrient index content list, construct the first nutrient index matrix, wherein the rows of the matrix represent different nutrient index types, and the columns represent the nutrient index content of each sample.

[0058] Through the above method steps, according to the k-level index relationship of the k-level index sample tree, the sample tree is gradually extracted and merged from bottom to top, ensuring that each level of node only retains the most representative nutrient index characteristics, thereby effectively reducing redundant data, retaining important nutrient index characteristics, and ensuring that the eigenvalue of the final root node can fully represent the nutrient index situation of the entire sample tree. At the same time, the complex tree structure is automatically processed through a recursive algorithm, simplifying the process of eigenvalue extraction and aggregation, and efficiently processing large-scale and complex data dimensional sample data, reducing manual intervention and computational complexity.

[0059] Further, the first nutrient index matrix is ​​received through the distiller's grains formula input channel of the nutritional value assessment layer, and the second nutrient index matrix is ​​received through the diet formula input channel of the nutritional value assessment layer to obtain a first nutritional value score value, which includes:

[0060] With the preset pregnancy period identifier of the preset animal as a restriction, a diet formula experimental data set, a distiller's grains formula experimental data set and a nutritional value index score value data set are collected; the diet formula experimental data set is indexed through the nutrient index retrieval layer to obtain a first nutrient index experimental matrix set; the distiller's grains formula experimental data set is indexed through the nutrient index retrieval layer to obtain a second nutrient index experimental matrix set; the nutritional value index score value data set is used as the supervised truth value of the BP neural network architecture, the first nutrient index experimental matrix set is used as the input data of the diet formula input channel of the BP neural network architecture, and the second nutrient index experimental matrix set is used as the input data of the distiller's grains formula input channel of the BP neural network architecture to train the nutritional value assessment layer.

[0061] Preferably, before the evaluation is performed through the nutritional value evaluation layer, the nutritional value evaluation layer is constructed based on the BP neural network. First, an experimental data set under the preset animal pregnancy identification restriction is collected, wherein the preset animal pregnancy identification is used to indicate the specific physiological stage of the animal (such as early pregnancy, mid-pregnancy, and late pregnancy), and the nutritional requirements for feed are different in different pregnancy stages. The obtained diet formula experimental data set contains the basic diet formula designed for animals in a specific pregnancy and its nutritional component analysis results, such as nutrients such as protein, energy, and minerals. The obtained distiller's grains formula experimental data set records the experimental data when the distiller's grains formula is used in feed, including the ratio of various distiller's grains ingredients and their nutrient information. The obtained nutritional value index score value data set contains the nutritional value scores of different diets and distiller's grains formula combinations evaluated based on the actual feeding experiment results, wherein the rating dimensions include digestible energy, metabolizable energy, amino acid digestibility, etc.

[0062] Then, through the nutrient index retrieval layer, relevant nutrient indexes are extracted from the diet formula experimental data set, the distiller's grains formula experimental data set, and the first nutrient index experimental matrix set and the second nutrient index experimental matrix set are generated. Among them, the first nutrient index experimental matrix set contains various nutrient components and their corresponding experimental data under different diet formulas, such as the content of protein, energy, vitamins, minerals, etc. The second nutrient index experimental matrix set represents the nutrient characteristics of the distiller's grains formula in the experiment, such as the content of dry matter, protein, energy, medium detergent fiber, amino acids, etc.

[0063] Specifically, the nutritional value index score data set is used as the supervised truth value of the BP neural network, that is, the target of network learning, and the first nutrient index experimental matrix set is input into the diet formula input channel of the BP neural network architecture to provide the nutrient data of the basic diet as training input. The second nutrient index experimental matrix set is input into the distiller's grains formula input channel of the BP neural network architecture to provide the nutrient data of the distiller's grains components.

[0064] Then, according to the input diet and distiller's grains nutrient index, the BP neural network is activated for forward calculation to generate a predicted nutritional value score, and the calculated score is compared with the actual nutritional value index score value data set to calculate the error between the predicted value and the true value. Then, the network weight is adjusted according to the error to gradually reduce the prediction error, and the above process is repeated for multiple iterative training to learn the mapping relationship between the nutritional value index score value data set and the diet and distiller's grains nutrient index to optimize the weight so that the predicted nutritional value score is closer and closer to the actual score value.

[0065] The above method steps train the nutritional value assessment layer through the BP neural network architecture according to the pregnancy mark of the target animal, so as to understand the nutritional needs of different stages of learning. The generated nutritional value assessment layer helps to improve the efficiency and accuracy of the evaluation of the distiller's grains formula and reduce the cost of optimization trial and error.

[0066] According to the nutritional value score set, in combination with the nutritional value parser and the basic diet formula matrix, the initial distiller's grains formula matrix set is optimized to obtain a recommended distiller's grains formula matrix sequence;

[0067] Specifically, the nutritional value score value set is the score data representing the nutritional value of each initial distiller's grains formula combined with the basic diet formula. It contains multiple indicators, such as digestible energy, metabolizable energy, amino acid digestibility, etc. The distiller's grains formula matrix sequence is a collection of multiple recommended distiller's grains formulas, which have been optimized by the nutritional value parser and the basic diet formula, and are sorted according to the optimized nutritional value score values. It is used to provide multiple recommended distiller's grains formulas for users to choose from.

[0068] According to the nutritional value score set, in combination with the nutritional value parser and the basic diet formula matrix, the initial distiller's grains formula matrix set is optimized to obtain a recommended distiller's grains formula matrix sequence, including:

[0069] According to the nutritional value score value set, a first nutritional value score value is extracted, wherein the first nutritional value score value includes a plurality of nutritional value index score values; according to the pregnancy period mark of the target animal, a plurality of nutritional value index weights are configured for the plurality of nutritional value indexes; the plurality of nutritional value index score values ​​are forward normalized and preprocessed to obtain a plurality of nutritional value index score value characteristic parameters; according to the plurality of nutritional value index weights and the plurality of nutritional value index score value characteristic parameters, a first fitness is obtained, and the fitness is added to the initial distiller's grains formula matrix fitness set; according to the initial distiller's grains formula matrix fitness set, the initial distiller's grains formula is selected from the initial distiller's grains formula matrix from large to small. A first number of initial distiller's grains formula matrices are extracted from a matrix set and set as a target distiller's grains formula matrix; a first number of initial distiller's grains formula matrices are extracted from the initial distiller's grains formula matrix set from small to large according to the fitness set of the initial distiller's grains formula matrices and set as an optimized distiller's grains formula matrix; the formulation of the optimized distiller's grains formula matrix is ​​adjusted close to the target distiller's grains formula matrix to obtain a new distiller's grains formula matrix, wherein the close adjustment refers to the close adjustment or exchange of the component weight portions of any several dimensions of the optimized distiller's grains formula matrix to the component weight portions of several dimensions of the target distiller's grains formula matrix; a loop is iterated a preset number of times, and a third number of recommended distiller's grains formula matrix sequences with larger fitness are output.

[0070] Specifically, the scores of multiple types of nutritional value indicators involved in the nutritional value score set are comprehensively evaluated according to the pregnancy mark of the target animal. Exemplarily, the nutritional value score set includes digestible energy score, metabolizable energy score, protein score, amino acid digestibility score, etc.

[0071] Optionally, animals at different stages of pregnancy have different nutritional requirements. For example, early pregnancy requires more energy and protein, while late pregnancy may require higher mineral content. According to the pregnancy mark of the target animal, a weight is assigned to each nutritional value indicator, and fitness is calculated through a weighted evaluation method.

[0072] Optionally, before performing fitness calculation, the nutritional value index score values ​​are subjected to forward normalization preprocessing to standardize the score values ​​of different indicators into the same dimension so that they can be weighted calculated.

[0073] Furthermore, according to the configured weights, the weighted summation of the characteristic parameters of the normalized nutritional value index score value is performed to calculate the fitness, and then the fitness value of each distiller's grains formula is recorded in the initial distiller's grains formula matrix fitness set as the basis for subsequent optimization and screening. Next, the initial distiller's grains formula matrix fitness set is sorted in descending order according to the fitness value, and the first few initial distiller's grains formulas with larger fitness are extracted and set as the target distiller's grains formula matrix. The distiller's grains formulas contained in the target distiller's grains formula matrix have the highest scores and are close to the optimization target. At the same time, the initial distiller's grains formula matrix fitness set is sorted in ascending order according to the fitness value, and the initial distiller's grains formulas with lower fitness are selected as the formulas that need to be optimized and set as the optimized distiller's grains formula matrix.

[0074] Furthermore, the formula in the optimized distiller's grains formula matrix is ​​moved closer to the target distiller's grains formula in terms of the weight of certain ingredients, gradually narrowing the gap with the high-fitness formula, and adjusting the optimized distiller's grains formula matrix. For example, if the weight of a certain ingredient in a target formula is 50g, and the optimized formula is 40g, the weight can be adjusted to 45g. Optionally, some ingredients can be exchanged or replaced to improve the randomness and diversity of the optimized nutrients and avoid falling into a local optimum.

[0075] Specifically, the above-mentioned close adjustment process is repeated, and multiple iterations are performed to optimize, and the wine lees formula is gradually improved to gradually improve its fitness. After each adjustment, the fitness of the optimized treatment formula is recalculated, and the target wine lees formula matrix and the optimized treatment wine lees formula matrix are updated. After the iteration is completed, the top third number of wine lees formulas with the largest fitness are selected as the recommended wine lees formula matrix.

[0076] The above-mentioned method steps improve the diversity of distiller's grains formula through multiple rounds of iterative optimization, thereby generating multiple distiller's grains formulas with higher adaptability. The entire process realizes the automation of formula optimization, reduces human intervention, improves the efficiency and accuracy of distiller's grains formula design, and ensures that it can maximize the nutritional needs of animals in practical applications.

[0077] Furthermore, according to the pregnancy period mark of the target animal, several nutritional value index weights are configured for several nutritional value indexes, including:

[0078] According to the pregnancy mark of the target animal, a target feeding status indicator set is matched through an animal feeding status indicator library pre-defined by a user of the server; based on the first target feeding status indicator of the target feeding status indicator set, a grey correlation analysis is performed on the several nutritional value indicators to obtain a first nutritional value correlation set; until the Qth target feeding status indicator based on the target feeding status indicator set, a grey correlation analysis is performed on the several nutritional value indicators to obtain a Qth nutritional value correlation set; the first nutritional value correlation set is added up to the Qth nutritional value correlation set to obtain a comprehensive correlation sum value; the first nutritional value indicator correlation set of the several nutritional value indicators is added to obtain a first nutritional value indicator correlation sum value; the ratio of the first nutritional value indicator correlation sum value to the comprehensive correlation sum value is calculated, set as a first nutritional value indicator weight, and added to the several nutritional value indicator weights.

[0079] Specifically, the animal feeding status index library pre-defined by the user contains a variety of indicators related to different pregnancy stages or feeding status of animals. Users can define feeding status indicators according to feeding goals (such as improving production performance or health status). The target feeding status indicator set matched from the indicator library according to the pregnancy identification of the target animal is used to reflect the physiological needs of the animal in the current pregnancy stage, exemplarily including energy needs, protein needs, mineral needs, vitamin needs, fiber needs, amino acid needs, etc.

[0080] Specifically, a first feeding status indicator (such as energy requirement) is selected from the target feeding status indicator set, and a correlation analysis is performed on the nutritional value indicator to generate a first nutritional value correlation set, which is used to represent the correlation between each nutritional value indicator (such as protein, energy, amino acid digestibility, etc.) and the feeding status indicator.

[0081] Furthermore, grey correlation analysis is performed on each target feeding status indicator in turn until the Qth feeding status indicator is reached, and Q nutritional value correlation degree sets are obtained.

[0082] Specifically, the association degree sets of all target feeding status indicators from the first to the Qth are accumulated to calculate the comprehensive sum of the association degrees. The comprehensive sum of the association degrees represents the overall association degree of all nutritional value indicators under the comprehensive influence of multiple feeding status indicators. At the same time, for each specific nutritional value indicator, its association degree in each feeding status indicator analysis is summed up, such as for the first nutritional value indicator association degree set, the sum of the first nutritional value indicator association degree is obtained. Furthermore, the ratio of the first nutritional value indicator association degree sum value to the comprehensive sum of the association degree is calculated to obtain the first nutritional value indicator weight.

[0083] Furthermore, for each nutritional value index, the above steps are repeated until the weights of all indicators are calculated and added, and the output is a set of nutritional value index weights. The above method steps, through grey correlation analysis, achieve a scientific evaluation of the impact of each nutritional value index on the current pregnancy of the target animal, and accurately configure the weight of each indicator, which helps to realize the automation of distiller's grains formula design, reduce manual intervention, and improve the efficiency and accuracy of feed design.

[0084] The recommended distiller's grains formula matrix sequence is sent to the visual interface of the user end for display feedback.

[0085] Specifically, the recommended distiller's grains recipe matrix sequence is returned to the user end, and visual feedback is provided to provide the user with multiple distiller's grains recipes with sequence relationships, so that the user can make adaptive choices based on their actual needs, thereby improving the user experience.

[0086] The method for optimizing distiller's grains formula data provided by the embodiment of the present invention has at least the following technical effects:

[0087] By responding to the distiller's grains formula request generated by the user end, which includes the pregnancy identification of the target animal, the basic diet formula matrix, and the initial distiller's grains formula matrix set; matching the appropriate nutritional value parser from the server's model library according to the pregnancy identification of the target animal; using the nutritional value parser, based on the basic diet formula matrix, traversing and parsing the initial distiller's grains formula matrix set to generate a nutritional value score value set; optimizing the initial distiller's grains formula matrix set according to the nutritional value score value set, combining the nutritional value parser and the basic diet formula matrix, and generating a recommended distiller's grains formula matrix sequence; displaying and feedback the recommended distiller's grains formula matrix sequence through the user end's visual interface. Thus, the technical effect of reducing the difficulty of formula optimization and improving the efficiency and adaptability of formula development is achieved.

[0088] Embodiment 2:

[0089] See also Figure 3 , Figure 3 Schematic diagram of an electronic device provided by an embodiment of the present invention. Figure 3As shown, an embodiment of the present invention provides an electronic device 500, including a memory 510, a processor 520, and a computer program 511 stored in the memory 510 and executable on the processor 520, wherein the processor 520 implements the following steps when executing the computer program 511: responding to a distiller's grains formula request generated by a user terminal, wherein the distiller's grains formula request includes a target animal pregnancy identifier, a basic diet formula matrix, and an initial distiller's grains formula matrix set; matching a nutritional value parser from a model library of a server according to the target animal pregnancy identifier; traversing the initial distiller's grains formula matrix set for parsing based on the basic diet formula matrix through the nutritional value parser to obtain a nutritional value score value set; optimizing the initial distiller's grains formula matrix set according to the nutritional value score value set, in combination with the nutritional value parser and the basic diet formula matrix, to obtain a recommended distiller's grains formula matrix sequence; and sending the recommended distiller's grains formula matrix sequence to a visualization interface of the user terminal for display feedback.

[0090] Embodiment three:

[0091] See also Figure 4 , Figure 4 A schematic diagram of an embodiment of a computer-readable storage medium provided in an embodiment of the present invention. Figure 4 As shown, this embodiment provides a computer-readable storage medium 600, on which a computer program 611 is stored. When the computer program 611 is executed by a processor, the following steps are implemented: responding to a distiller's grains formula request generated by a user end, wherein the distiller's grains formula request includes a target animal pregnancy identification, a basic diet formula matrix and an initial distiller's grains formula matrix set; matching a nutritional value parser from a model library of a server according to the target animal pregnancy identification; using the nutritional value parser, based on the basic diet formula matrix, traversing the initial distiller's grains formula matrix set for parsing to obtain a nutritional value score value set; optimizing the initial distiller's grains formula matrix set according to the nutritional value score value set, in combination with the nutritional value parser and the basic diet formula matrix, to obtain a recommended distiller's grains formula matrix sequence; and sending the recommended distiller's grains formula matrix sequence to a visualization interface of the user end for display feedback.

[0092] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0093] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0094] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0095] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0096] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0097] Although preferred embodiments of the present invention have been described, additional changes and modifications may occur to these embodiments once those skilled in the art understand the basic inventive concepts.

[0098] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention belong to the scope of the present invention and its equivalent technologies, the present invention is also intended to include these changes and variations.

Claims

1. A method for optimizing distiller's grains recipe data, characterized in that: Application servers, including: Responding to a distiller's grains formula request generated by a user end, wherein the distiller's grains formula request includes a target animal pregnancy stage identifier, a basic diet formula matrix, and an initial distiller's grains formula matrix set; According to the pregnancy period identifier of the target animal, a nutritional value parser is matched from the model library of the server; By means of the nutritional value parser, based on the basic diet formula matrix, the initial distiller's grains formula matrix set is traversed for parsing to obtain a nutritional value score value set; According to the nutritional value score set, in combination with the nutritional value parser and the basic diet formula matrix, the initial distiller's grains formula matrix set is optimized to obtain a recommended distiller's grains formula matrix sequence; Sending the recommended distiller's grains formula matrix sequence to the visual interface of the user terminal for display feedback; According to the nutritional value score set, in combination with the nutritional value parser and the basic diet formula matrix, the initial distiller's grains formula matrix set is optimized to obtain a recommended distiller's grains formula matrix sequence, including: Extracting a first nutritional value score value according to the nutritional value score value set, wherein the first nutritional value score value includes a plurality of nutritional value index score values; According to the pregnancy period mark of the target animal, a plurality of nutritional value index weights are configured for the plurality of nutritional value indexes; Performing forward normalization preprocessing on the plurality of nutritional value index score values ​​to obtain characteristic parameters of the plurality of nutritional value index score values; According to the weights of the plurality of nutritional value indicators and the characteristic parameters of the plurality of nutritional value indicator scores, a first fitness is obtained, and added into the initial distiller's grains formula matrix fitness set; According to the initial distiller's grains formula matrix fitness set, extracting a first number of initial distiller's grains formula matrices from the initial distiller's grains formula matrix set from large to small, and setting them as target distiller's grains formula matrices; According to the initial distiller's grains formula matrix fitness set, extracting a first number of initial distiller's grains formula matrices from the initial distiller's grains formula matrix set from small to large, and setting them as optimized distiller's grains formula matrices; The formulation of the optimized lees formula matrix is ​​adjusted toward the target lees formula matrix to obtain a new lees formula matrix, wherein the adjustment refers to adjusting or exchanging the weight portions of components in any number of dimensions of the optimized lees formula matrix toward the weight portions of components in any number of dimensions of the target lees formula matrix; Iterate the loop a preset number of times, and output the recommended distiller's grains formula matrix sequence with a third quantity of greater fitness; Among them, according to the pregnancy period mark of the target animal, several nutritional value index weights are configured for several nutritional value indexes, including: According to the pregnancy period identifier of the target animal, a target feeding status indicator set is matched through an animal feeding status indicator library pre-defined by a user of the server; Based on a first target feeding status indicator of the target feeding status indicator set, performing a grey correlation analysis on the plurality of nutritional value indicators to obtain a first nutritional value correlation set; Until a Qth target feeding state indicator of the target feeding state indicator set is used to perform a grey correlation analysis on the plurality of nutritional value indicators to obtain a Qth nutritional value correlation set; Adding the first nutritional value association degree set up to the Qth nutritional value association degree set to obtain a comprehensive sum of association degrees; Adding the first nutritional value indicator correlation degree sets of the plurality of nutritional value indicators to obtain a first nutritional value indicator correlation degree sum value; The ratio of the sum of the correlation degrees of the first nutritional value indicators to the comprehensive sum of the correlation degrees is calculated and set as the first nutritional value indicator weight, which is added to the weights of the plurality of nutritional value indicators.

2. The method according to claim 1, characterized in that The nutritional value parser is used to traverse the initial distiller's grains formula matrix set for parsing based on the basic diet formula matrix to obtain a nutritional value score value set, including: The nutritional value analyzer includes a nutrient index retrieval layer and a nutritional value assessment layer; Extracting a first initial distiller's grains formula matrix from the initial distiller's grains formula matrix set; The first initial distiller's grains formula matrix is ​​indexed by the nutrient index retrieval layer to obtain a first nutrient index matrix; At the same time, the basic diet formula matrix is ​​indexed through the nutrient index retrieval layer to obtain a second nutrient index matrix; The first nutrient index matrix is ​​received through the distiller's grains formula input channel of the nutritional value assessment layer, and the second nutrient index matrix is ​​received through the diet formula input channel of the nutritional value assessment layer to obtain a first nutritional value score value; The first nutritional value score value is added to the nutritional value score value set.

3. The method according to claim 2, characterized in that The first initial distiller's grains formula matrix is ​​indexed by the nutrient index retrieval layer to obtain a first nutrient index matrix, including: The first initial distiller's grains formula matrix includes a distiller's grains formula component type list and a distiller's grains component type weight portion list; Matching the distiller's grains formula component type list through a weight tolerance threshold library pre-defined by a user of the server to obtain a weight tolerance threshold list, calibrating the distiller's grains component type weight list to obtain a distiller's grains component type weight range list; Performing primary sample indexing based on the lees component type weight interval list to obtain a primary index sample set, wherein when the sample lees formula component type is the same as the lees component type and falls into the lees component type weight interval list, it is selected as the primary index sample; Based on the weight tolerance threshold list, traverse and calibrate the first-level index sample set to obtain a first-level distiller's grains component type weight interval list set; Perform sample indexing based on the first-level distiller's grains component type weight interval list set until a k-level index sample set is obtained, where 10≥k≥2, and k is an integer; Taking the first initial distiller's grains recipe matrix as a root node, taking the first-level index sample set as a first-level leaf node, and then taking the k-level index sample set as a k-level leaf node, constructing a k-level index sample tree; Recursively fit the k-level index sample tree to obtain the first nutrient indicator matrix.

4. The method according to claim 3, characterized in that Recursively fitting the k-level index sample tree to obtain the first nutrient indicator matrix includes: Construct a recursive fitting rule: According to the k-level index sample tree, obtain an i+1-level subsample set of the i-level first sample of the i-level index sample set; Clustering the nutrient index type of the i+1 level sub-sample set to obtain a j-th attribute nutrient index measurement value set; When the quantity ratio of the j-th attribute nutrient index measurement value set to the i+1-level sub-sample set is less than or equal to the quantity ratio threshold, the j-th attribute nutrient index is deleted; When the quantity ratio of the j-th attribute nutrient index measurement value set to the i+1-level subsample set is greater than the quantity ratio threshold, a mode analysis is performed on the j-th attribute nutrient index measurement value set to obtain the j-th attribute nutrient index characteristic value; In the k-level index sample tree, the characteristic value of the j-th attribute nutrient index is stored in the i-level first sample node, and after deleting the i+1-level sub-sample set node, the recursive fitting is continued; Based on the recursive fitting rule, the k-level index sample tree is aggregated from the k-level leaf node to the root node to obtain a root node nutrient indicator type list and a root node nutrient indicator content list; The first nutrient indicator matrix is ​​constructed according to the root node nutrient indicator type list and the root node nutrient indicator content list.

5. The method according to claim 2, characterized in that The first nutrient index matrix is ​​received through the distiller's grains formula input channel of the nutritional value assessment layer, and the second nutrient index matrix is ​​received through the diet formula input channel of the nutritional value assessment layer to obtain a first nutritional value score value, which includes: With the preset animal and preset pregnancy period identification as the restriction, the diet formula experimental data set, distiller's grains formula experimental data set and nutritional value index score value data set were collected; Indexing the diet formula experimental data set through the nutrient index retrieval layer to obtain a first nutrient index experimental matrix set; The distiller's grains formula experimental data set is indexed through the nutrient index retrieval layer to obtain a second nutrient index experimental matrix set; The nutritional value evaluation layer is trained by using the nutritional value index score data set as the supervised truth value of the BP neural network architecture, using the first nutrient index experimental matrix set as the input data of the diet formula input channel of the BP neural network architecture, and using the second nutrient index experimental matrix set as the input data of the distiller's grains formula input channel of the BP neural network architecture.

6. An electronic device, characterized in that: include: Memory for storing computer software programs; A processor is used to read and execute the computer software program, thereby implementing the optimization processing method of the wine lees formula data described in any one of claims 1-5.

7. A non-transitory computer-readable storage medium, characterized in that: The storage medium stores a computer software program, which, when executed by a processor, implements the method for optimizing the processing of distiller's grains formula data as described in any one of claims 1 to 5.

Images