Method and system for predicting amino acid utilization rate of multi-stage feed for laying hens

Abstract

The application provides a laying hen multi-stage feed amino acid utilization rate prediction method and system, relates to the field of bioinformatics, and comprises the following steps: obtaining chemical component determination values and standard ileal amino acid digestibility in vivo determination values of feed raw material samples in different growth stages of laying hens; performing prediction factor screening according to the chemical component determination values and the standard ileal amino acid digestibility in vivo determination values to obtain a prediction factor combination; performing prediction model construction according to the prediction factor combination to obtain a candidate prediction equation; performing optimization according to the candidate prediction equation to screen an optimized equation; performing model verification according to the optimized equation to obtain a target prediction model; and performing digestibility prediction according to the target prediction model to obtain a predicted standard ileal amino acid digestibility value. The application realizes accurate prediction of the standard ileal amino acid digestibility based on in-vitro detection data, and effectively overcomes the limitations of traditional in-vivo determination methods.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of bioinformatics, and specifically relates to a method and system for predicting the amino acid utilization rate of multi-stage feed for laying hens. BACKGROUND

[0002] In the field of precise feeding of laying hens, accurate evaluation of the amino acid utilization rate of feed is the key to achieving nutritional optimization and reducing feeding costs, and the standard ileal amino acid digestibility is the core indicator for evaluating the biological efficacy of amino acids. At present, the acquisition of this indicator mainly relies on traditional in vivo determination method. Although the determination result of this method is accurate, it has problems such as long test cycle, high cost, and great impact on experimental animal welfare, which makes it difficult to meet the demand for rapid evaluation of a large number of feed raw materials in actual production. Although the in vitro prediction method developed in recent years has improved the detection efficiency to some extent, the model is mostly based on static nutrient component data, and the dynamic differences in metabolic characteristics of laying hens at different physiological stages and the complex effects of intestinal microbial community and endogenous amino acid loss on the digestion process are not fully considered, which limits the universality and accuracy of the prediction model. Especially when facing different egg production cycles and different raw material combinations, the prediction deviation is large, which makes it difficult to support the actual application of precise feed formulation.

[0003] Based on the above-mentioned shortcomings of the prior art, there is an urgent need for a method and system for predicting the amino acid utilization rate of multi-stage feed for laying hens. SUMMARY

[0004] The purpose of the present application is to provide a method and system for predicting the amino acid utilization rate of multi-stage feed for laying hens to improve the above-mentioned problems. In order to achieve the above-mentioned purpose, the technical scheme adopted by the present application is as follows:

[0005] In a first aspect, the present application provides a method and system for predicting the amino acid utilization rate of multi-stage feed for laying hens, comprising:

[0006] obtaining the chemical component determination value and the in vivo determination value of the standard ileal amino acid digestibility of the feed raw material sample of the laying hen at different growth stages;

[0007] According to the chemical component determination value and the in vivo determination value of the standard ileal amino acid digestibility, a prediction factor screening is performed, the correlation between the nutrient components of the feed raw material and the metabolic characteristics of the laying hen at different egg production cycles is dimensionally mined through Akaike information criterion analysis, and a prediction factor combination is obtained;

[0008] According to the prediction factor combination, a prediction model is constructed, and a multiple regression relationship between the amino acid flow characteristics of the laying hen intestinal chyme and the prediction factor combination is established to obtain a candidate prediction equation;

[0009] According to the candidate prediction equation, optimization is performed, the determination coefficients and biological rationalities of each equation at different growth stages of the laying hens are compared, and an optimized equation is screened out;

[0010] According to the optimized equation, model verification is performed, the relative deviation of the predicted value and the in-vivo measured value is calculated, a verification mechanism is constructed in combination with the endogenous amino acid loss correction of the laying hens, and a target prediction model is obtained;

[0011] According to the target prediction model, digestibility prediction is performed, the predicted factor content of the laying hen feed raw material to be measured is input into the target prediction model, and the predicted standard ileal amino acid digestibility value of the current feed formula is obtained.

[0012] In a second aspect, the application further provides a method and system for predicting the amino acid utilization rate of multi-stage feed of laying hens, comprising:

[0013] An acquisition module is configured to acquire chemical component measured values and standard ileal amino acid digestibility in-vivo measured values of feed raw material samples at different growth stages of the laying hens;

[0014] A screening module is configured to perform prediction factor screening according to the chemical component measured values and the standard ileal amino acid digestibility in-vivo measured values, to perform dimensionality reduction mining on the correlation between the nutritional components of the feed raw material and the metabolic characteristics of the laying hens in the egg-laying cycle by Akaike information criterion analysis, and to obtain a prediction factor combination;

[0015] A construction module is configured to perform prediction model construction according to the prediction factor combination, to obtain a candidate prediction equation by establishing a multiple regression relationship between the ileal chyme amino acid flow characteristics of the laying hens and the prediction factor combination;

[0016] An optimization module is configured to perform optimization according to the candidate prediction equation, to compare the determination coefficients and biological rationalities of each equation at different growth stages of the laying hens, and to screen out an optimized equation;

[0017] A verification module is configured to perform model verification according to the optimized equation, to calculate the relative deviation of the predicted value and the in-vivo measured value, to construct a verification mechanism in combination with the endogenous amino acid loss correction of the laying hens, and to obtain a target prediction model;

[0018] A prediction module is configured to perform digestibility prediction according to the target prediction model, to input the prediction factor content of the laying hen feed raw material to be measured into the target prediction model, and to obtain the predicted standard ileal amino acid digestibility value of the current feed formula.

[0019] The application has the following beneficial effects:

[0020] The application realizes accurate prediction of standard ileal amino acid digestibility based on in-vitro detection data by establishing a dynamic correlation model of metabolism characteristics of the laying hen egg-laying cycle and feed nutritional components, combining a multi-stage screening verification mechanism and an endogenous amino acid loss compensation algorithm, and effectively overcomes the limitations of traditional in-vivo determination methods. BRIEF DESCRIPTION OF DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0022] Figure 1 A flowchart of a laying hen multi-stage feed amino acid utilization rate prediction method described in an embodiment of the present application;

[0023] Figure 2 A structural diagram of a laying hen multi-stage feed amino acid utilization rate prediction system described in an embodiment of the present application;

[0024] Figure 3 A structural diagram of a laying hen multi-stage feed amino acid utilization rate prediction device described in an embodiment of the present application.

[0025] In the figure, 800 is a laying hen multi-stage feed amino acid utilization rate prediction device; 801 is a processor; 802 is a memory; 803 is a multimedia component; 804 is an I / O interface; 805 is a communication component; 901 is an acquisition module; 902 is a screening module; 903 is a construction module; 904 is an optimization module; 905 is a verification module; and 906 is a prediction module. DETAILED DESCRIPTION

[0026] In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the following will combine the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, but not all the embodiments. The components of the embodiments of the present application described and shown in the drawings can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0027] It should be noted that similar reference numerals and letters refer to like items in the accompanying drawings and that, once an item is defined in one drawing, it should not require further defining or explaining in subsequent drawings. Also, in the description of the present application, the terms "first", "second", and the like are used only to distinguish one item from another, and do not imply or suggest relative importance.

[0028] Embodiment 1

[0029] The embodiment provides a method and system for predicting the utilization rate of amino acids in multi-stage feed for laying hens.

[0030] Referring to Figure 1 , the method includes steps S100 to S600.

[0031] Step S100, obtaining chemical composition determination values and standard ileal amino acid digestibility in-vivo determination values of feed raw material samples at different growth stages of laying hens;

[0032] It can be understood that, in the practice of precise feeding of laying hens, although the traditional in-vivo determination method can obtain accurate standard ileal amino acid digestibility data, it has limitations such as long test cycle and high cost. In view of these pain points, step S100 designs a systematic data collection scheme: first, representative laying hen feed raw material samples are selected, covering raw materials of different origins, batches and processing technologies, to ensure the diversity and representativeness of the samples; second, ileal chyme samples are systematically collected at different growth stages of laying hens (such as the brooding period, the growing period, the peak egg production period, the molting period, etc.), and standardized determination procedures are used to obtain chemical composition determination values, including conventional nutritional ingredients (such as crude protein, crude fat, crude fiber, etc.) and specific indicators (such as neutral detergent fiber, acid detergent fiber, mineral content, etc.); at the same time, the standard ileal amino acid digestibility value is determined through a strictly controlled in-vivo test to ensure the accuracy and reliability of the data. This step establishes a complete data set covering the entire growth cycle of laying hens, which not only contains the chemical property data of the raw materials but also obtains the corresponding biological potency data, laying a foundation for subsequent establishment of a precise prediction model.

[0033] Step S200, performing prediction factor screening according to the chemical composition determination values and the standard ileal amino acid digestibility in-vivo determination values, and performing dimensionality reduction mining on the correlation between the nutritional ingredients of the feed raw materials and the metabolic characteristics of the laying hen egg production cycle through Akaike information criterion analysis to obtain a combination of prediction factors;

[0034] It should be noted that the Akaike information criterion is used to perform dimensionality reduction mining on the multi-dimensional data, which can effectively identify key prediction factors closely related to the metabolic characteristics of the laying hen egg production cycle, and through quantitative analysis of the dynamic correlation strength between each nutritional ingredient and the digestibility, the purpose of accurately selecting a combination of core indicators from a large number of potential influencing factors is achieved.

[0035] Step S300, according to the prediction factor combination, a prediction model is constructed by establishing a multiple regression relationship between the ileal digesta amino acid flow characteristics of laying hens and the prediction factor combination, and a candidate prediction equation is obtained;

[0036] It can be understood that, in this step, a multiple regression model is constructed based on the screened prediction factor combination, which particularly considers the unique rules of the ileal digesta amino acid flow of laying hens, and forms a prediction equation framework that can reflect the actual digestive physiological process by establishing a mathematical relationship between the prediction factor and the digestibility.

[0037] Step S400, optimization according to the candidate prediction equation, by comparing the determination coefficients and biological rationality of each equation in different growth stages of laying hens, an optimized equation is screened;

[0038] It should be noted that, in this step, the candidate prediction equation is optimized and screened in multiple dimensions, and the prediction performance difference of each equation in different growth stages is evaluated, which fully considers the influence of the physiological state change of laying hens on the applicability of the model, and ensures that the optimized equation obtained finally has good statistical characteristics and conforms to the biological logic.

[0039] Step S500, model verification according to the optimized equation, by calculating the relative deviation of the predicted value and the in vivo measured value, and combining the endogenous amino acid loss correction to construct a verification mechanism, a target prediction model is obtained;

[0040] It can be understood that, in this step, the verification mechanism combines the prediction deviation analysis with the endogenous amino acid loss correction, and by constructing a multi-level verification system, the reliability and accuracy of the model in actual application are effectively improved, and the defects of the traditional prediction method in considering the complex physiological factors are significantly improved.

[0041] Step S600, according to the target prediction model, the digestibility is predicted, by inputting the prediction factor content of the feed raw material of the laying hen to be tested into the target prediction model, the predicted standard ileal amino acid digestibility value of the current feed formula is obtained.

[0042] It should be noted that, in this step, the target prediction model is applied to the digestibility prediction of the actual feed formula, realizing the transformation from the theoretical model to the practical application, and the design of this step reflects the applicability of the method in the actual production environment, and provides effective technical support for the precise feeding of laying hens.

[0043] Further, step S200 includes steps S210 to S230.

[0044] Step S210, dynamic correlation analysis is performed according to the chemical component determination value and the standard ileum amino acid digestion rate in-vivo determination value, a time series phase synchronization model of the laying rate fluctuation of the laying hen and the metabolic demand of the amino acid is established, and a nutrition factor that responds synchronously with the physiological change in the laying period is identified;

[0045] Step S220, combination optimization processing is performed according to the nutrition factor, a modular grouping based on a clustering coefficient is constructed by analyzing the functional correlation strength between factors, and a minimum factor set with functional synergy in the peak laying period and the molting period of the laying hen is screened out;

[0046] Step S230, biological verification processing is performed according to the minimum factor set, a multi-level interaction network model between the ileum microbial community of the laying hen and the prediction factor is established, and a prediction factor combination representing the amino acid digestion characteristics in the laying stage of the laying hen is obtained.

[0047] Specifically, step S210 first dynamically aligns the laying rate fluctuation curve of the laying hen with the change in the metabolic demand of the amino acid by establishing a time series phase synchronization model. This method can capture the time sequence correlation characteristics between the nutrition factor and the physiological period, and identify the key nutrition factors that fluctuate synchronously with the change in the laying period. On this basis, step S220 further adopts a modular grouping algorithm based on a clustering coefficient, calculates the functional correlation strength between different nutrition factors, classifies the factors with synergistic effect into the same functional module, and uses the minimum spanning tree principle in graph theory to screen out a minimum factor set that can cover the needs of key physiological stages such as the peak laying period and the molting period of the laying hen. Finally, step S230 correlates the minimum factor set screened out with the ileum microbial community structure of the laying hen by constructing a multi-level interaction network model. The network model contains multiple interaction relationships between microbial populations and nutrition factors, and can verify the biological rationality of the prediction factor combination from the system level, so as to ensure that the final prediction factor combination not only has statistical significance, but also truly reflects the specific amino acid digestion and metabolism characteristics in the laying stage of the laying hen.

[0048] Further, step S300 includes step S310 to step S330.

[0049] Step S310, dynamic modeling processing is performed according to the prediction factor combination, a time series phase adjustment of the ileum chyme amino acid flow rate in the laying period of the laying hen is introduced, and a primary regression equation based on time dimension correction is constructed;

[0050] Step S320, structure optimization processing is performed according to the primary regression equation, a dynamic prediction framework based on quantile regression is established by analyzing the distribution characteristics of the amino acid absorption efficiency in different laying stages;

[0051] Step S330, biological constraint processing is performed according to the dynamic prediction framework, and a candidate prediction equation conforming to the physiological characteristics of the laying hen digestion is obtained by integrating the coupling relationship between the microbial metabolic pathway of the laying hen ileum and the amino acid digestion process.

[0052] Specifically, step S310 first aligns the change law of the ileal chyme amino acid flow rate in the laying hen laying cycle with the prediction factor data through a time series phase adjustment technique. This processing method can effectively eliminate the time dimension deviation caused by physiological cycle fluctuations, and the primary regression equation constructed has a preliminary time dynamic characteristic. On this basis, step S320 uses a quantile regression method to analyze the distribution characteristics of the amino acid absorption efficiency at different laying stages. This method can capture the change law at different quantile points of the data distribution, thereby establishing a dynamic prediction framework that can adapt to the characteristics of each physiological stage of the laying hen. Step S330 further integrates the coupling relationship between the microbial metabolic pathway of the laying hen ileum and the amino acid digestion process, and adds biological constraint conditions to the prediction equation by establishing a correlation model between microbial community function and digestion efficiency, so that the candidate prediction equation obtained finally not only has mathematical statistical significance, but also can truly reflect the physiological characteristics of the laying hen digestive system. This series of processing processes embodies the progressive optimization idea from time dimension correction to distribution characteristic analysis, and then to biological mechanism constraint, ensuring that the prediction model not only conforms to the mathematical law, but also is close to the actual physiological process.

[0053] For a preferred embodiment of the present application, for any growth stage of poultry, after determining each group of candidate prediction factors, a primary regression equation of poultry at the growth stage is constructed based on each group of candidate prediction factors, as shown in Table 1.

[0054] Table 1 Primary regression equation of standard ileal lysine digestibility of secondary powder raw material

[0055]

[0056] In the prediction method of the amino acid utilization rate of the multi-stage feed of the laying hen, the derivation of the candidate prediction equation is based on a multiple linear regression model. First, the prediction factors (such as crude ash, acid detergent fiber, etc.) that have a significant influence on the standard ileal amino acid digestibility (IAAD) are selected from the chemical components of the feed raw material by the Akaike information criterion. The selection of these factors takes into account the metabolic characteristics of the laying hen laying cycle, for example, during the laying peak period, the amino acid requirement is higher, so the factors need to reflect the digestion characteristics of energy and protein. SID

[0057] The general multiple linear regression model is as follows:

[0058] ;

[0059] ​In the formula, β0 is the intercept term; β i X is the regression coefficient of the i-th predictor, where i = 1, 2, ..., k; i Represents the measured value of the predictor; This is the error term.

[0060] Specifically, in Table 1, each equation represents the standard ileal leucine digestibility of the secondary wheat flour raw material. SID The predictive model for this study included crude ash (representing the inorganic content of feed, affecting digestibility), acid detergent fiber (representing cellulose content, reducing digestibility), neutral detergent fiber (representing hemicellulose and cellulose, affecting digestion), bulk density (feed density, potentially affecting digestion rate), calcium (a mineral involved in metabolic regulation), albumin (a protein component affecting amino acid supply), and total energy (total energy in feed, reflecting metabolic potential). The coefficients of these factors were obtained by fitting experimental data, where R-squared and p-values ​​were used to evaluate the goodness of fit and significance of the model.

[0061] Next, based on the primary regression equation, the model structure was optimized using quantile regression. Quantile regression is used to capture the distribution characteristics of amino acid uptake efficiency in different laying stages of hens (such as peak laying and molting), rather than just the conditional mean. This allows the model to adapt to different quantiles of the distribution (such as...). =0.25, 0.5, 0.75), thus better reflecting physiological fluctuations. The general form of the quantile regression model is:

[0062] ;

[0063] In the formula, This represents the standard ileal amino acid digestibility under a given predictor factor X. Quantile conditional value; β0( ) represents the quantiles The intercept term below; β1( ) represents the quantiles The coefficient of the i-th predictor represents the marginal effect of the factor at different locations in the distribution (e.g., low, medium, and high digestibility levels); the coefficient is estimated by minimizing the check function, which enables the model to capture outliers and asymmetric distributions.

[0064] Taking equation A1 in Table 1 as an example, we can extend it to a quantile regression framework. For example, for quantiles... =0.5 (median), the quantile regression equation can be written as:

[0065] ;

[0066] In the formula, represents the median of standard ileal amino acid digestibility under given predictors; X1 is the measured value of crude ash; X2 is the measured value of acid detergent fiber; X3 is the measured value of neutral detergent fiber; X4 is the measured value of bulk density; β0(0.5) is the baseline digestibility at the median quantile, representing the baseline level of laying hens in the typical laying stage; β1(0.5) is the coefficient of crude ash at the median quantile, representing the strength of the influence of inorganic matter content on digestibility; β2(0.5) is the coefficient of acid detergent fiber at the median quantile, representing the inhibitory effect of cellulose on digestion; β3(0.5) is the coefficient of neutral detergent fiber at the median quantile, representing the promoting effect of hemicellulose on microbial activity; β4(0.5) is the coefficient of bulk density at the median quantile, representing the effect of feed density on chyme flow.

[0067] Next, based on the quantile regression framework, biological constraints are introduced to integrate the coupling relationship between the ileal microbial metabolic pathway of laying hens and the amino acid digestion process. This is achieved by adding microbial-related variables or constraint conditions to make the equation conform to physiological reality. For example, the microbial metabolic pathway can affect the digestion efficiency of amino acids, so a microbial activity factor is introduced as an interaction term or an additional predictor. Biological constraints can be applied through constraint optimization, such as ensuring that the predicted value is within the physiologically feasible range (such as SID between 0-100%).

[0068] Based on the quantile regression equation, the constraint processing can be represented as a modified equation:

[0069] ;

[0070] In the formula, M is a microbial activity factor (such as ileal microbial community abundance); is the coefficient of the microbial activity factor at the quantile , reflecting the direct enhancing effect of microbial metabolism on digestibility; is the interaction term coefficient, representing the synergistic effect of the predictor microbial activity.

[0071] Further, step S400 includes steps S410 to S430.

[0072] Step S410, according to the candidate prediction equation, a multi-stage screening process is carried out, by establishing a decision model based on dynamic programming to analyze the metabolic characteristic differences of laying hens in the peak egg production period and the molting period, to obtain a preliminary screening equation set;

[0073] Step S420, according to the preliminary screening equation set, a biological rationality verification process is carried out, by constructing a phase matching model of the amino acid metabolic pathway of laying hens and the prediction equation to analyze the physiological fitness, to obtain a verified equation set;

[0074] Step S430, according to the verified equation set, an optimization process is performed, a compensation mechanism of endogenous amino acid loss of laying hens is introduced to establish an equation correction model, and an optimized equation is obtained.

[0075] Specifically, step S410 adopts a dynamic programming algorithm to perform multi-stage screening on the candidate prediction equation. The algorithm describes the change rule of the metabolic characteristics of laying hens from the egg laying peak period to the molting period by establishing a state transition equation, and evaluates the fitness of the prediction equation at each physiological stage node, so as to screen an equation subset that can adapt to the change of different physiological states. The state transition equation is expressed as:

[0076] ;

[0077] In the formula, S t is the metabolic state index (dimensionless) of the laying hen at time t, reflecting the current metabolic efficiency; S t+1 is the metabolic state index of the laying hen at time t+1; represents the egg laying stage identifier at time t; represents the egg laying stage identifier at time t+1; κ is a feed input influence coefficient, representing the immediate effect of feed change on the metabolic state; I t represents the feed input index at time t; η is a stage change influence coefficient, representing the adjustment amplitude of the metabolic state caused by the conversion of the egg laying stage.

[0078] Step S420 constructs a phase matching model on this basis, compares and analyzes the output result of the prediction equation with the actual amino acid metabolic pathway in the laying hen, calculates the phase difference between the predicted value and the actual physiological value in the time dimension, evaluates the degree of coincidence between the equation output and the biological process, and ensures that the screened equation meets the digestive physiological characteristics of the laying hen. The phase matching model quantifies the prediction deviation by time series alignment, and is expressed as:

[0079] ;

[0080] In the formula, is the optimal time offset (unit: day), which is the offset that maximizes the cross-correlation, and reflects the phase difference between the prediction and the actual value; is the time offset, which represents the time delay or advance of the predicted value relative to the actual value; represents the actual measured SID value at time t; represents the predicted standard ileal amino acid digestibility at time t, which comes from the output of the candidate equation; T is the total time, representing the length of the observation period.

[0081] Step S430 further introduces an endogenous amino acid loss compensation mechanism, a quantitative relationship model between loss amount and feed composition is established to systematically correct the prediction equation, which effectively solves the systematic deviation problem caused by ignoring endogenous loss in traditional prediction methods, so that the final optimized equation can more accurately reflect the true amino acid digestibility of laying hens. The whole optimization process embodies the progressive optimization idea from mathematical screening to biological verification and then to physiological mechanism correction, which ensures that the prediction model has statistical reliability and conforms to physiological laws. The quantitative relationship model between loss amount and feed composition quantifies the influence of feed composition on endogenous loss by linear relationship, which is expressed as:

[0082] ;

[0083] In the formula, E is the endogenous amino acid loss, which represents the loss of endogenous secreted amino acids in the digestion process of laying hens; is the acid detergent fiber content in the feed; is the neutral detergent fiber content in the feed; is the crude protein content in the feed; λ0 is the baseline endogenous loss, which represents the minimum loss without fiber and protein; λ1, λ2, λ3 represent the influence coefficients of each component;

[0084] In a preferred embodiment of the present application, the content corresponding to the prediction factor is brought into the standard ileal amino acid digestibility prediction equation of any growth stage to calculate the standard ileal amino acid digestibility, as shown in Table 2, the measured average value is 79.95%, and the relative deviation is between 2.08% and 2.55%. Among them, the relative deviation of candidate prediction equation A03 is only 2.08%, so equation A03 is the prediction equation with the highest accuracy. But the significance of A03 does not meet the requirements, so A04 is the optimal prediction equation.

[0085] Table 2 Comparison of predicted and measured values of standard ileal amino acid digestibility of secondary powder raw materials

[0086]

[0087] Further, step S500 includes step S510 to step S530.

[0088] Step S510, deviation distribution analysis processing is performed according to the optimized equation, a deviation probability model is established by statistically analyzing the deviation distribution characteristics of the prediction values and in vivo measured values in different laying stages, and a stage deviation analysis result is obtained;

[0089] Step S520, metabolic compensation processing is performed according to the stage deviation analysis result, dynamic compensation correction is performed through the establishment of a metabolic pathway correlation network between endogenous amino acid loss and digestibility deviation of laying hens, and a verification index after metabolic compensation is obtained.

[0090] Step S530, according to the verification index after metabolic compensation, the model convergence judgment processing is carried out, the model parameter optimization is carried out through setting the multi-objective convergence condition based on biological rationality, and the target prediction model is obtained.

[0091] Specifically, step S510 first analyzes the prediction deviation of the optimization equation at different egg production stages by establishing a deviation probability model. This method uses a Gaussian mixture model to fit the deviation distribution of the predicted value and the in vivo measured value, which can accurately identify the concentration trend and dispersion characteristics of the deviation distribution, especially the deviation pattern specific to key physiological stages such as the egg production peak period and the molting period. On this basis, step S520 constructs a metabolic pathway correlation network of endogenous amino acid loss and digestion rate deviation. This network model takes factors such as ileal microbial metabolites and digestive enzyme activity as network nodes, determines the contribution of each factor to the digestion rate deviation by calculating the correlation strength between nodes, and then realizes targeted dynamic compensation correction. Step S530 finally sets multi-objective convergence conditions based on biological rationality. These conditions include the allowed range of prediction error, the physiologically feasible interval, and the limitation of model complexity. Through multiple rounds of iterative optimization, the model parameters meet the requirements of statistical accuracy and physiological rationality at the same time, and finally obtain the target prediction model that has both prediction accuracy and conforms to the physiological characteristics of laying hens.

[0092] Further, step S600 includes step S610 to step S630.

[0093] Step S610, according to the target prediction model, the input parameter preprocessing is carried out, by synchronizing the predicted factor content of the feed raw material to be tested with the current egg production cycle of laying hens, calculating the time weight coefficient of each predicted factor in a specific egg production stage, and obtaining the spatiotemporal calibrated input parameter set;

[0094] Step S620, according to the input parameter set, the digestion rate calculation process is carried out, by establishing a dynamic response surface between the metabolic state of laying hens and the predicted factor, solving the optimal digestion rate prediction value based on the gradient descent method, and obtaining the preliminary prediction result;

[0095] Step S630, according to the preliminary prediction result, the physiological compensation processing is carried out, by constructing a feedback regulation network of endogenous amino acid loss of laying hens and feed composition, using an iterative approximation algorithm to adaptively correct the prediction result, and obtaining the standard ileal amino acid digestion rate prediction value.

[0096] Specifically, step S610 first synchronizes the predicted factor content of the feed raw material to be tested with the current egg laying cycle of the laying hen by a time series analysis method, extracts the fundamental frequency and harmonic components of the egg laying cycle by using Fourier transform technology, calculates the time weight coefficient of each predicted factor at different physiological stages, and realizes the space-time calibration of the input parameters; step S620 establishes a dynamic response surface model based on the calibrated parameter set, the model uses a radial basis function neural network to construct a nonlinear mapping relationship between the predicted factors and the metabolic state, finds the optimal solution on the multidimensional response surface by using a conjugate gradient descent algorithm, and obtains the digestibility prediction value that meets the current physiological state; step S630 further constructs a feedback regulation network of endogenous amino acid loss and feed composition, the network takes the feed ingredient decomposition rate, chyme flow rate and other parameters as nodes, uses an adaptive genetic algorithm for iterative optimization, and finally obtains the standard ileal amino acid digestibility prediction value that accurately reflects the real digestion condition of the laying hen through physiological adaptability correction of the prediction result.

[0097] By comparing the prediction method provided by the embodiment of the present application with the traditional in vivo determination method, the traditional in vivo determination method needs to be tested by animal experiments, use expensive determination devices, purchase and employ poultry breeders to raise poultry, and measure the standard ileal amino acid digestibility based on animal experiments. Taking the determination of the standard ileal amino acid digestibility of 66 raw materials as an example, as shown in Table 3, the traditional standard ileal amino acid digestibility determination method costs about 132,000 yuan from the preparation of the experiment to the end of the experiment, takes 30-60 days, needs farm rental fees, farm management fees, employee fees of breeding personnel and test personnel, and purchase fees of amino acid determination devices and other equipment, and consumes a large amount of manpower, material resources and time. However, the in vitro prediction method provided by the embodiment of the present application only needs to measure 1-4 key predicted factors in the secondary powder after the target standard ileal amino acid digestibility prediction model is constructed, and can directly calculate the prediction value of the standard ileal amino acid digestibility without animal experiments, and the determination of the above 66 samples (30 raw material samples + 36 chyme) only needs to cost about 1,100 yuan, which is 0.83% of the traditional cost, and the determination period is 1-4 days, which is 1.67%~3.33% of the traditional determination method period, greatly saving manpower, material resources, financial resources and time. In the future, if the method is loaded with near infrared spectroscopy (NIRS), the detection cost will be further reduced.

[0098] Table 3 Comparison of the prediction method of the embodiment of the present application with the traditional method

[0099]

[0100] Compared with the traditional in vivo method, the in vitro prediction method of the embodiment of the present application does not need to use animal experiments, saves the farm rental fee, animal purchase and feeding cost, avoids the harm of force-feeding method to animal welfare, greatly improves the sample evaluation quantity per unit time, saves the evaluation period, and therefore greatly saves manpower, material resources, financial resources and time. In addition, since the prediction factor range screened by the embodiment of the present application is wide, the prediction accuracy of the standard ileal amino acid digestibility is significantly improved. In addition, the prediction method of the embodiment of the present application is simple and easy to implement, compared with the traditional Excel calculation method, after embedding the prediction equation into the poultry raw material net energy prediction program, a large number of popularization can be carried out through the WeChat mini program or other mobile phone applications and computer software, and the prediction method has high use value.

[0101] Embodiment 2

[0102] As Figure 2 shown, the present embodiment provides a prediction system for egg chicken multi-stage feed amino acid utilization rate, the system comprises:

[0103] The acquisition module 901 is used for acquiring the chemical composition determination value and the standard ileal amino acid digestibility in vivo determination value of the feed raw material sample of the egg chicken at different growth stages.

[0104] The screening module 902 is used for screening the prediction factor according to the chemical composition determination value and the standard ileal amino acid digestibility in vivo determination value, performing dimensionality reduction mining on the correlation between the nutritional ingredients of the feed raw material and the metabolism characteristics of the egg chicken during the egg laying period through Akaike information criterion analysis, and obtaining a prediction factor combination.

[0105] The construction module 903 is used for constructing a prediction model according to the prediction factor combination, establishing a multiple regression relationship between the egg chicken ileal chyme amino acid flow characteristics and the prediction factor combination, and obtaining a candidate prediction equation.

[0106] The optimization module 904 is used for optimizing the candidate prediction equation, screening an optimized equation by comparing the determination coefficients and biological rationality of each equation at different growth stages of the egg chicken.

[0107] The verification module 905 is used for verifying the model according to the optimized equation, constructing a verification mechanism by calculating the relative deviation of the predicted value and the in vivo determination value, and combining the endogenous amino acid loss correction, and obtaining a target prediction model.

[0108] The prediction module 906 is used for predicting the digestibility according to the target prediction model, inputting the prediction factor content of the feed raw material to be tested into the target prediction model, and obtaining the predicted standard ileal amino acid digestibility value of the current feed formula.

[0109] In one specific embodiment of the present application, the screening module 902 comprises:

[0110] The first screening unit is configured to perform dynamic correlation analysis according to the chemical component determination value and the standard ileum amino acid digestion rate in vivo determination value, identify a nutritional factor that is synchronously responsive to physiological changes in the egg laying period by establishing a time series phase synchronization model of the fluctuation of the laying rate of the laying hen and the metabolic demand of the amino acid, and perform dynamic correlation analysis according to the chemical component determination value and the standard ileum amino acid digestion rate in vivo determination value.

[0111] The second screening unit is configured to perform combination optimization processing according to the nutritional factor, filter a minimum factor set having functional cooperativity in the peak egg laying period and the molting period of the laying hen by analyzing functional correlation strength between factors and constructing a modular grouping based on a clustering coefficient, and perform combination optimization processing according to the nutritional factor.

[0112] The third screening unit is configured to perform biological verification processing according to the minimum factor set, obtain a combination of prediction factors representing the amino acid digestion characteristics in the egg laying stage of the laying hen by establishing a multi-level interaction network model between the ileum microbial community of the laying hen and the prediction factor, and perform biological verification processing according to the minimum factor set.

[0113] In one specific embodiment of the present application, the construction module 903 includes:

[0114] The first construction unit is configured to perform dynamic modeling processing according to the combination of prediction factors, construct a primary regression equation based on time dimension correction by introducing time series phase adjustment of the ileum chyme amino acid flow rate in the egg laying period of the laying hen, and perform dynamic modeling processing according to the combination of prediction factors.

[0115] The second construction unit is configured to perform structure optimization processing according to the primary regression equation, establish a dynamic prediction framework based on quantile regression by analyzing the distribution characteristics of the amino acid absorption efficiency in different egg laying stages, and perform structure optimization processing according to the primary regression equation.

[0116] The third construction unit is configured to perform biological constraint processing according to the dynamic prediction framework, obtain a candidate prediction equation that conforms to the physiological characteristics of the laying hen by integrating the coupling relationship between the ileum microbial metabolic pathway of the laying hen and the amino acid digestion process, and perform biological constraint processing according to the dynamic prediction framework.

[0117] In one specific embodiment of the present application, the optimization module 904 includes:

[0118] The first optimization unit is configured to perform multi-stage screening processing according to the candidate prediction equation, obtain a preliminary screening equation set by establishing a decision model based on dynamic programming to analyze the metabolic characteristic differences between the peak egg laying period and the molting period of the laying hen, and perform multi-stage screening processing according to the candidate prediction equation.

[0119] The second optimization unit is configured to perform biological rationality verification processing according to the preliminary screening equation set, obtain a verified equation set by constructing a phase matching model of the amino acid metabolic pathway of the laying hen and the prediction equation to analyze physiological fitness, and perform biological rationality verification processing according to the preliminary screening equation set.

[0120] The third optimization unit is configured to perform optimization processing according to the verified equation set, establish an equation correction model by introducing a compensation mechanism for endogenous amino acid loss of the laying hen, and obtain an optimized equation.

[0121] In one specific embodiment of the present application, the verification module 905 includes:

[0122] The first verification unit is configured to perform bias distribution analysis processing according to the optimized equation, establish a bias probability model by statistically analyzing bias distribution characteristics of predicted values and in-vivo measured values at different egg-laying stages, and obtain a stage-by-stage bias analysis result.

[0123] The second verification unit is configured to perform metabolic compensation processing according to the stage-by-stage bias analysis result, perform dynamic compensation correction by establishing a metabolic pathway correlation network between endogenous amino acid loss and digestion rate bias of the laying hen, and obtain a verification index after metabolic compensation.

[0124] The third verification unit is configured to perform model convergence determination processing according to the verification index after metabolic compensation, perform model parameter optimization by setting multi-objective convergence conditions based on biological rationality, and obtain a target prediction model.

[0125] In one specific embodiment of the present application, the prediction module 906 includes:

[0126] The first prediction unit is configured to perform input parameter preprocessing according to the target prediction model, perform phase synchronization between predicted factor content of a feed material to be tested and a current egg-laying period of the laying hen, calculate time weight coefficients of each predicted factor at a specific egg-laying stage, and obtain a set of input parameters after time and space calibration.

[0127] The second prediction unit is configured to perform digestion rate calculation processing according to the set of input parameters, establish a dynamic response surface between metabolic states of the laying hen and the predicted factors, solve an optimal digestion rate prediction value based on a gradient descent method, and obtain a preliminary prediction result.

[0128] The third prediction unit is configured to perform physiological compensation processing according to the preliminary prediction result, construct a feedback regulation network between endogenous amino acid loss of the laying hen and feed composition, and perform physiological adaptability correction on the prediction result by using an iterative approximation algorithm, to obtain a standard ileal amino acid digestion rate prediction value.

[0129] Example 3

[0130] Corresponding to the method embodiment above, the present embodiment also provides a laying hen multi-stage feed amino acid utilization rate prediction device. The laying hen multi-stage feed amino acid utilization rate prediction device described below can be mutually referred to with the laying hen multi-stage feed amino acid utilization rate prediction method described above.

[0131] Figure 3FIG. 8 is a block diagram of an egg-laying hen multi-stage feed amino acid utilization rate prediction device 800 according to an exemplary embodiment. As shown in Figure 3 The egg-laying hen multi-stage feed amino acid utilization rate prediction device 800 can include a processor 801, a memory 802, as shown. The egg-laying hen multi-stage feed amino acid utilization rate prediction device 800 can also include one or more of a multimedia component 803, an I / O interface 804, and a communication component 805.

[0132] The processor 801 is configured to control overall operation of the device 800. The memory 802 is configured to store data and applications used by at least one processor 801 in the device 800. The screen 803 can be a touch screen, and the audio component can be configured to output and / or input audio signals. For example, the audio component can include a microphone for receiving external audio signals. The received audio signals can be further stored in the memory 802 or transmitted through the communication component 805. The audio component also includes at least one speaker for outputting audio signals. The I / O interface 804 provides an interface between the processor 801 and other interface modules, which can be a keyboard, mouse, button, etc. These buttons can be virtual buttons or physical buttons. The communication component 805 is configured to enable wired or wireless communication between the device 800 and other devices. The wireless communication, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G or 4G, or a combination of one or more of them, so the corresponding communication component 805 can include a Wi-Fi module, a Bluetooth module, an NFC module.

[0133] In an exemplary embodiment, a device 800 for predicting the amino acid utilization of multi-stage feed for laying hens can be implemented by one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs), Field Programmable Gate Arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic elements for executing the above-mentioned method for predicting the amino acid utilization of multi-stage feed for laying hens.

[0134] In another exemplary embodiment, a computer-readable storage medium including program instructions that, when executed by a processor, implement the steps of the above-mentioned method for predicting the amino acid utilization of multi-stage feed for laying hens is also provided. For example, the computer-readable storage medium can be the above-mentioned memory 802 including program instructions that can be executed by the processor 801 of the device 800 for predicting the amino acid utilization of multi-stage feed for laying hens to complete the above-mentioned method for predicting the amino acid utilization of multi-stage feed for laying hens.

[0135] The above merely provides the specific implementation of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of the changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application.

Claims

1. A method of predicting the amino acid utilization of a multi-stage layer feed, characterized by, The method comprises the following steps: obtaining the chemical composition determination value and the standard ileal amino acid digestibility in-vivo determination value of the feed raw material sample of the laying hen at different growth stages; performing predictive factor screening according to the chemical composition determination value and the standard ileal amino acid digestibility in-vivo determination value, performing dimensionality reduction mining on the correlation between the nutritional components of the feed raw material and the metabolic characteristics of the laying hen during the egg production cycle by Akaike information criterion analysis, and obtaining a predictive factor combination; performing predictive model construction according to the predictive factor combination, establishing a multiple regression relationship between the characteristics of the laying hen ileal chyme amino acid flow and the predictive factor combination, and obtaining a candidate predictive equation; performing optimization according to the candidate predictive equation, comparing the determination coefficients and biological rationalities of each equation at different growth stages of the laying hen, and screening to obtain an optimized equation; performing model verification according to the optimized equation, calculating the relative deviation of the predicted value and the in-vivo determination value, and combining the endogenous amino acid loss correction to construct a verification mechanism, and obtaining a target predictive model; performing digestibility prediction according to the target predictive model, inputting the predictive factor content of the laying hen feed raw material to be tested into the target predictive model, and obtaining the predicted standard ileal amino acid digestibility value of the current feed formula. In the method for predicting the amino acid utilization rate of the multi-stage feed of the laying hen, the derivation of the candidate predictive equation is based on a multiple linear regression model, and the general form of the multiple linear regression model is: ； where β0is an intercept term; β i is a regression coefficient for the i-th predictor, where i = 1, 2,..., k; X i represents a measured value of the predictor; is an error term; Based on the primary regression equation, the model is structurally optimized by quantile regression method, and the quantile regression is used to capture the distribution characteristics of the amino acid absorption efficiency of the laying hen at different egg production stages, and the general form of the quantile regression model is: ； where Q SID (τ|X) denotes the conditional value of the τ quantile of the standard ileal amino acid digestibility given the predictor X; β0(τ) is the intercept term at quantile τ; β1(τ) is the coefficient of the i-th predictor at quantile τ, representing the marginal effect of this predictor at different locations of the distribution; the coefficients are estimated by minimizing a check function, enabling the model to capture outliers and asymmetric distributions; According to the target predictive model, the predictive factor content of the laying hen feed raw material to be tested is input into the target predictive model to obtain the predicted standard ileal amino acid digestibility value of the current feed formula, which comprises: According to the target predictive model, the predictive factor content of the feed raw material to be tested is phase-synchronized with the current egg production cycle of the laying hen, the time weight coefficient of each predictive factor at a specific egg production stage is calculated, and a spatiotemporal calibrated input parameter set is obtained; According to the input parameter set, the digestibility calculation process is performed, the dynamic response surface between the metabolic state of the laying hen and the predictive factor is established, the optimal digestibility prediction value is solved based on the gradient descent method, and a preliminary prediction result is obtained; According to the preliminary prediction result, the physiological compensation processing is performed, the feedback regulation network of the endogenous amino acid loss of the laying hen and the feed composition is constructed, the iterative approximation algorithm is adopted to correct the prediction result in terms of physiological adaptability, and the standard ileal amino acid digestibility prediction value is obtained.

2. The method for predicting the amino acid utilization rate of multi-stage feed for laying hens according to claim 1, characterized in that, According to the chemical composition determination value and the standard ileal amino acid digestibility in-vivo determination value, the predictive factor screening comprises the following steps: According to the chemical composition determination value and the standard ileal amino acid digestibility in-vivo determination value, dynamic correlation analysis is performed, a time series phase synchronization model of the laying hen egg production rate fluctuation and the amino acid metabolic demand is established, and the nutritional factors that respond synchronously with the physiological changes during the egg production cycle are identified. According to the nutritional factor, a combination optimization process is performed, a functional correlation strength between factors is analyzed, a modular grouping based on a clustering coefficient is constructed, and a minimum factor set with functional synergy in the egg laying peak period and the molting period of the laying hen is screened out; According to the minimum factor set, a biological verification process is performed, a multi-level interaction network model between the cecal microbial community of the laying hen and the prediction factor is established, and a prediction factor combination representing the amino acid digestion characteristics in the egg laying stage of the laying hen is obtained.

3. The method for predicting the amino acid utilization rate of multi-stage feed for laying hens according to claim 1, characterized in that, According to the prediction factor combination, a prediction model is constructed, including: According to the prediction factor combination, a dynamic modeling process is performed, a time series phase adjustment of the amino acid flow rate of the cecal chyme in the egg laying cycle of the laying hen is introduced, and a primary regression equation based on time dimension correction is constructed; According to the primary regression equation, a structure optimization process is performed, the distribution characteristics of the amino acid absorption efficiency in different egg laying stages are analyzed, and a dynamic prediction framework based on quantile regression is established; According to the dynamic prediction framework, a biological constraint process is performed, the coupling relationship between the cecal microbial metabolic pathway of the laying hen and the amino acid digestion process is integrated, and a candidate prediction equation conforming to the digestion physiological characteristics of the laying hen is obtained.

4. The method for predicting the amino acid utilization rate of multi-stage feed for laying hens according to claim 1, characterized in that, According to the candidate prediction equation, optimization is performed, including: According to the candidate prediction equation, a multi-stage screening process is performed, a decision model based on dynamic programming is established to analyze the metabolic characteristic differences between the egg laying peak period and the molting period of the laying hen, and a preliminary screening equation set is obtained; According to the preliminary screening equation set, a biological rationality verification process is performed, a phase matching model of the amino acid metabolic pathway of the laying hen and the prediction equation is constructed to analyze the physiological fitness, and a verified equation set is obtained; According to the verified equation set, an optimization process is performed, a compensation mechanism for endogenous amino acid loss of the laying hen is introduced to establish an equation correction model, and an optimized equation is obtained.

5. The method for predicting the amino acid utilization rate of multi-stage feed for laying hens according to claim 1, characterized in that, According to the optimized equation, model verification is performed, including: According to the optimized equation, a deviation distribution analysis process is performed, a deviation probability model is established by statistically analyzing the deviation distribution characteristics of the predicted value and the in vivo measured value in different egg laying stages, and a stage deviation analysis result is obtained; According to the stage deviation analysis result, a metabolic compensation process is performed, a metabolic pathway association network between the endogenous amino acid loss and the digestion rate deviation of the laying hen is established for dynamic compensation correction, and a verification index after metabolic compensation is obtained; According to the verification index after metabolic compensation, a model convergence determination process is performed, a multi-objective convergence condition based on biological rationality is set for model parameter optimization, and a target prediction model is obtained.

6. A predictive system for amino acid utilization rate in multi-stage feed for laying hens, characterized in that, including: An acquisition module is configured to acquire chemical composition measurement values of feed raw material samples at different growth stages of laying hens and standard in vivo measurement values of ileal amino acid digestibility; A screening module is configured to perform prediction factor screening based on the chemical composition measurement values and the standard in vivo measurement values of ileal amino acid digestibility, to perform dimensionality reduction mining on the correlation between the nutritional ingredients of the feed raw materials and the metabolic characteristics in the egg laying cycle of the laying hen by Akaike information criterion analysis, and to obtain a prediction factor combination; The construction module is configured to perform prediction model construction according to the combination of the prediction factors, to obtain a candidate prediction equation by establishing a multiple regression relationship between the flow characteristics of the egg hen ileum chyme amino acid and the combination of the prediction factors; The optimization module is configured to perform optimization according to the candidate prediction equation, to obtain an optimized equation by comparing the determination coefficients and biological rationalities of each equation in different growth stages of the egg hen; The verification module is configured to perform model verification according to the optimized equation, to obtain a target prediction model by calculating the relative deviation of the predicted value and the in-vivo measured value, and combining the endogenous amino acid loss correction to construct a verification mechanism; The prediction module is configured to perform digestion rate prediction according to the target prediction model, to obtain the predicted standard ileum amino acid digestion rate value of the current feed formula by inputting the prediction factor content of the to-be-tested egg hen feed raw material into the target prediction model; In the prediction method for the multi-stage feed amino acid utilization rate of the egg hen, the derivation of the candidate prediction equation is based on a multiple linear regression model, and the general form of the multiple linear regression model is: ； wherein β0is an intercept term; β i is a regression coefficient for the i-th predictor, where i = 1, 2,..., k; X i denotes a measured value of the predictor; is an error term; Based on the primary regression equation, the model is structurally optimized by a quantile regression method, the quantile regression is used to capture the distribution characteristics of the amino acid absorption efficiency of the egg hen in different egg production stages, and the general form of the quantile regression model is: ； where Q SID (τ|X) denotes the conditional value of the τ quantile of the standard ileal amino acid digestibility given the predictor X; β0(τ) is the intercept term at quantile τ; β1(τ) is the coefficient of the i-th predictor at quantile τ, representing the marginal effect of this predictor at different locations of the distribution; the coefficients are estimated by minimizing a check function, enabling the model to capture outliers and asymmetric distributions; According to the target prediction model, the digestion rate prediction is performed by inputting the prediction factor content of the to-be-tested egg hen feed raw material into the target prediction model to obtain the predicted standard ileum amino acid digestion rate value of the current feed formula, which includes: According to the target prediction model, the input parameter preprocessing is performed by synchronizing the prediction factor content of the to-be-tested feed raw material with the current egg laying period, calculating the time weight coefficient of each prediction factor in a specific egg laying stage, and obtaining a spatiotemporal calibrated input parameter set; According to the input parameter set, the digestion rate calculation processing is performed by establishing a dynamic response surface between the egg hen metabolic state and the prediction factor, solving the optimal digestion rate prediction value based on the gradient descent method, and obtaining a preliminary prediction result; According to the preliminary prediction result, the physiological compensation processing is performed by constructing a feedback regulation network of the endogenous amino acid loss of the egg hen and the feed composition, and the iterative approximation algorithm is used to perform physiological adaptability correction on the prediction result to obtain a standard ileum amino acid digestion rate prediction value.

7. The system for predicting the amino acid utilization of layers' multi-stage feed according to claim 6, wherein, The screening module includes: The first screening unit is configured to perform dynamic correlation analysis according to the chemical composition measured value and the standard ileum amino acid digestion rate in-vivo measured value, to identify the nutrient factors that respond synchronously with the physiological changes in the egg laying period by establishing a time series phase synchronization model of the egg hen laying rate fluctuation and the amino acid metabolic demand; The second screening unit is configured to perform combination optimization processing according to the nutrient factors, to obtain a minimum factor set with functional synergy in the egg laying peak period and the molting period by analyzing the functional correlation strength between the factors and constructing a modular grouping based on a clustering coefficient. The third screening unit is configured to perform biological verification processing according to the minimum factor set, to obtain a combination of prediction factors representing the amino acid digestion characteristics of the laying hens in the egg production stage by establishing a multi-level interaction network model between the cecal microbial community of the laying hens and the prediction factors.

8. The system for predicting the amino acid utilization of layers' multi-stage feed according to claim 6, wherein, The construction module comprises: The first construction unit is configured to perform dynamic modeling processing according to the combination of prediction factors, to construct a primary regression equation based on time dimension correction by introducing time sequence phase adjustment of the amino acid flow rate of the cecal chyme of the laying hens in the egg production cycle; The second construction unit is configured to perform structure optimization processing according to the primary regression equation, to establish a dynamic prediction framework based on quantile regression by analyzing the distribution characteristics of the amino acid absorption efficiency in different egg production stages; The third construction unit is configured to perform biological constraint processing according to the dynamic prediction framework, to obtain a candidate prediction equation conforming to the physiological characteristics of the laying hens by integrating the coupling relationship between the cecal microbial metabolic pathways and the amino acid digestion process of the laying hens.

9. The system for predicting the amino acid utilization of layers' multi-stage feed according to claim 6, wherein, The optimization module comprises: The first optimization unit is configured to perform multi-stage screening processing according to the candidate prediction equation, to obtain a preliminary screening equation set by establishing a decision model based on dynamic programming to analyze the metabolic characteristic differences between the egg production peak period and the molting period of the laying hens; The second optimization unit is configured to perform biological rationality verification processing according to the preliminary screening equation set, to obtain a verified equation set by constructing a phase matching model of the amino acid metabolic pathways of the laying hens and the prediction equation to analyze the physiological fitness; The third optimization unit is configured to perform optimization processing according to the verified equation set, to obtain an optimized equation by introducing a compensation mechanism for the endogenous amino acid loss of the laying hens to establish an equation correction model.

10. The system for predicting the amino acid utilization of layers' multi-stage feed according to claim 6, wherein, The verification module comprises: The first verification unit is configured to perform bias distribution analysis processing according to the optimized equation, to obtain a stage-by-stage bias analysis result by establishing a bias probability model based on the bias distribution characteristics of the prediction values and the in vivo measured values in different egg production stages; The second verification unit is configured to perform metabolic compensation processing according to the stage-by-stage bias analysis result, to obtain a metabolic compensation verified index by establishing a metabolic pathway association network between the endogenous amino acid loss and the digestion rate bias of the laying hens for dynamic compensation correction; The third verification unit is configured to perform model convergence determination processing according to the metabolic compensation verified index, to obtain a target prediction model by setting multi-objective convergence conditions based on biological rationality for model parameter optimization.

Images