Laying hen farming simulation system based on digital twinning

Abstract

The application discloses a kind of based on digital twinning egg chicken type breeding simulation system, it is related to intelligent breeding technical field, including data acquisition unit, for obtaining multidimensional state parameters, and output to biological state quantification unit;Biological state quantification unit is used to solve out individual metabolic rate index and group aggregation entropy;Time lag characteristic analysis unit is used to quantify and generate biological response time lag complexity index;Regulation and control instruction generation unit is used to calculate temperature demand reference value, and further generate final temperature setting target value;Model self-optimization unit is used to predict the biological state of next time, and compare with the actual observation state after regulation and control execution, generate prediction error, and the weight and coefficient preset in system are iterated and updated.The application quantifies egg chicken biological state and response characteristics accurately, realizes accurate, flexible and can continuously self-improve intelligent environmental regulation and control by combining model self-optimization, significantly improves breeding efficiency and animal welfare.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent breeding, in particular to an egg chicken breeding simulation system based on digital twinning. BACKGROUND

[0002] Modern egg chicken breeding industry is rapidly developing towards scale, intensification and automation. In order to protect the health and production performance of egg chickens, accurate control of the breeding environment becomes crucial. The existing technical solution mainly relies on deploying a sensor network inside the breeding facility to monitor temperature, humidity, key gas concentration and other macro physical environment parameters in real time. Based on these monitoring data, the environment control system (such as fans, wet curtains, heaters, etc.) can execute the preset control logic, for example, automatically start and stop the equipment when the temperature is lower or higher than a certain static threshold, so as to maintain the environment indicators within a relatively stable interval. This automated environment monitoring and control technology has replaced part of the management relying on human experience to a certain extent, and has improved the efficiency of breeding management. It is an indispensable technical foundation for current intensive breeding.

[0003] The existing technical solution has significant limitations. First, the core basis of its control is the physical environment parameters, which essentially regards the complex biological population as a passive responder to environmental changes, ignoring the internal physiological state, behavioral needs and diversity of egg chickens as living beings. The management mode is a one-size-fits-all mode based on age, which cannot meet the individualized environmental needs of individuals in the same flock due to differences in body weight, health and egg production level. Second, its control logic is usually a rigid "set-execution" mode, lacking consideration of the response characteristics of biological systems, and the control action itself may become a new stress source due to its abruptness. In addition, the stress state of the flock has traditionally relied on sampling detection of individuals. This method not only has a lag and interferes with animals, but also cannot reflect the overall real-time state of the group. Finally, if a model is included in the existing system, the model is usually static and does not change once it is established, which cannot adapt to the dynamic changes of the flock due to growth, health status changes or seasonal changes, resulting in the model losing accuracy over time and affecting the long-term effectiveness of control. SUMMARY

[0004] The purpose of the present application is to provide an egg chicken breeding simulation system based on digital twinning, which solves the problems in the background art.

[0005] To solve the above technical problems, the present application provides an egg chicken breeding simulation system based on digital twinning, comprising: a data acquisition unit for acquiring multi-dimensional state parameters in the breeding facility in real time and outputting to a biological state quantification unit, the multi-dimensional state parameters including physiological information of individual egg chickens, overall spatial position distribution of the flock and physical environment parameters;

[0006] a biological state quantification unit configured to calculate an individual metabolic rate index and a population aggregation entropy based on the physiological information and the overall spatial position distribution obtained by the data acquisition unit;

[0007] a time lag feature analysis unit configured to quantify and generate a biological response time lag complexity index based on the change process of the population aggregation entropy after the environmental regulation in the historical data;

[0008] a regulation instruction generation unit configured to calculate a temperature demand reference value based on the individual metabolic rate index and the population aggregation entropy generated by the biological state quantification unit, and further generate a final temperature setting target value in combination with the biological response time lag complexity index generated by the time lag feature analysis unit;

[0009] a model self-optimization unit configured to predict the biological state at the next moment based on the temperature setting target value and the current biological state, compare the predicted biological state with the actual observed state after the regulation execution, generate a prediction error, and iteratively update the preset weight and coefficient in the system according to the prediction error.

[0010] Preferably, the calculation process of the individual metabolic rate index by the biological state quantification unit comprises:

[0011] standardizing the collected body weight, egg production rate and age data to generate dimensionless physiological index values; and weighting and summing the dimensionless physiological index values and the respective preset weights to generate the individual metabolic rate index.

[0012] Preferably, the calculation process of the population aggregation entropy by the biological state quantification unit comprises:

[0013] virtually dividing the breeding space into a plurality of grid areas; identifying and counting the proportion of the number of chickens distributed in each grid area in the total number by an image analysis algorithm to generate a regional distribution proportion; and calculating the regional distribution proportion in a preset information entropy formula to generate the population aggregation entropy.

[0014] Preferably, the generation process of the biological response time lag complexity index by the time lag feature analysis unit comprises:

[0015] obtaining a time variation curve of the population aggregation entropy after a historical environmental regulation event; determining a time when the absolute value of the derivative of the time variation curve is continuously lower than a preset threshold to calculate a lag time of the biological response; normalizing the lag time to generate a dimensionless lag coefficient; and weighting and summing the dimensionless lag coefficient to generate the biological response time lag complexity index.

[0016] Preferably, the calculation process of the temperature demand reference value by the regulation instruction generation unit comprises:

[0017] acquire a preset standard temperature value, an ideal metabolic rate index, and an ideal aggregation entropy;

[0018] calculate a deviation between the current average metabolic rate index and the ideal metabolic rate index, and combine a preset temperature sensitivity coefficient to generate a metabolic rate temperature compensation amount;

[0019] calculate a deviation between the current group aggregation entropy and the ideal aggregation entropy, and combine a preset entropy sensitivity coefficient to generate an aggregation entropy temperature compensation amount;

[0020] correct the standard temperature value by using the metabolic rate temperature compensation amount and the aggregation entropy temperature compensation amount to generate a temperature demand reference value.

[0021] Preferably, the generation process of the temperature setting target value by the regulation instruction generation unit comprises:

[0022] substitute the biological response time lag complexity index into a preset index attenuation model to generate a regulation intensity factor; calculate a difference between the temperature demand reference value and the actually collected actual temperature to generate a temperature target deviation; multiply the temperature target deviation by the regulation intensity factor, and sum the actual temperature to generate the temperature setting target value of the next control time step.

[0023] Preferably, the prediction process of the biological state at the next time by the model self-optimization unit comprises:

[0024] construct a state transition prediction model; use the individual metabolic rate index and the group aggregation entropy observed at the current time, and the temperature setting target value generated by the regulation instruction generation unit as inputs; and perform operation through the state transition prediction model to output the predicted values of the individual metabolic rate index and the group aggregation entropy at the next time.

[0025] Preferably, the iterative updating process of the model self-optimization unit comprises:

[0026] after the regulation is executed, actually observe the individual metabolic rate index and the group aggregation entropy at the next time step; compare the actually observed values with the predicted values to calculate a weighted deviation to generate a prediction error; use the prediction error as a loss function, and adopt an optimization algorithm to minimize the loss function as the goal to periodically update all preset adjustable weights and coefficients in the system.

[0027] Compared with the prior art, the present application has the following beneficial effects:

[0028] 1. The present application realizes the deepening of the cognition of the breeding object by constructing a new biological state quantitative index, which is no longer limited to monitoring the external environment, but through the individual metabolic rate index, the key physiological information such as body weight, egg production rate and age affecting energy demand is integrated into a calculable quantitative index, so that the real needs of different individuals can be understood and responded, at the same time, the introduced group aggregation entropy can convert the macro spatial distribution pattern of the chicken group into a dynamic index for evaluating the overall comfort and stress level through non-contact image analysis method, realizing the real-time and global perception of the group welfare state.

[0029] 2. The present application establishes a dynamic and flexible closed-loop control decision mechanism, the target of the control is no longer the fixed industry standard, but a temperature reference value closely fitting the current real biological needs of the chicken group is dynamically calculated and generated according to the real-time individual metabolic rate and group aggregation entropy feedback, and the hysteresis and complexity of the biological group responding to the environmental change are innovatively quantified, and the strength and rhythm of the control instruction are dynamically adjusted based on this, realizing a gradual flexible compensation strategy, which effectively avoids the secondary stress on the chicken group caused by environmental mutation, and improves the stability and safety of the control process.

[0030] 3. Through the built-in state transition prediction model, the present application can pre-act the biological consequences that may be caused by the control decision before its execution, after obtaining the real observation data after the control, the error between the prediction and the reality will be calculated, and the optimal algorithm will be used to update all the key model parameters in the system reversely, this closed-loop mechanism of "prediction-verification-correction" enables the digital twin model to continuously learn and evolve, and dynamically approximates the real situation of the physical farm, thereby ensuring the long-term adaptability, accuracy and robustness of the system under different growth periods, different health conditions and variable external environment of the chicken group. BRIEF DESCRIPTION OF DRAWINGS

[0031] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, brief descriptions will be given to the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor;

[0032] Figure 1 The logical block diagram of the system of the present application. DETAILED DESCRIPTION

[0033] With reference to the drawings of the embodiments of the present application, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, and not all the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work are within the protection scope of the present application.

[0034] Please refer to Figure 1 The present application provides a kind of based on digital twin's egg chicken type breeding simulation system, comprising: data acquisition unit, for real-time acquisition in the multi-dimensional state parameter of breeding facility, and output to biological state quantification unit, multi-dimensional state parameter includes the physiological information of egg chicken individual, the overall spatial position distribution of chicken group and physical environment parameter;

[0035] Biological state quantification unit is used to calculate individual metabolic rate index and group aggregation entropy according to physiological information and overall spatial position distribution obtained by data acquisition unit.

[0036] Time lag feature analysis unit is used to quantify and generate biological response time lag complexity index based on the change process of group aggregation entropy in historical data after environmental regulation.

[0037] Control instruction generation unit is used to calculate temperature demand reference value based on individual metabolic rate index and group aggregation entropy generated by biological state quantification unit, and further generate final temperature setting target value in combination with biological response time lag complexity index generated by time lag feature analysis unit.

[0038] Model self-optimization unit is used to predict biological state at next time based on temperature setting target value and current biological state, compare with actual observation state after regulation execution, generate prediction error, and iteratively update preset weight and coefficient in system according to prediction error.

[0039] The present embodiment provides an egg chicken type breeding simulation system based on digital twin; the system runs in a computing device configured with a processor and a memory, and aims to realize accurate perception, dynamic simulation and closed-loop feedback regulation of egg chicken breeding environment through a series of closely coupled computing units; the system as a whole constitutes a self-consistent technical closed loop, including data acquisition unit, biological state quantification unit, time lag feature analysis unit, control instruction generation unit and model self-optimization unit.

[0040] A data acquisition unit, which aims to provide real-time, multi-dimensional data input for the entire digital twin system, is the starting point of the mapping from the physical world to the digital space; in this embodiment, this unit is realized through a sensor network deployed in the breeding facility; specifically, it uses a top-mounted machine vision system combined with individual radio frequency identification tags to capture and record the physiological information of selected sample individual hens, which in this embodiment refers to core data reflecting the individual's health and production status, mainly including body weight, age, and egg production rate; at the same time, the top-mounted imaging device is used for continuous shooting to obtain the overall spatial position distribution of the chicken population; in addition, a standard environmental sensor array is deployed to continuously monitor the physical environmental parameters in the breeding space, including temperature, humidity, and the concentration of key gases such as ammonia and carbon dioxide; all collected raw data is structured and output to the subsequent unit in real time;

[0041] A biological state quantification unit, which aims to convert the collected raw, heterogeneous data into standardized, computable indicators that can represent the biological state of hens; in this embodiment, this unit receives physiological information and spatial position distribution data from the data acquisition unit; it contains two core algorithm modules: one, based on the body weight, egg production rate, and age of individual hens, a comprehensive individual metabolic rate index is calculated; the second, based on the overall spatial position distribution of the chicken population, the group aggregation entropy is calculated through information entropy theory; these two indices together constitute a digital description of the current biological state of hens;

[0042] A time lag feature analysis unit, which aims to quantify the hysteresis and complexity of the biological system (chicken population) in response to environmental control instructions, providing key evidence for subsequent adaptive control strategies; this unit is not based on instantaneous data, but through in-depth analysis of the environmental control events stored in the historical database and the subsequent biological state change process; specifically, it analyzes how long it takes for the group aggregation entropy representing the group stress level to reach a new stable state after a temperature or ventilation adjustment; this process is quantified to generate a biological response time lag complexity index;

[0043] A control instruction generation unit, which aims to calculate the optimal environmental control target value based on the current biological state and the response characteristics of the system; in this embodiment, this unit is the key link to achieve adaptive control; it first calculates a theoretical temperature demand benchmark value based on the individual metabolic rate index and the group aggregation entropy generated by the biological state quantification unit, which reflects the ideal temperature of the chicken population under the current state; then, it combines the biological response time lag complexity index generated by the time lag feature analysis unit to dynamically adjust the intensity and speed of control, further generating a final temperature setting target value that will be executed in the next control cycle;

[0044] A model self-optimization unit aims to ensure that the system model can continuously learn and improve itself, ensuring its accuracy and adaptability in the long run. In this embodiment, this unit forms a complete prediction-feedback-correction closed loop. After the generation of control instructions and before their execution, the unit will predict the biological state of the chicken flock (i.e., individual metabolic rate index and flock aggregation entropy) at the next moment based on the current biological state and the target temperature setting value to be executed, through an internal state transition prediction model. After the actual control is executed, the system compares the actual observed state at the next moment with the previous predicted value, generating a prediction error. This error is used as a loss function to drive an optimization algorithm to periodically update all pre-set and adjustable weights and coefficients in the system, achieving self-calibration and continuous optimization of the model.

[0045] Through the cooperative work of the above five units, this embodiment constructs a complete digital twin closed-loop system from perception to decision-making to optimization. Compared with the traditional farming mode relying on static feeding standards or artificial experience, this system can accurately quantify the comprehensive biological state of individual and group of laying hens in real time, and innovatively incorporates the hysteresis of biological response into the decision-making model. This makes environmental control no longer a simple "setting-execution", but an intelligent and flexible process that can predict results, adapt to objects, and self-correct. Ultimately, the system can ensure the welfare of laying hens, reduce stress response, significantly improve feed conversion rate and egg production performance, maximize farming benefits, and provide solid technical support for standardization and intelligentization of the farming process.

[0046] Embodiment 2:

[0047] The individual metabolic rate index calculation process of the biological state quantification unit includes:

[0048] The collected body weight, egg production rate, and age data are standardized to generate dimensionless physiological index values. The dimensionless physiological index values are weighted and summed with their respective pre-set weights to generate the individual metabolic rate index.

[0049] This embodiment is a specific implementation of the individual metabolic rate index calculation process in the biological state quantification unit of embodiment 1. This process aims to unify raw physiological data in different physical units into a dimensionless index framework to achieve standardized comparison of metabolic levels among different individuals.

[0050] To achieve the above object, the biological state quantification unit calculates the individual metabolic rate index, which first standardizes the collected core physiological data of the laying hen, including body weight, egg laying rate and age. The standardization process is a mathematical transformation that eliminates the numerical differences caused by different units and dimensions, and makes the data comparable. In this embodiment, the relative deviation of each real-time physiological data from the ideal value at the corresponding age is calculated to generate dimensionless physiological index values.

[0051] Specifically, the standardization process can be calculated by the following formula:

[0052] ;

[0053] is the dimensionless physiological index value;

[0054] is the collected real-time physiological data (e.g. current body weight or egg laying rate);

[0055] is the ideal physiological data value corresponding to the current age, which is pre-set in the system and can be derived from authoritative feeding management guidelines or historical data statistics of the best production batch;

[0056] Based on this, the system weights and sums these dimensionless physiological index values with their respective pre-set weights to generate the final individual metabolic rate index. This calculation process is defined by a pre-set individual metabolic rate index model. The individual metabolic rate index model is a linear weighted model based on the recognized principle of animal energy metabolism, which considers that the total metabolic level of an organism is a weighted linear combination of its basic physiological indicators. Its mathematical expression is:

[0057] ;

[0058] M is the individual metabolic rate index, which is a dimensionless comprehensive value that quantifies the comprehensive energy metabolism level of an individual in the current physiological state;

[0059] m, r, a are dimensionless physiological index values corresponding to the standardized body weight, egg laying rate and age, respectively. Their data is derived from the standardized processing of the original data obtained by the data acquisition unit;

[0060] are preset weights, which are dimensionless weight coefficients corresponding to the body weight, the egg laying rate and the age, respectively; the role of the preset weights is to represent the relative contribution of each physiological index to the total metabolic level; the initial values of the preset weights are set according to the published research results of poultry physiology and nutrition, and the sum of the preset weights is constrained to be 1, and the preset weights can be iteratively updated in the subsequent model self-optimization process;

[0061] Through the above embodiments, the present application provides a refined quantitative means for the physiological state of laying hens; compared with the traditional method of rough grouping management according to the age, the individual metabolic rate index can more comprehensively and dynamically reflect the actual energy demand difference caused by the body weight difference and the egg laying level difference; this gain effect makes the subsequent environmental temperature regulation more targeted, which can meet the real needs of individuals with different metabolic levels, thereby improving the uniformity of the group while avoiding energy waste or production performance loss caused by one-size-fits-all management.

[0062] Embodiment 3:

[0063] The solving process of the group aggregation entropy by the biological state quantification unit includes:

[0064] A plurality of grid regions are virtually divided in the breeding space; through an image analysis algorithm, the proportion of the number of chickens distributed in each grid region to the total number is recognized and counted to generate a regional distribution proportion; the regional distribution proportion is substituted into a preset information entropy formula to calculate the group aggregation entropy.

[0065] This embodiment is a specific implementation of the solving process of the group aggregation entropy in the biological state quantification unit of embodiment 1; the process aims to convert the descriptive image of the spatial distribution of the chicken group captured by the camera device into a quantitative numerical index that can indicate the overall comfort and stress level of the group;

[0066] To achieve this purpose, the solving process of the group aggregation entropy by the biological state quantification unit starts with virtually grid processing the two-dimensional plan view of the breeding space in the digital twin model of the system; the virtual grid processing refers to dividing the whole chicken coop bird's-eye view into a plurality of equal-sized, non-overlapping grid regions in the software layer, which aims to provide discrete spatial units for the subsequent counting of the number of chickens; the total number N of the grid is a parameter that needs to be preset, and its setting needs to balance the spatial resolution and the calculation efficiency;

[0067] Subsequently, the system processes the collected real-time image of the chicken group through a preset image analysis algorithm; the image analysis algorithm in this embodiment refers to a target recognition and counting model based on computer vision technology, such as YOLO or similar algorithm, which can accurately identify each chicken in the image and determine the grid area where it is located after training on chicken feature data; through this algorithm, the system can identify and count the number of chickens distributed in each grid area, and calculate the proportion of the number in the total number of the chicken group, to generate the regional distribution proportion of each grid ;

[0068] Finally, the system substitutes the regional distribution proportion of all grids into a preset information entropy formula to calculate the group aggregation entropy; the calculation process is defined by the group aggregation entropy model; the group aggregation entropy model is a model that borrows the concept of entropy in information theory to quantify the degree of disorder of the system; in this embodiment, the more uniform the distribution of the chicken group, the higher the degree of disorder and the higher the entropy value, representing a more comfortable and relaxed state of the chicken group; on the contrary, if the chicken group is aggregated in a local area, the distribution is more ordered, the entropy value is lower, and it may represent that there is environmental stress; the mathematical expression is:

[0069] ;

[0070] H is the group aggregation entropy, which is a dimensionless value used as an ordinal index in the system; its function is to quantify the uniformity of the spatial distribution of the chicken group, thereby indirectly evaluating the macro stress state of the group; the value is calculated by the formula;

[0071] N is the total number of grid areas, which is a dimensionless integer; its value is preset according to the size of the breeding space and the required accuracy during system initialization;

[0072] is the regional distribution proportion, which is the proportion of the number of chickens distributed in the i-th grid area to the total number, and is a dimensionless probability value; its data is derived from the statistical calculation results of the real-time image through the image analysis algorithm;

[0073] The gain effect brought by this embodiment is to provide a non-contact, low-cost and real-time group stress state evaluation method; traditional stress evaluation often relies on sampling to determine blood physiological indicators, which has a lag, stress interference and cannot cover the whole group; and the group aggregation entropy can capture signals (such as aggregation behavior caused by excessive local wind speed or unsuitable temperature) in the early stage of environmental stress through the quantification of macro behavior patterns, so that the system can make more proactive and active environmental intervention, effectively prevent large-scale stress events, and ensure animal welfare and production stability.

[0074] Example 4:

[0075] The process by which the time delay feature analysis unit generates the biological response time delay complexity index includes:

[0076] Obtain the time-varying curve of population aggregation entropy after historical environmental regulation events; determine the moment when the absolute value of the derivative of the time-varying curve is continuously lower than a preset threshold to calculate the lag time of the biological response; normalize the lag time to generate a dimensionless lag coefficient; and perform a weighted summation of the dimensionless lag coefficients to generate the biological response time lag complexity index.

[0077] This embodiment is a specific implementation of the biological response time delay complexity index generation process in the time delay feature analysis unit of Embodiment 1; the purpose of this process is to quantify the degree of sluggishness or viscousness exhibited by chicken flocks as a complex biological system when their behavior pattern recovers to a new steady state after being stimulated by the external environment.

[0078] To achieve this quantification process, the system has a built-in control event log module for recording key control events. This log is used when the control command generation unit outputs the temperature setpoint value. Compared with the current actual temperature absolute value of the difference Exceeding the preset trigger threshold (For example When the control command is executed continuously for at least one control cycle, the system marks this as an effective temperature control event and records the start time of the event, the control type (heating / cooling), and the relevant state parameters before and after the control.

[0079] The time-delay characteristic analysis unit is activated at a preset analysis period (e.g., every 24 hours) or after a sufficient number (e.g., 10) of similar events have accumulated in the event log, and performs retrospective analysis on the historical data in the log. This unit first filters out specific types of regulatory events (such as temperature regulation) from the log, and extracts the time change curve of the population aggregation entropy H recorded subsequently.

[0080] Subsequently, in order to objectively determine the moment when the flock's behavior reaches a new steady state, the system calculates the derivative of the time-varying curve. The absolute value; a preset threshold Introduced here, it is an extremely small positive number. Its technical principle is that after reaching a steady state, the population aggregation entropy fluctuates naturally within a very small range, at which point its rate of change should approach zero. The specific value of this threshold can be set based on statistical analysis of a large amount of time-series data on the population aggregation entropy of chickens under normal conditions, taking the 95th or 99th percentile of the absolute value of the rate of change to ensure that only statistically significant changes are considered the end of the state transition. Therefore, the system continuously monitors... When the value is always below the preset threshold for a preset steady state judgment duration , it is determined that the chicken behavior has reached a new steady state. The steady state judgment duration is a system preset parameter, and its value should be greater than a control period, for example, it can be set to 3 control periods in total to ensure that the observed stability is not a random transient phenomenon; the time elapsed from the moment when the environmental regulation instruction starts to be executed to this steady state moment is calculated as the biological response lag time;

[0081] This process is performed for different types of environmental disturbances (such as temperature, humidity, air quality), and the respective lag times are obtained; then, the lag times are normalized, and the normalization aims to eliminate the magnitude difference of the lag times caused by different disturbance events and convert them into unified and dimensionless lag coefficients;

[0082] Specifically, the normalization can use the maximum-minimum normalization method, and its calculation formula is:

[0083] ;

[0084] is the dimensionless lag coefficient corresponding to a specific disturbance (such as temperature, humidity, or air quality);

[0085] is the biological response lag time measured this time under the disturbance;

[0086] and are preset reference values of the maximum and minimum lag times that the type of disturbance can cause, which are statistically obtained from a large amount of historical data. Through this method, lag times of different magnitudes can be uniformly mapped into the interval of 0 to 1;

[0087] To enable those skilled in the art to reproduce, a statistical calibration method for determining and is provided: again, the effective regulation event logs accumulated in the system initialization calibration period or the early running stage (such as the first 30 days) are used. From the logs, a data set of biological response lag times of the same type of disturbance (such as temperature disturbance) is extracted. After statistical analysis of the data set and removing outliers, the 95th percentile of the data set is set as , and the 5th percentile is set as . These two values should also be updated regularly based on a larger amount of historical data as the system runs for a long time.

[0088] Finally, these dimensionless lag coefficients are weighted and summed to generate a comprehensive biological response time lag complexity index; this calculation is defined by a preset biological response time lag complexity index model; the biological response time lag complexity index model is a linear weighted model designed to integrate the response delay of multiple environmental factors; its mathematical expression is:

[0089] ;

[0090] L is the biological response time lag complexity index, which is a dimensionless comprehensive index; its role is to quantify the overall response sensitivity of the current chicken population to environmental changes;

[0091] is a dimensionless lag coefficient, which represents the normalized response lag time caused by temperature, humidity and air quality disturbance respectively; its data comes from the results of standardizing the corresponding lag time in historical data;

[0092] is the preset weight, which is the dimensionless weight corresponding to the above three lag coefficients respectively; its role is to represent the importance of different environmental factors on the overall state of the chicken population; its initial value is set according to the general recognition of the threat of each environmental factor to the health of the chicken population, for example, the weight of air quality is usually given a higher value; the sum of these weights is 1, and can be adjusted through the model self-optimization unit;

[0093] The innovation of this embodiment is that it no longer regards the chicken population as an ideal object that responds instantaneously to the control instruction, but for the first time quantifies its biological inertia; the resulting gain effect is that the system can adjust the aggressiveness of its control strategy according to the complexity index L; when the chicken population is in a complex stress state and shows a high response time lag (high L value), the system will adopt a more gentle and gradual control method, effectively avoiding the dramatic oscillation of environmental parameters caused by excessive control, thereby improving the robustness and stability of the entire closed-loop control system.

[0094] Example 5:

[0095] The calculation process of the temperature demand reference value by the control instruction generation unit includes:

[0096] Obtain the preset standard temperature value, ideal metabolic rate index and ideal aggregation entropy;

[0097] Calculate the deviation of the current average metabolic rate index from the ideal metabolic rate index, and generate a metabolic rate temperature compensation amount in combination with the preset temperature sensitivity coefficient;

[0098] Calculate the deviation of the current population aggregation entropy from the ideal aggregation entropy, and generate an aggregation entropy temperature compensation amount in combination with the preset entropy sensitivity coefficient;​

[0099] The standard temperature value is corrected by using the metabolic rate temperature compensation amount and the aggregation entropy temperature compensation amount to generate a temperature demand reference value.

[0100] This embodiment is a specific implementation of the temperature demand reference value calculation process in the regulation instruction generation unit of embodiment 1. The core purpose of this calculation process is to make individualized and dynamic corrections based on the recommended temperature of the industry standard breeding guide and according to the real-time biological state of the current flock, so as to obtain a theoretically optimal temperature value.

[0101] To achieve this calculation, the regulation instruction generation unit needs to obtain a series of preset reference parameters first in the calculation process of the temperature demand reference value. The standard temperature value is obtained by consulting the recommended environmental temperature from authoritative breeding management guidelines according to the age stage of the laying hens. The ideal metabolic rate index and the ideal aggregation entropy are determined by finding the statistical average values of the metabolic rate and the aggregation entropy associated with the highest egg production rate and the lowest culling rate under strictly controlled experimental conditions. They represent the biological state indicators of the flock under the best production performance.

[0102] To enable skilled personnel to reproduce, a system initialization calibration method for determining the ideal biological state indicators and is provided: at the initial stage of system deployment, a calibration period of days (for example, 14 days) is set. During this period, the environmental control adopts traditional static breeding standards based on age, and the system continuously collects and records the physiological information of all laying hens and the flock aggregation entropy. After the calibration period ends, the day or days with the best overall production performance (such as the highest average egg production rate and the lowest culling rate) during this period are analyzed. The statistical average values of the average individual metabolic rate index and the average flock aggregation entropy measured during these days are used as the preset ideal metabolic rate index and the ideal aggregation entropy .

[0103] Next, the system compares the real-time feedback biological state indicators with the ideal values, calculates the deviation, and generates the corresponding temperature compensation amount combined with the preset sensitivity coefficient. Specifically, it includes:

[0104] Calculate the deviation between the average metabolic rate index of the current flock and the ideal metabolic rate index , and generate the metabolic rate temperature compensation amount by combining the deviation with a preset temperature sensitivity coefficient .

[0105] Similarly, calculate the current group aggregation entropy H and the ideal aggregation entropy. The deviation between them, combined with the preset entropy sensitivity coefficient This generates the temperature compensation amount for the aggregation entropy.

[0106] Finally, the system uses these two calculated temperature compensation values ​​to double-correct the standard temperature value to generate the final temperature requirement benchmark value. This calculation logic is defined by a preset temperature requirement benchmark value model. This model is a feedback-based control model, whose physical meaning is that when the actual state of the flock deviates from the optimal state, compensation is made by adjusting the temperature. Its mathematical expression is:

[0107] ;

[0108] It is the baseline value for temperature demand, in degrees Celsius (°C); its function is to provide a theoretical target for dynamic regulation in the next stage; this value is calculated by this formula.

[0109] It is a standard temperature value, in degrees Celsius (°C); its source is a pre-set industry standard or feeding guide.

[0110] These are indicators of the current biological state, namely the average metabolic rate index and the population aggregation entropy of the current flock, both of which are dimensionless; their data come from the real-time calculation results of the biological state quantification unit.

[0111] These are indicators of ideal biological state, namely the ideal metabolic rate index and the ideal aggregation entropy, both of which are dimensionless; their values ​​are derived from the optimal values ​​calibrated through previous experiments and preset in the system.

[0112] This is the sensitivity coefficient, and the unit is degrees Celsius (°C) to ensure that the physical dimensions on both sides of the formula are consistent. The physical meaning of is the amount of temperature compensation required for the average metabolic rate index to deviate from the ideal value by one unit. The physical meaning is similar; their initial values ​​are set based on thermodynamics and animal behavior knowledge, and can be adjusted through the model's self-optimizing unit;

[0113] Through this embodiment, the core basis of environmental regulation is changed from static and universal feeding standards to dynamic and targeted biological needs. The resulting gain effect is to greatly improve the fine level of environmental management. For example, when the average metabolic rate of the chicken flock is higher than the ideal value (possibly due to high feed intake and high activity), the system will automatically calculate a slightly lower temperature demand baseline value to help dissipate heat. When the flock aggregation entropy is lower than the ideal value (aggregation behavior occurs), the system will calculate a higher temperature demand baseline value to alleviate cold stress. This real-time, feedback-based correction ensures accurate matching of environmental supply and biological needs, thereby improving energy utilization efficiency and animal production performance.

[0114] Embodiment 6:

[0115] The generation process of the temperature setting target value by the regulation instruction generation unit includes:

[0116] The biological response time lag complexity index is substituted into a preset exponential decay model to generate a regulation intensity factor. The difference between the temperature demand baseline value and the actual temperature currently collected is calculated to generate a temperature target deviation. The temperature target deviation is multiplied by the regulation intensity factor, and the actual temperature is summed to generate a temperature setting target value for the next control time step.

[0117] This embodiment is a specific implementation of the final temperature setting target value generation process in the regulation instruction generation unit. After calculating the theoretical temperature demand baseline value, an actual executable specific temperature setting value for the next control time step is generated in combination with the response characteristics of the system. It embodies a prudent and intelligent flexible control strategy.

[0118] This process first introduces a regulation intensity factor The purpose is to dynamically adjust the amplitude of the control action according to the degree of sluggishness of the biological system response. This factor is generated by a preset exponential decay model based on the biological response time lag complexity index L calculated in the previous step. The technical principle of the exponential decay model is that when the system response is complex and sluggish (L value is high), the regulation intensity should be significantly reduced to avoid excessive intervention. Its mathematical expression is:

[0119]

[0120] is the regulation intensity factor, which is a dimensionless factor between 0 and 1. Its role is to act as a scaling factor for the regulation step size.

[0121] L is the biological response time lag complexity index, which is a dimensionless input value from the time lag characteristic analysis unit.

[0122] ​k is the attenuation constant, a positive dimensionless number; its role is to adjust the degree of influence of L value on the control intensity; its value is determined by regression analysis on historical control effect data, and can be updated by the model self-optimization unit;

[0123] After obtaining the control intensity factor, the system calculates the temperature demand reference value and the difference between the actual temperature collected by the current sensor , which is defined as the temperature target deviation;

[0124] Finally, multiply the temperature target deviation by the control intensity factor to get the actual adjustment amount, and sum it with the current actual temperature to generate the temperature set target value of the next control time step; the calculation process is defined by the final control instruction model:

[0125] ;

[0126] is the temperature set target value, unit: Celsius (℃); this is the final output of this unit, which will be sent to the underlying environmental control hardware (such as air conditioner, heater) for execution instruction;

[0127] is the actual temperature, unit: Celsius (℃); its data comes from the real-time reading of the temperature sensor in the data acquisition unit;

[0128] is the control intensity factor, which is a dimensionless value calculated in the previous step;

[0129] is the temperature demand reference value, unit: Celsius (℃);

[0130] The technical scheme of this embodiment integrates the biological response time lag characteristic and the dynamic demand reference, thereby realizing a synergistic gain effect; it not only determines the final target of the control (determined by ), but more importantly determines the process way to achieve the target (determined by and L); its technical effect is to realize a gradual and flexible compensation strategy; when the chicken population is stable and responds quickly (L value is low, tending to 1), the system will quickly approach the ideal temperature , ensuring the control efficiency; when the chicken population is fragile and responds slowly (L value is high, tending to 0), the system only makes small changes in a gradual adjustment manner based on the current temperature, thereby effectively avoiding the secondary stress on the chicken population caused by environmental mutations; this adaptive control rhythm greatly enhances the safety and stability of the system.

[0131] Example 7:

[0132] The prediction process of the model self-optimization unit for the next time biological state is as follows:

[0133] A state transition prediction model is constructed; the individual metabolic rate index and the population aggregation entropy observed at the current time, and the temperature setting target value generated by the control instruction generation unit are taken as inputs; the state transition prediction model is operated to output the predicted values of the individual metabolic rate index and the population aggregation entropy at the next time.

[0134] This embodiment is a specific implementation of the prediction process of the model self-optimization unit in Example 1 for the next time biological state; the prediction function is the logical premise of the whole system to be able to self-optimize and closed-loop learning, and its purpose is to predict the biological consequences that the decision may bring based on the current known state and the control decision to be executed;

[0135] To realize this prediction, a state transition prediction model is constructed in the system; the core function of this model is to fit the complex dynamic relationship between environmental input (temperature regulation) and biological state output (change of metabolic rate and aggregation entropy); this model can be trained based on historical data, and the specific implementation can use a multiple linear regression model, or a more complex machine learning model such as a recurrent neural network (RNN), which is particularly good at processing time series data;

[0136] The inputs of this model include three key variables: the average individual metabolic rate index of the chicken population observed at the current time t and the population aggregation entropy , and the temperature setting target value just calculated by the control instruction generation unit, which will be executed at the next time step.

[0137] Substitute these three input variables into the state transition prediction model for operation, and the model will output a set of predicted values, i.e. the predicted values of the individual metabolic rate index and the population aggregation entropy at the next time ; this prediction process can be expressed by the following function relationship:

[0138] ;

[0139] is the predicted biological state, and is the dimensionless predicted value of the individual metabolic rate index and the population aggregation entropy at the next time by the model, respectively;

[0140] is the current biological state, which is the dimensionless biological state index actually observed at the current time t, provided by the biological state quantification unit;

[0141] is the control instruction, is the current calculated temperature setting target value (℃) to be executed at the next time step, provided by the control instruction generation unit;

[0142] is the state transition prediction function, representing the specific mathematical or algorithmic structure of the state transition prediction model, used to calculate the output;

[0143] In a preferred embodiment, the state transition prediction function and can be constructed using a multiple linear regression model, whose mathematical expression is:

[0144] ;

[0145] and is the predicted value;

[0146] and are model inputs;

[0147] are trainable parameters of the model, whose initial values are determined by regression analysis on historical data and are iteratively updated in the subsequent model self-optimization unit;

[0148] is a dimensionless parameter, while the coefficient has the dimension of the inverse of temperature (e.g., 1 / ℃) to ensure dimensional consistency on both sides of the equation;

[0149] The gain effect of this embodiment is to give the system the ability of forward-looking prediction; without this prediction function, the system can only know the effect after control; through the built-in prediction model, the system has a reasonable expectation of the consequences of the decision at the same time; this provides accurate quantitative targets for subsequent error calculation and model parameter self-correction, which is the core technical link to realize the transition from simple open-loop control or lagging feedback control to truly predictive and adaptive control.

[0150] Example 8:

[0151] The iterative updating process of the model self-optimization unit includes:

[0152] After the control is executed, the individual metabolic rate index and the population aggregation entropy at the next time step are actually observed; the actual observed value is compared with the predicted value to calculate the weighted deviation and generate the prediction error; the prediction error is used as the loss function, and an optimization algorithm is used to minimize the loss function as the goal to periodically update all preset adjustable weights and coefficients within the system.

[0153] This embodiment is a specific implementation of the iterative update process in the model self-optimization unit; this process is the closed-loop endpoint of the entire digital twin system to achieve long-term autonomous learning, adapt to environmental changes and maintain high-precision control capabilities; its core lies in quantitatively comparing the model's predictive performance with the results in the real world, and using this deviation to drive the model's self-improvement in reverse.

[0154] The iterative update process is triggered after the control cycle is completed; the system first observes and calculates the next time step using the data acquisition unit and the biological state quantification unit. The true biological state, i.e., the metabolic rate index of the actual observed individual. And actual observed population aggregation entropy ;

[0155] The system then compares these two actual observations with the predicted values ​​generated by the model at the previous time step. and The predictions are compared, and a weighted bias is calculated to generate a comprehensive prediction error. This error is calculated using a pre-defined prediction error model. This prediction error model quantifies the accuracy of the model's predictions, and its output value is typically referred to as a loss function in machine learning. Its mathematical expression is:

[0156] ;

[0157] Prediction error is a dimensionless numerical value that represents the overall deviation between the model's prediction and the actual result.

[0158] These are actual observed values, namely the metabolic rate index and entropy value actually observed within the next time step after the regulation is implemented, both of which are dimensionless;

[0159] These are model predictions, corresponding values ​​predicted by the model based on control commands, and both are dimensionless.

[0160] It is the error weight, a dimensionless preset weight coefficient. Its function is to balance the importance of the two indicators, metabolic rate and aggregation entropy, in the total error calculation. Their sum does not have to be 1 and can be adjusted through the model's self-optimization unit.

[0161] Finally, the system uses this calculated prediction error As a loss function, and with a preset optimization algorithm, an iterative update is performed on all preset adjustable weights and coefficients in the system to minimize the loss function; the optimization algorithm usually uses gradient descent or its variants (such as Adam optimizer); the algorithm calculates the partial derivative (i.e. gradient) of the loss function with respect to all preset adjustable parameters in the system, and adjusts each parameter slightly in the opposite direction of its gradient. The adjustable parameters include all preset weights, coefficients, and non-physical constants defined in the embodiments of the present specification, such as the weights in the individual metabolic rate index model ; the weights in the biological response time lag complexity index model ; the sensitivity coefficient in the temperature demand reference value model ; the attenuation constant k in the regulation intensity factor model; and the error weight in the prediction error model , etc., so that the prediction error in the next iteration is expected to be reduced; this process is repeated periodically;

[0162] This iterative update mechanism is the core guarantee for the long-term robustness and high accuracy of the present application, which cooperates with the prediction function of the model self-optimization unit to form a complete learning closed loop; the significant technical effect brought by it is the adaptability of the system; as time goes on, no matter whether the chicken population enters a new growth cycle (physiological parameter changes), encounters sudden disease challenges (stress pattern changes), or the breeding environment changes seasonally, the system can automatically adjust its internal model parameters through continuous "prediction-verification-correction" cycles to continuously fit the latest biological-environmental dynamic relationship; this makes the technical solution of the present application different from static expert systems, and builds a dynamic model that can continuously iterate and optimize itself according to the measured data, so as to always maintain the effectiveness and advancement of the control strategy under various complex and changing actual breeding conditions.

[0163] The technical solution of the present application realizes a fundamental breakthrough in the perception, decision-making and control of the traditional breeding mode by building a digital twin-based layer breeding simulation system; its beneficial effects and technical progress are not a single improvement, but a complete and self-consistent closed loop that deeply integrates biology, information theory and control theory.

[0164] The prior art is usually limited to monitoring and passive control of macro-physical environmental parameters such as temperature and humidity in the breeding facility; this approach ignores the internal state and demand diversity of the biological body as the core of regulation, and essentially simplifies the complex biological population into a passive responder to the environment.

[0165] The present solution realizes a leap in cognition by introducing two core biological state quantification indicators:​

[0166] Firstly, the establishment of individual metabolic rate index, its physical meaning is to build a dynamic energy demand model for each individual layer, the key physiological information (body weight, egg production level, age) that directly affect energy consumption are dimensionless and linear superposition; This formula is not simply a numerical addition, but based on the basic principles of animal energy metabolism, different dimensions of physiological data are mapped to a scalar field that can represent individual energy demand; This makes the system evolve from the extensive management of the group average to the fine management of individual differences, providing a computable physical basis for meeting the unique environmental needs of different individuals at different life stages;

[0167] Secondly, the introduction of group aggregation entropy, its theoretical basis comes from the description of system disorder in statistical mechanics; It is innovatively used here to describe the macroscopic order parameter of the spatial distribution of the chicken flock; A uniform and loose distribution of the chicken flock corresponds to a high entropy state, which means that the system is in a lower energy and more stable comfortable state in physics; On the contrary, the aggregation behavior of the chicken flock leads to a sharp peak in the local area, the system entropy value decreases, which marks the formation of an ordered structure, and this ordered structure is often driven by external environmental stress (such as local low temperature or air flow), which is a direct macroscopic manifestation of group stress; Through this index, the system can detect the overall welfare level and stress state of the whole group in real time and non-invasively through non-contact image analysis, which is a great progress in real-time, globality and non-invasiveness compared to the traditional stress judgment method which relies on individual sampling detection.

[0168] The traditional environmental control logic is a "set-execution" mode based on static threshold, for example, turn on the heating when the temperature is lower than 20℃; This control method is rigid and lagging, and does not consider the response characteristics of the organism to the regulation action itself, which may become a new stressor;

[0169] This scheme constructs a multi-level adaptive compensation regulation mechanism, its progressiveness lies in:

[0170] Firstly, the regulation target is no longer fixed, but a dynamically generated temperature demand reference value; In essence, it is a feedback control law, which takes the industry standard temperature as the reference, uses the deviation between the current biological state ( ) and the ideal state ( ) as the error signal, and performs negative feedback correction through sensitivity coefficients and ; This ensures that the regulation target always closely follows the real biological needs of the chicken flock;

[0171] Secondly, the regulation process is flexible, and the inertia of the biological system is quantified; the construction of the biological response time delay complexity index L is a quantification of the response function characteristics of the biological population as a complex system; on this basis, the regulation strength factor is modeled as an exponential decay form; the physical connotation of this design is that when the system response is sluggish and the complexity is high (L value is large), tends to 0, the regulation behavior will become extremely gentle; the final temperature setting target value calculation formula is like an intelligent adjusted damper; when tends to 1 (fast system response), the system will quickly converge to the target when tends to 0 (slow system response), the adjustment step is extremely small each time, so that the optimal state is smoothly and stably reached without causing severe oscillation of the system; this is fundamentally different from the step-by-step and impact control of the prior art.

[0172] The effectiveness of any model-based system is based on the premise that the model accurately reflects the real world; the model in the prior art is usually static and does not change once it is established; however, the biological population itself is dynamically changing (growth, aging, disease), and the environment is also continuously changing, so the fixed model will inevitably lose accuracy over time;

[0173] The core progress of the present scheme lies in the design of the model self-optimization unit, which gives the system the ability to evolve; this mechanism predicts the consequences of the regulation behavior through a state transition prediction model, and after obtaining the real observation value , the prediction error between the two is calculated; this error is used as a loss function to periodically iteratively update all adjustable parameters (including metabolic rate model weights , temperature correction coefficients , time delay model weights , etc.) in the system through optimization algorithms (such as gradient descent);

[0174] This means that the present scheme does not construct a fixed expert system, but an intelligent agent that can continuously perform the "hypothesis-verification-learning" scientific cycle; it can continuously correct itself to make its internal digital twin model more and more close to the real dynamics of the physical entity, thereby ensuring the accuracy, robustness and adaptability to unknown changes of the system in the long-term running process; this is an advanced feature that the prior art does not have and is fundamentally different.

[0175] The above merely describes the preferred embodiments of the present application, and is not intended to limit the present application in other forms. Any person skilled in the art can make changes or modifications to the above disclosed technical contents into equivalent embodiments with equivalent changes, and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present application, without departing from the technical solution content of the present application, still belongs to the protection scope of the technical solution of the present application.

Claims

1. A digital twin-based simulation system for egg-laying hen farming, characterized in that, include: The data acquisition unit is used to acquire multidimensional state parameters in the breeding facility in real time and output them to the biological state quantification unit. The multidimensional state parameters include the physiological information of individual laying hens, the overall spatial distribution of the flock, and physical environmental parameters. The biological state quantification unit is used to calculate the individual metabolic rate index and the population aggregation entropy based on the physiological information and overall spatial distribution obtained by the data acquisition unit. The time-delay feature analysis unit is used to quantify and generate a biological response time-delay complexity index based on the change process of population aggregation entropy after environmental regulation in historical data. The regulation instruction generation unit is used to calculate the temperature demand baseline value based on the individual metabolic rate index and population aggregation entropy generated by the biological state quantification unit, and further generate the final temperature setting target value by combining the biological response time delay complexity index generated by the time delay characteristic analysis unit. The model self-optimization unit is used to predict the biological state at the next moment based on the target temperature setting and the current biological state, compare it with the actual observed state after the regulation is implemented, generate prediction error, and iteratively update the preset weights and coefficients in the system based on the prediction error. The biological state quantification unit's process for calculating the population aggregation entropy includes: Multiple grid areas are virtually divided within the breeding space; image analysis algorithms are used to identify and count the proportion of chickens distributed in each grid area to the total number of chickens, in order to generate the regional distribution ratio; the regional distribution ratio is then substituted into a preset information entropy formula for calculation to generate the group aggregation entropy. The process by which the time delay feature analysis unit generates the biological response time delay complexity index includes: Obtain the time change curve of population aggregation entropy after historical environmental regulation events; determine the moment when the absolute value of the derivative of the time change curve is continuously lower than a preset threshold to calculate the lag time of the biological response; normalize the lag time to generate a dimensionless lag coefficient; and perform a weighted summation of the dimensionless lag coefficients to generate the biological response time lag complexity index. Its mathematical expression is: ； It is a biological response time-delay complexity index, a dimensionless comprehensive index; its function is to quantify the overall response sensitivity of the current chicken flock to environmental changes. These are dimensionless lag coefficients, representing the normalized response lag time caused by temperature, humidity, and air quality disturbances, respectively; their data are derived from the results of standardizing the corresponding lag times in historical data. The process by which the control command generation unit generates the target temperature value includes: The biological response time delay complexity index is substituted into the preset exponential decay model to generate the control intensity factor; the difference between the temperature demand baseline value and the currently collected actual temperature is calculated to generate the temperature target deviation; the temperature target deviation is multiplied by the control intensity factor and summed with the actual temperature to generate the temperature set target value for the next control time step. This calculation process is defined by the final control command model: ； This is the temperature target value, in degrees Celsius (°C); this is the final output of this unit, which will be sent to the underlying environmental control hardware (such as air conditioners and heaters). This is the actual temperature, expressed in degrees Celsius (°C); the data comes from the real-time readings of the temperature sensor in the data acquisition unit. It is the regulation intensity factor, which is the dimensionless value calculated in the previous step; It is the baseline value for temperature requirements, and the unit is degrees Celsius (°C).

2. The egg-laying hen farming simulation system based on digital twins according to claim 1, characterized in that, The biological state quantification unit's calculation process for the individual metabolic rate index includes: The collected data on body weight, egg production rate, and age were standardized to generate dimensionless physiological index values. The dimensionless physiological index values ​​were then weighted and summed with their respective preset weights to generate an individual metabolic rate index.

3. The egg-laying hen farming simulation system based on digital twins according to claim 1, characterized in that, The calculation process of the temperature requirement baseline value by the control command generation unit includes: Obtain the preset standard temperature value, ideal metabolic rate index, and ideal aggregation entropy; Calculate the deviation between the current average metabolic rate index and the ideal metabolic rate index, and generate the metabolic rate temperature compensation amount by combining it with the preset temperature sensitivity coefficient. Calculate the deviation between the current population aggregation entropy and the ideal aggregation entropy, and combine it with the preset entropy sensitivity coefficient to generate the aggregation entropy temperature compensation amount; The standard temperature value is corrected by using the metabolic rate temperature compensation and the aggregation entropy temperature compensation to generate the temperature requirement benchmark value.

4. The egg-laying hen farming simulation system based on digital twins according to claim 1, characterized in that, The process by which the model's self-optimizing unit predicts the biological state at the next moment is as follows: A state transition prediction model is constructed. The individual metabolic rate index and population aggregation entropy observed at the current moment, as well as the temperature target value generated by the control command generation unit, are used as inputs. The state transition prediction model is used to perform calculations and output the predicted values ​​of the individual metabolic rate index and population aggregation entropy at the next moment.

5. A digital twin-based egg-laying hen farming simulation system according to claim 1 or 4, characterized in that, The iterative update process of the model self-optimization unit includes: After the regulation is completed, the individual metabolic rate index and population aggregation entropy are observed at the next time step. The actual observed values ​​are compared with the predicted values, and the weighted bias is calculated to generate the prediction error. The prediction error is used as the loss function, and an optimization algorithm is used to periodically update all preset adjustable weights and coefficients in the system with the goal of minimizing the loss function.

Images