A method and device for simulating laying hen breeding system based on digital twin
Through digital twin technology and neural network model, the egg production and health status of laying hen farms are predicted, which solves the problem of difficulty in predicting returns in the existing technology and maximizes the economic benefits of the farms.
Patent Information
- Application Number
- CN202410641218.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-05-22
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-05-22
AI Technical Summary
The prior art is difficult to effectively predict the total income of laying hen farms, especially the balance between meat production and egg production, which makes it difficult to predict returns.
Using the simulation method of laying hen breeding system based on digital twins, the breeding farming system is established by establishing a neural network model, using multiple historical environmental data and health data to predict egg yield and health coefficients, thereby determining the predicted economic returns, and the farm is carried out based on the prediction results. Regulation.
Accurate prediction of the egg production and health status of laying hen farms has been achieved, and the economic benefits of the farms have been improved by balancing egg production and meat production.
Smart Images

Figure CN118484998B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of scientific breeding, and in particular to a method and device for simulating a laying hen breeding system based on digital twins. Background Art
[0002] With the continuous advancement of agricultural technology, laying hen breeding equipment is also constantly developing and improving. From traditional free-range breeding to modern industrial breeding, laying hen breeding equipment has undergone tremendous changes.
[0003] Modern laying hen breeding equipment has achieved comprehensive optimization from feed production, chicken house design, breeding management to disease prevention and control. First of all, modern breeding equipment uses scientifically formulated feed to ensure that laying hens obtain comprehensive and balanced nutrition. Secondly, the design of the chicken house fully considers the living habits and growth needs of laying hens, providing a comfortable living environment for laying hens. In addition, breeding management has been automated and intelligent, and the growth status of laying hens can be understood in real time through monitoring equipment, and the feeding strategy can be adjusted in time. Finally, disease prevention and control measures have been greatly strengthened, effectively reducing the incidence of diseases in laying hens.
[0004] The main income of a farm includes meat production and egg production. The output of these two parts changes with the changes in the farm environment and has a certain correlation. It is difficult to predict them at the same time. Therefore, how to predict the total income of a farm is an urgent problem to be solved. Summary of the invention
[0005] The embodiments of the present application provide a method and device for simulating a laying hen breeding system based on digital twins to improve the above-mentioned problems.
[0006] In order to achieve the above objectives, this application adopts the following technical solutions:
[0007] In the first aspect, the embodiment of the present application proposes a method for simulating a laying hen breeding system based on digital twins, the method comprising:
[0008] Determine the current breeding situation of the target farm, wherein the current breeding situation includes a precise feeding plan, an intelligent condition detection plan, and a precise feeding intelligent solution based on the digital twin system, wherein the precise feeding plan includes regulating the target farm based on the digital twin system.
[0009] In conjunction with the first aspect, the method includes:
[0010] Acquire multiple historical environmental data groups and multiple historical egg production data of the target farm in multiple breeding cycles, the historical environmental data groups including temperature, humidity and breeding density, and acquire a first training set based on the multiple historical environmental data groups and the multiple historical egg production data, wherein one historical environmental data group corresponds to one group of historical environmental data groups;
[0011] Establishing a first neural network training model, importing the first training set into the first neural network training model and performing iterative training, wherein the first neural network training model is used to output predicted egg production data based on the input environmental data set;
[0012] Acquire a plurality of historical health data, and acquire a plurality of historical health coefficients based on the historical health data, wherein the historical health data is used to characterize the overall health status of all breeding individuals in the target breeding farm, and acquire a second training set based on the plurality of historical health coefficients and the plurality of historical environmental data groups;
[0013] Establishing a second neural network training model, importing the second training set into the second neural network training model and performing iterative training, the second neural network training model is used to output a predicted health coefficient according to the input environmental data group;
[0014] Based on different data gradients, a plurality of different environmental data groups are generated, and the plurality of environmental data groups are input into a first neural network training model and a second neural network training model, and a plurality of predicted egg production data and a plurality of predicted health coefficients are obtained according to output results of the first neural network training model and the second neural network training model;
[0015] Based on the predicted egg production data and the predicted health coefficient, a plurality of predicted economic benefit data are obtained, and one of the plurality of predicted economic benefit data is determined as the target economic benefit data;
[0016] The target farm is regulated according to the environmental data set corresponding to the target economic benefit.
[0017] In combination with the first aspect, in some embodiments, a first neural network training model is established, the first training set is imported into the first neural network training model and iterative training is performed, and the first neural network training model is used to output predicted egg production data based on the input environmental data group, including:
[0018] Obtaining predicted egg production data based on the first neural network training model;
[0019] Determining historical environmental data corresponding to the predicted egg production data based on the data input into the first neural network training model;
[0020] Compare the input environmental data set with the historical environmental data set and obtain a first error value;
[0021] The first error value is compared with the first preset value. If the first error value is less than or equal to the error value, the training is stopped and the first neural network training model is output.
[0022] In combination with the first aspect, in some implementations, comparing the input environment data set with the historical environment data set and obtaining a first error value includes:
[0023] Determine a first reference value of the input data based on the input environmental data set, satisfying:
[0024]
[0025] in, The first reference value of the input data, Enter the value of the temperature, is the input humidity value, is the input stocking density value, , , is the quantization constant;
[0026] First reference value based on input data The second reference value of the corresponding historical environmental data set Get a first error value.
[0027] In combination with the first aspect, in some implementations, based on a first reference value of the input data The second reference value of the corresponding historical environmental data set Obtaining a first error value includes:
[0028] The first reference value With the second reference value Compare, if the first benchmark value Greater than the second reference value , then the first error value and the first reference value With the second reference value The relationship between them satisfies:
[0029] =( - )
[0030] Among them, the first error value is .
[0031] In combination with the first aspect, in some implementations, based on a first reference value of the input data The second reference value of the corresponding historical environmental data set Obtaining a first error value includes:
[0032] The first reference value With the second reference value Compare, if the first benchmark value Less than the second reference value , then the first error value and the first reference value With the second reference value The relationship between them satisfies:
[0033] =( - )
[0034] Among them, the first error value is .
[0035] In combination with the first aspect, in some embodiments, a plurality of historical health data are obtained, and a plurality of historical health coefficients are obtained based on the historical health data, wherein the historical health data are used to characterize the overall health status of all farmed individuals in the target farm, and a second training set is obtained based on the plurality of historical health coefficients and a plurality of historical environmental data groups, including:
[0036] Determine a target breeding cycle and a breeding cycle before the target breeding cycle from a plurality of breeding cycles, and obtain first image data of the target breeding cycle and second image data of a breeding cycle before the target breeding cycle;
[0037] A historical health data corresponding to the target period is acquired based on the first image data and the second image data.
[0038] In combination with the first aspect, in some implementations, acquiring historical health data corresponding to a target period based on the first image data and the second image data includes:
[0039] Acquire a first body data group based on the first image data, the first body data group including body length data, fur color data and motion amplitude data, and acquire a second body data group based on the second image data;
[0040] Determine the historical health data corresponding to the target period based on the first body data group and the second body data group, and the historical health data satisfy:
[0041]
[0042] in, is the weight corresponding to the change in body length data of different individuals, is the weight corresponding to the change in fur color data of different individuals, is the weight corresponding to the change in the movement amplitude data of different individuals, Indicates total Change in item length data; Indicates total Change in fur color data; Indicates total The weight corresponding to the change in the action amplitude data.
[0043] In combination with the first aspect, in some implementations, a second neural network training model is established, the second training set is imported into the second neural network training model and iterative training is performed, and the second neural network training model is used to output a predicted health coefficient according to the input environmental data group, including:
[0044] Obtaining a health coefficient based on a second neural network training model;
[0045] Determining historical environmental data corresponding to the health coefficient based on data input into the second neural network training model;
[0046] Compare the input environmental data set with the historical environmental data set and obtain a second error value;
[0047] The second error value is compared with the second preset value. If the second error value is less than or equal to the error value, the training is stopped and the second neural network training model is output.
[0048] In combination with the first aspect, in some embodiments, based on the predicted egg production data and the predicted health coefficient, a plurality of predicted economic benefit data are obtained, and one of the plurality of predicted economic benefit data is determined as the target economic benefit data, including:
[0049] Obtaining predicted meat production data based on predicted health coefficients;
[0050] The predicted economic benefit data is determined based on the predicted meat production data and the predicted egg production data.
[0051] The second aspect of the embodiment of the present application proposes a laying hen breeding system simulation device based on digital twins, and the device is configured as follows:
[0052] Acquire multiple historical environmental data groups and multiple historical egg production data of the target farm in multiple breeding cycles, the historical environmental data groups including temperature, humidity and breeding density, and acquire a first training set based on the multiple historical environmental data groups and the multiple historical egg production data, wherein one historical environmental data group corresponds to one group of historical environmental data groups;
[0053] Establishing a first neural network training model, importing the first training set into the first neural network training model and performing iterative training, wherein the first neural network training model is used to output predicted egg production data based on the input environmental data set;
[0054] Acquire a plurality of historical health data, and acquire a plurality of historical health coefficients based on the historical health data, wherein the historical health data is used to characterize the overall health status of all breeding individuals in the target breeding farm, and acquire a second training set based on the plurality of historical health coefficients and the plurality of historical environmental data groups;
[0055] Establishing a second neural network training model, importing the second training set into the second neural network training model and performing iterative training, the second neural network training model is used to output a predicted health coefficient according to the input environmental data group;
[0056] Based on different data gradients, a plurality of different environmental data groups are generated, and the plurality of environmental data groups are input into a first neural network training model and a second neural network training model, and a plurality of predicted egg production data and a plurality of predicted health coefficients are obtained according to output results of the first neural network training model and the second neural network training model;
[0057] Based on the predicted egg production data and the predicted health coefficient, a plurality of predicted economic benefit data are obtained, and one of the plurality of predicted economic benefit data is determined as the target economic benefit data;
[0058] The target farm is regulated according to the environmental data set corresponding to the target economic benefit.
[0059] In conjunction with the second aspect, in some embodiments, the apparatus is configured to:
[0060] A first neural network training model is established, the first training set is imported into the first neural network training model and iterative training is performed, and the first neural network training model is used to output predicted egg production data according to the input environmental data group, including:
[0061] Obtaining predicted egg production data based on the first neural network training model;
[0062] Determining historical environmental data corresponding to the predicted egg production data based on the data input into the first neural network training model;
[0063] Compare the input environmental data set with the historical environmental data set and obtain a first error value;
[0064] The first error value is compared with the first preset value. If the first error value is less than or equal to the error value, the training is stopped and the first neural network training model is output.
[0065] In conjunction with the second aspect, in some embodiments, the apparatus is configured to:
[0066] Comparing the input environmental data set with the historical environmental data set and obtaining a first error value includes:
[0067] Determine a first reference value of the input data based on the input environmental data set, satisfying:
[0068]
[0069] in, The first reference value of the input data, Enter the value of the temperature, is the input humidity value, is the input stocking density value, , , is the quantization constant;
[0070] First reference value based on input data The second reference value of the corresponding historical environmental data set Get a first error value.
[0071] In conjunction with the second aspect, in some embodiments, the apparatus is configured to:
[0072] First reference value based on input data The second reference value of the corresponding historical environmental data set Obtaining a first error value includes:
[0073] The first reference value With the second reference value Compare, if the first benchmark value Greater than the second reference value , then the first error value and the first reference value With the second reference value The relationship between them satisfies:
[0074] =( - )
[0075] Among them, the first error value is .
[0076] In conjunction with the second aspect, in some embodiments, the apparatus is configured to:
[0077] First reference value based on input data The second reference value of the corresponding historical environmental data set Obtaining a first error value includes:
[0078] The first reference value With the second reference value Compare, if the first benchmark value Less than the second reference value , then the first error value and the first reference value With the second reference value The relationship between them satisfies:
[0079] =( - )
[0080] Among them, the first error value is .
[0081] In conjunction with the second aspect, in some embodiments, the apparatus is configured to:
[0082] Acquire multiple historical health data, and acquire multiple historical health coefficients based on the historical health data, wherein the historical health data is used to characterize the overall health status of all breeding individuals in the target farm, and acquire a second training set based on the multiple historical health coefficients and multiple historical environmental data groups, including:
[0083] Determine a target breeding cycle and a breeding cycle before the target breeding cycle from a plurality of breeding cycles, and obtain first image data of the target breeding cycle and second image data of a breeding cycle before the target breeding cycle;
[0084] A historical health data corresponding to the target period is acquired based on the first image data and the second image data.
[0085] In conjunction with the second aspect, in some embodiments, the apparatus is configured to:
[0086] Acquiring historical health data corresponding to a target period based on the first image data and the second image data includes:
[0087] Acquire a first body data group based on the first image data, the first body data group including body length data, fur color data and motion amplitude data, and acquire a second body data group based on the second image data;
[0088] Determine the historical health data corresponding to the target period based on the first body data group and the second body data group, and the historical health data satisfy:
[0089]
[0090] in, is the weight corresponding to the change in body length data of different individuals, is the weight corresponding to the change in fur color data of different individuals, is the weight corresponding to the change in the movement amplitude data of different individuals, Indicates total Change in item length data; Indicates total Change in fur color data; Indicates total The weight corresponding to the change in the action amplitude data.
[0091] In conjunction with the second aspect, in some embodiments, the apparatus is configured to:
[0092] A second neural network training model is established, and the second training set is imported into the second neural network training model and iterative training is performed. The second neural network training model is used to output a predicted health coefficient according to the input environmental data group, including:
[0093] Obtaining a health coefficient based on a second neural network training model;
[0094] Determining historical environmental data corresponding to the health coefficient based on data input into the second neural network training model;
[0095] Compare the input environmental data set with the historical environmental data set and obtain a second error value;
[0096] The second error value is compared with the second preset value. If the second error value is less than or equal to the error value, the training is stopped and the second neural network training model is output.
[0097] In conjunction with the second aspect, in some embodiments, the apparatus is configured to:
[0098] Based on the predicted egg production data and the predicted health coefficient, a plurality of predicted economic benefit data are obtained, and one of the plurality of predicted economic benefit data is determined as the target economic benefit data, including:
[0099] Obtaining predicted meat production data based on predicted health coefficients;
[0100] The predicted economic benefit data is determined based on the predicted meat production data and the predicted egg production data.
[0101] A third aspect of an embodiment of the present invention provides an electronic device, the electronic device comprising:
[0102] At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method proposed in the first aspect of the embodiment of the present invention.
[0103] A fourth aspect of an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method provided in the first aspect of the embodiment of the present invention.
[0104] In summary, the above method and device have the following technical effects:
[0105] The present application proposes a method for simulating a laying hen breeding system based on digital twins. First, a first neural network model for outputting predicted egg production data is established through multiple historical environmental data groups and multiple historical egg production data. Then, a second neural network prediction model for outputting predicted health coefficients is established through multiple historical environmental data groups and multiple historical health data. Since the health coefficient is highly correlated with meat production, the predicted economic benefit data can be determined by predicting meat production data and predicted egg production data. Finally, the target farm is regulated according to the environmental data group corresponding to the target economic benefit. The present application proposes a method for simulating a laying hen breeding system based on digital twins. By utilizing a neural network model, the egg production and health coefficient under different environmental data are reasonably predicted, and the predicted meat production is obtained through the health coefficient. According to the results of the prediction of different environmental data, the egg production and meat production are balanced to achieve the highest economic benefit for the farm. BRIEF DESCRIPTION OF THE DRAWINGS
[0106] Figure 1 A schematic flow chart of a method for simulating a laying hen farming system based on digital twins proposed in an embodiment of the present application. DETAILED DESCRIPTION
[0107] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0108] The present application embodiment proposes a method for simulating a laying hen breeding system based on digital twins, and the method includes:
[0109] Determine the current breeding situation of the target farm, wherein the current breeding situation includes a precise feeding plan, an intelligent condition detection plan, and a precise feeding intelligent solution based on the digital twin system, wherein the precise feeding plan includes regulating the target farm based on the digital twin system.
[0110] For details, please refer to Figure 1 , including the following steps:
[0111] S101: Obtain multiple historical environmental data groups and multiple historical egg production data of the target farm in multiple breeding cycles, the historical environmental data groups including temperature, humidity and breeding density, and obtain a first training set based on the multiple historical environmental data groups and the multiple historical egg production data, wherein one historical environmental data group corresponds to a group of historical environmental data groups.
[0112] It is understandable that the impact of temperature and humidity on egg production is significant. Within the appropriate temperature and humidity range, egg production can often reach a higher level. And breeding density is also a factor that cannot be ignored. Too high breeding density may lead to the deterioration of the breeding environment, which in turn affects egg production. In multiple breeding cycles, for the target farm, collecting historical environmental data sets and historical egg production data can be used as a basis for analyzing breeding efficiency and optimizing breeding methods. Specifically, the collected historical environmental data sets include key indicators such as temperature, humidity, and breeding density. In order to further study the relationship between these data and egg production, a neural network calculation method can be used. Therefore, these historical data are used as the first training set for subsequent training of the neural network. This training set is generated based on multiple historical environmental data sets and multiple historical egg production data. Each historical environmental data set corresponds to a set of historical egg production data. Such a pairing relationship can more accurately analyze the impact of environmental factors on egg production.
[0113] Specifically, in this embodiment, the collected historical data can be cleaned and preprocessed to remove outliers and erroneous data to ensure the accuracy and reliability of the data. Then, statistical analysis and machine learning algorithms are used to conduct in-depth mining and analysis of these data.
[0114] S102: Establish a first neural network training model, import the first training set into the first neural network training model and perform iterative training, the first neural network training model is used to output predicted egg production data based on the input environmental data group.
[0115] Specifically, first, the predicted egg production data can be obtained based on the first neural network training model, and then the historical environmental data corresponding to the predicted egg production data can be determined based on the data input into the first neural network training model.
[0116] Specifically, a fully connected neural network training model commonly used in deep learning can be used. This model can model the nonlinear relationship between input features and output values through calculations of multiple layers of neurons.
[0117] Specifically, as an implementation method, let the feature vector be x=[ , , ],in , and Represent the normalized temperature, humidity and breeding density respectively, and the neural network training model is assumed to be f( ),in Represents the model parameters so that the model output value f( ) can be close to the true value y of the egg production data. Then the training process of the neural network training model can be achieved by minimizing the loss function, that is:
[0118] L( ) =
[0119] Where N represents the number of training samples. and Represent the input features and output values of the i-th sample respectively. This loss function can measure the error between the model prediction value and the true value. The model parameters can be updated through optimization algorithms such as back propagation algorithm and gradient descent. , thereby improving the accuracy of the model.
[0120] In order to ensure the accuracy of the neural network model, after this, the input environmental data set can be compared with the historical environmental data set, and a first error value can be obtained. It can be understood that the larger the first error value, the lower the accuracy of the neural network training model. Therefore, a first error value can be set manually, and then the first error value can be compared with the first preset value. If the first error value is less than or equal to the error value, the training is stopped and the first neural network training model is output.
[0121] As for how to obtain the first error value, since the environment data array group contains multiple data, in this embodiment, a first benchmark value representing the volume of the entire environment data group and a second benchmark value representing the historical environment data group can be introduced to determine the error value from the difference between the benchmark values.
[0122] Specifically, in this embodiment, the first reference value of the input data may be determined based on the input environment data group, satisfying:
[0123]
[0124] in, The first reference value of the input data, Enter the value of the temperature, is the input humidity value, is the input stocking density value, , , is a quantization constant. Then based on the first reference value of the input data The second reference value of the corresponding historical environmental data set Get a first error value.
[0125] As an implementation method, the first reference value With the second reference value Compare, if the first benchmark value Greater than the second reference value , then the first error value and the first reference value With the second reference value The relationship between them satisfies:
[0126] =( - )
[0127] Among them, the first error value is .
[0128] Of course, if the first reference value Less than the second reference value , then the first error value and the first reference value With the second reference value The relationship between them satisfies:
[0129] =( - )
[0130] In this way, it can be ensured that the error value is always greater than 0. It is understandable that when the error value is less than the preset value, the training process of the neural network training model is stopped. This is because when the error value is less than a certain threshold, the model has predicted the data accurately enough, and continuing to train the model will not be effective and may cause problems such as overfitting.
[0131] S103: Obtain multiple historical health data, and obtain multiple historical health coefficients based on the historical health data, wherein the historical health data is used to characterize the overall health status of all farmed individuals in the target farm, and obtain a second training set based on the multiple historical health coefficients and multiple historical environmental data groups.
[0132] Understandably, in modern farming, in order to better understand the overall health of the farm and predict future health trends, a large amount of historical health data needs to be collected and analyzed.
[0133] Specifically, first, we need to collect the historical health data of all farmed individuals in the target farm. Then, we need to calculate the historical health coefficient based on these historical health data. The historical health data is used to characterize the overall health status of all farmed individuals in the target farm.
[0134] As an implementation method, an image recognition method can be used to obtain the health coefficient. How to use images for feature recognition has been disclosed in the relevant technology and will not be repeated here. As an implementation method, a target breeding cycle and a breeding cycle before the target breeding cycle can be determined from multiple breeding cycles, and the first image data of the target breeding cycle and the second image data of the breeding cycle before the target breeding cycle can be obtained, and then a historical health data corresponding to the target cycle can be obtained based on the first image data and the second image data.
[0135] Specifically, first, a first body data group can be obtained based on the first image data, the first body data group including body length data, fur color data and motion amplitude data, and a second body data group can be obtained based on the second image data. Then, historical health data corresponding to the target period can be determined based on the first body data group and the second body data group. The historical health data satisfy:
[0136]
[0137] in, is the weight corresponding to the change in body length data of different individuals, is the weight corresponding to the change in fur color data of different individuals, is the weight corresponding to the change in the movement amplitude data of different individuals, Indicates total Change in item length data; Indicates total Change in fur color data; Indicates total The weight corresponding to the change in the action amplitude data.
[0138] A second training set can be constructed through multiple historical health data to train the prediction model. This second training set can help us build a model that can predict the health trend of farmed individuals. By inputting current environmental data and health data of farmed individuals, this model can output a predicted health coefficient, thereby providing valuable information about future health trends.
[0139] S104: Establish a second neural network training model, import the second training set into the second neural network training model and perform iterative training, the second neural network training model is used to output a predicted health coefficient based on the input environmental data group.
[0140] It can be understood that the process of establishing the neural network training model is similar to step S102 and will not be described in detail here.
[0141] S105: Based on different data gradients, multiple different environmental data groups are generated, and the multiple environmental data groups are input into the first neural network training model and the second neural network training model, and multiple predicted egg production data and multiple predicted health coefficients are obtained according to the output results of the first neural network training model and the second neural network training model.
[0142] S106: Based on the predicted egg production data and the predicted health coefficient, a plurality of predicted economic benefit data are obtained, and one of the plurality of predicted economic benefit data is determined as the target economic benefit data.
[0143] S107: Regulate the target farm according to the environmental data group corresponding to the target economic benefit.
[0144] It is understandable that, since the health coefficient is highly correlated with meat production, according to relevant technical records, the health coefficient and meat production are positively correlated. Therefore, the predicted meat production data can be obtained based on the predicted health coefficient, and then the predicted economic benefit data can be determined based on the predicted meat production data and the predicted egg production data. By applying multiple sets of different data to obtain multiple predicted economic benefit data, the highest benefit solution can be selected to regulate the farm environment to maximize profits.
[0145] The present application proposes a method for simulating a laying hen breeding system based on digital twins. First, a first neural network model for outputting predicted egg production data is established through multiple historical environmental data groups and multiple historical egg production data. Then, a second neural network prediction model for outputting predicted health coefficients is established through multiple historical environmental data groups and multiple historical health data. Since the health coefficient is highly correlated with meat production, the predicted economic benefit data can be determined by predicting meat production data and predicted egg production data. Finally, the target farm is regulated according to the environmental data group corresponding to the target economic benefit. The present application proposes a method for simulating a laying hen breeding system based on digital twins. By utilizing a neural network model, the egg production and health coefficient under different environmental data are reasonably predicted, and the predicted meat production is obtained through the health coefficient. According to the results of the prediction of different environmental data, the egg production and meat production are balanced to achieve the highest economic benefit for the farm.
[0146] Based on the same inventive concept, the present application also proposes a laying hen breeding system simulation device based on digital twins, which is configured as follows:
[0147] Acquire multiple historical environmental data groups and multiple historical egg production data of the target farm in multiple breeding cycles, the historical environmental data groups including temperature, humidity and breeding density, and acquire a first training set based on the multiple historical environmental data groups and the multiple historical egg production data, wherein one historical environmental data group corresponds to one group of historical environmental data groups;
[0148] Establishing a first neural network training model, importing the first training set into the first neural network training model and performing iterative training, wherein the first neural network training model is used to output predicted egg production data based on the input environmental data set;
[0149] Acquire a plurality of historical health data, and acquire a plurality of historical health coefficients based on the historical health data, wherein the historical health data is used to characterize the overall health status of all breeding individuals in the target breeding farm, and acquire a second training set based on the plurality of historical health coefficients and the plurality of historical environmental data groups;
[0150] Establishing a second neural network training model, importing the second training set into the second neural network training model and performing iterative training, the second neural network training model is used to output a predicted health coefficient according to the input environmental data group;
[0151] Based on different data gradients, a plurality of different environmental data groups are generated, and the plurality of environmental data groups are input into a first neural network training model and a second neural network training model, and a plurality of predicted egg production data and a plurality of predicted health coefficients are obtained according to output results of the first neural network training model and the second neural network training model;
[0152] Based on the predicted egg production data and the predicted health coefficient, a plurality of predicted economic benefit data are obtained, and one of the plurality of predicted economic benefit data is determined as the target economic benefit data;
[0153] The target farm is regulated according to the environmental data set corresponding to the target economic benefit.
[0154] In some embodiments, the apparatus is configured to:
[0155] A first neural network training model is established, the first training set is imported into the first neural network training model and iterative training is performed, and the first neural network training model is used to output predicted egg production data according to the input environmental data group, including:
[0156] Obtaining predicted egg production data based on the first neural network training model;
[0157] Determining historical environmental data corresponding to the predicted egg production data based on the data input into the first neural network training model;
[0158] Compare the input environmental data set with the historical environmental data set and obtain a first error value;
[0159] The first error value is compared with the first preset value. If the first error value is less than or equal to the error value, the training is stopped and the first neural network training model is output.
[0160] In some embodiments, the apparatus is configured to:
[0161] Comparing the input environmental data set with the historical environmental data set and obtaining a first error value includes:
[0162] Determine a first reference value of the input data based on the input environmental data set, satisfying:
[0163]
[0164] in, The first reference value of the input data, Enter the value of the temperature, is the input humidity value, is the input stocking density value, , , is the quantization constant;
[0165] First reference value based on input data The second reference value of the corresponding historical environmental data set Get a first error value.
[0166] In some embodiments, the apparatus is configured to:
[0167] First reference value based on input data The second reference value of the corresponding historical environmental data set Obtaining a first error value includes:
[0168] The first reference value With the second reference value Compare, if the first benchmark value Greater than the second reference value , then the first error value and the first reference value With the second reference value The relationship between them satisfies:
[0169] =( - )
[0170] Among them, the first error value is .
[0171] In some embodiments, the apparatus is configured to:
[0172] First reference value based on input data The second reference value of the corresponding historical environmental data set Obtaining a first error value includes:
[0173] The first reference value With the second reference value Compare, if the first benchmark value Less than the second reference value , then the first error value and the first reference value With the second reference value The relationship between them satisfies:
[0174] =( - )
[0175] Among them, the first error value is .
[0176] In some embodiments, the apparatus is configured to:
[0177] Acquire multiple historical health data, and acquire multiple historical health coefficients based on the historical health data, wherein the historical health data is used to characterize the overall health status of all breeding individuals in the target farm, and acquire a second training set based on the multiple historical health coefficients and multiple historical environmental data groups, including:
[0178] Determine a target breeding cycle and a breeding cycle before the target breeding cycle from a plurality of breeding cycles, and obtain first image data of the target breeding cycle and second image data of a breeding cycle before the target breeding cycle;
[0179] A historical health data corresponding to the target period is acquired based on the first image data and the second image data.
[0180] In some embodiments, the apparatus is configured to:
[0181] Acquiring historical health data corresponding to a target period based on the first image data and the second image data includes:
[0182] Acquire a first body data group based on the first image data, the first body data group including body length data, fur color data and motion amplitude data, and acquire a second body data group based on the second image data;
[0183] Determine the historical health data corresponding to the target period based on the first body data group and the second body data group, and the historical health data satisfy:
[0184]
[0185] in, is the weight corresponding to the change in body length data of different individuals, is the weight corresponding to the change in fur color data of different individuals, is the weight corresponding to the change in the movement amplitude data of different individuals, Indicates total Change in item length data; Indicates total Change in fur color data; Indicates total The weight corresponding to the change in the action amplitude data.
[0186] In some embodiments, the apparatus is configured to:
[0187] A second neural network training model is established, and the second training set is imported into the second neural network training model and iterative training is performed. The second neural network training model is used to output a predicted health coefficient according to the input environmental data group, including:
[0188] Obtaining a health coefficient based on a second neural network training model;
[0189] Determining historical environmental data corresponding to the health coefficient based on data input into the second neural network training model;
[0190] Compare the input environmental data set with the historical environmental data set and obtain a second error value;
[0191] The second error value is compared with the second preset value. If the second error value is less than or equal to the error value, the training is stopped and the second neural network training model is output.
[0192] In some embodiments, the apparatus is configured to:
[0193] Based on the predicted egg production data and the predicted health coefficient, a plurality of predicted economic benefit data are obtained, and one of the plurality of predicted economic benefit data is determined as the target economic benefit data, including:
[0194] Obtaining predicted meat production data based on predicted health coefficients;
[0195] The predicted economic benefit data is determined based on the predicted meat production data and the predicted egg production data.
[0196] The present application proposes a simulation device for laying hen breeding system based on digital twins. First, a first neural network model for outputting predicted egg production data is established through multiple historical environmental data groups and multiple historical egg production data. Then, a second neural network prediction model for outputting predicted health coefficients is established through multiple historical environmental data groups and multiple historical health data. Since the health coefficient is highly correlated with meat production, the predicted economic benefit data can be determined by predicting meat production data and predicted egg production data. Finally, the target farm is regulated according to the environmental data group corresponding to the target economic benefit. The present application proposes a simulation device for laying hen breeding system based on digital twins. By using a neural network model, the egg production and health coefficient under different environmental data are reasonably predicted, and the predicted meat production is obtained through the health coefficient. According to the results of the prediction of different environmental data, the egg production and meat production are balanced to achieve the highest economic benefit for the farm.
[0197] Based on the same inventive concept, an embodiment of the present application further proposes an electronic device, the electronic device comprising:
[0198] At least one processor; and, a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the laying hen breeding system simulation method based on digital twins according to an embodiment of the present application.
[0199] In addition, to achieve the above-mentioned purpose, an embodiment of the present application also proposes a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the laying hen breeding system simulation method based on digital twins of the embodiment of the present application.
[0200] The following is a detailed introduction to the various components of electronic equipment:
[0201] The processor is the control center of the electronic device, which can be a processor or a general term for multiple processing elements. For example, the processor is one or more central processing units (CPUs), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention, such as one or more microprocessors (digital signal processors, DSPs), or one or more field programmable gate arrays (field programmable gate arrays, FPGAs).
[0202] Optionally, the processor can perform various functions of the electronic device by running or executing a software program stored in the memory, and calling data stored in the memory.
[0203] The memory is used to store the software program for executing the solution of the present invention, and the execution is controlled by the processor. The specific implementation method can refer to the above method embodiment, which will not be repeated here.
[0204] Optionally, the memory may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory may be integrated with the processor, or may exist independently and be coupled to the processor through an interface circuit of the electronic device, which is not specifically limited in the embodiments of the present invention.
[0205] A transceiver is used to communicate with a network device or a terminal device.
[0206] Optionally, the transceiver may include a receiver and a transmitter, wherein the receiver is used to implement a receiving function, and the transmitter is used to implement a sending function.
[0207] Optionally, the transceiver may be integrated with the processor, or may exist independently and be coupled to the processor via an interface circuit of the router, which is not specifically limited in the embodiment of the present invention.
[0208] In addition, the technical effects of the electronic device can refer to the technical effects of the data transmission method described in the above method embodiment, and will not be repeated here.
[0209] It should be understood that the processor in the embodiment of the present invention may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.
[0210] It should also be understood that the memory in the embodiments of the present invention may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct RAM bus RAM (DR RAM).
[0211] The above embodiments can be implemented in whole or in part by software, hardware (such as circuits), firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.
[0212] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.
[0213] In the present invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.
[0214] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.
[0215] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.
Claims
1. A method for simulating laying hen breeding system based on digital twin, characterized in that: The method comprises: Determine the current breeding situation of the target farm, wherein the current breeding situation includes a precise feeding plan, an intelligent condition detection plan, and a precise feeding intelligent solution based on the digital twin system, wherein the precise feeding plan includes regulating the target farm based on the digital twin system; Acquire multiple historical environmental data groups and multiple historical egg production data of the target breeding farm in multiple breeding cycles, wherein the historical environmental data groups include temperature, humidity and breeding density, and acquire a first training set based on the multiple historical environmental data groups and the multiple historical egg production data, wherein one historical environmental data group corresponds to one group of historical environmental data groups; Establishing a first neural network training model, importing the first training set into the first neural network training model and performing iterative training, wherein the first neural network training model is used to output predicted egg production data based on the input environmental data set; Acquire the predicted egg production data based on the first neural network training model; Determining the historical egg production environment data corresponding to the predicted egg production data based on the data input into the first neural network training model; Compare the input environment data group with the historical environment data group and obtain a first error value; The first error value is compared with a first preset value. If the first error value is less than or equal to the error value, the training is stopped and the first neural network training model is output; a plurality of historical health data are obtained, and a plurality of historical health coefficients are obtained based on the historical health data, wherein the historical health data is used to characterize the overall health status of all breeding individuals in the target breeding farm, and a second training set is obtained based on the plurality of historical health coefficients and the plurality of historical environmental data groups; Establishing a second neural network training model, importing the second training set into the second neural network training model and performing iterative training, wherein the second neural network training model is used to output a predicted health coefficient based on the input environmental data group; Based on different data gradients, a plurality of different environmental data groups are generated, and the plurality of environmental data groups are input into the first neural network training model and the second neural network training model, and a plurality of predicted egg production data and a plurality of predicted health coefficients are obtained according to output results of the first neural network training model and the second neural network training model; Based on the predicted egg production data and the predicted health coefficient, a plurality of predicted economic benefit data are obtained, and one of the plurality of predicted economic benefit data is determined as the target economic benefit data; The target breeding farm is regulated according to the environmental data group corresponding to the target economic benefit.
2. A method for simulating laying hen breeding systems based on digital twins according to claim 1, characterized in that: Comparing the input environment data set with the historical environment data set and obtaining a first error value includes: Determining a first reference value of input data based on the input environmental data group satisfies: ,in, the first reference value of the input data, Enter the temperature value, is the value of the humidity you enter, is the value of the stocking density entered, , , is the quantization constant; The first reference value based on the input data The second reference value of the historical environment data set corresponding to The first error value is obtained.
3. A method for simulating laying hen breeding system based on digital twin according to claim 2, characterized in that: The first reference value based on the input data The second reference value of the historical environment data set corresponding to Obtaining the first error value includes: The first reference value With the second reference value Compare, if the first reference value Greater than the second reference value , then the first error value and the first reference value With the second reference value The relationship between them satisfies: =( - ) , wherein the first error value is .
4. The method for simulating a laying hen breeding system based on digital twin according to claim 3, characterized in that: The first reference value based on the input data The second reference value of the historical environment data set corresponding to Obtaining the first error value includes: The first reference value With the second reference value If the first reference value Less than the second reference value , then the first error value and the first reference value With the second reference value The relationship between them satisfies: =( - ) , wherein the first error value is .
5. The method for simulating laying hen breeding system based on digital twin according to claim 1, characterized in that: Acquire a plurality of historical health data, and acquire a plurality of historical health coefficients based on the historical health data, wherein the historical health data is used to characterize the overall health status of all breeding individuals in the target breeding farm, and acquire a second training set based on the plurality of the historical health coefficients and the plurality of the historical environmental data groups, including: Determine a target breeding cycle and a breeding cycle before the target breeding cycle from the plurality of breeding cycles, and obtain first image data of the target breeding cycle and second image data of a breeding cycle before the target breeding cycle; The historical health data corresponding to the target breeding cycle is obtained based on the first image data and the second image data.
6. The method for simulating laying hen breeding system based on digital twin according to claim 5, characterized in that: Acquiring the historical health data corresponding to the target breeding cycle based on the first image data and the second image data includes: Acquire a first body data group based on the first image data, the first body data group including body length data, fur color data and motion amplitude data, and acquire a second body data group based on the second image data; The historical health data corresponding to the target breeding cycle is determined based on the first body data group and the second body data group, and the historical health data satisfy: in, is the weight corresponding to the change in body length data of different individuals, is the weight corresponding to the change in fur color data of different individuals, is the weight corresponding to the change in the movement amplitude data of different individuals, Indicates total Change in item length data; Indicates total Change in fur color data; Indicates total The weight corresponding to the change in the action amplitude data.
7. The method for simulating laying hen breeding system based on digital twin according to claim 1, characterized in that: Establishing a second neural network training model, importing the second training set into the second neural network training model and performing iterative training, wherein the second neural network training model is used to output a predicted health coefficient according to the input environmental data group, including: Acquire the health coefficient based on the second neural network training model; Determining the historical environmental data corresponding to the health coefficient according to the data input into the second neural network training model; Compare the input environment data group with the historical environment data group and obtain a second error value; The second error value is compared with a second preset value. If the second error value is less than or equal to the error value, the training is stopped and the second neural network training model is output.
8. The method for simulating laying hen breeding system based on digital twin according to claim 1, characterized in that: Based on the predicted egg production data and the predicted health coefficient, a plurality of predicted economic benefit data are obtained, and one of the plurality of predicted economic benefit data is determined as the target economic benefit data, including: Acquiring predicted meat production data based on the predicted health coefficient; The predicted economic benefit data is determined based on the predicted meat production data and the predicted egg production data.
