Chicken flock average weight estimation method and system based on multi-feature deep fusion learning

Abstract

The invention discloses a chicken flock average weight estimation method and system based on multi-feature deep fusion learning, and the method comprises the steps: collecting the chicken flock average weight through an intelligent collection device, processing the chicken flock average weight data and chicken flock related information through a data processing program, and storing the processed data into a database of a server; creating a chicken flock average weight estimation model based on historical chicken flock average weight data and chicken flock related priori knowledge, and predicting chicken flock weight; prediction is displayed through a data visualization module, data information is converted into graphs and charts which are easy to analyze, and therefore the trend and law of the data are more clearly informed. The accuracy of livestock and poultry weight prediction is effectively improved, and technical support is provided for making a chicken flock selling plan and a breeding plan of an enterprise.

Description

Technical Field

[0001] The present invention relates to the field of smart farming technology, and more specifically, to a method and system for estimating the average weight of a flock of chickens based on prior knowledge and multi-feature deep fusion learning. Background Art

[0002] In recent years, with the growth of global population and the improvement of living standards, livestock and poultry meat production has continued to grow, and livestock and poultry farming has become one of the core pillars of the agricultural economy. In the breeding process, livestock and poultry weight is one of the focuses of attention, which is directly related to the production performance and economic benefits of livestock and poultry farming. With the deep integration of new generation information technology and livestock and poultry farming, various automatic livestock and poultry weight prediction algorithms have been proposed. According to the different technical means used, livestock and poultry weight prediction algorithms can be roughly divided into growth curve-based methods, vision-based methods and time series-based methods.

[0003] The growth curve model has achieved good results in predicting livestock and poultry weight. However, in actual application scenarios, due to the sparsity, high dynamics and diversity of commercial livestock and poultry data, it is difficult for this method to capture the inherent correlation information in the data, so the prediction accuracy is limited. Vision-based livestock and poultry weight estimation avoids stress reactions and injuries to livestock and poultry, and significantly improves measurement efficiency. However, most of these methods focus on the extraction of visual features of broilers, ignoring the impact of differences in broiler species, age, gender, etc. on their weight. In addition, factors such as broiler posture changes and lighting environment will also affect the accuracy of weight estimation. Therefore, the development of a flock weight estimation method that integrates growth curve models and deep networks is of great significance to intelligent management and intelligent decision-making in the breeding industry. Summary of the invention

[0004] In order to solve the above technical problems, the present invention proposes a chicken group average weight estimation method and system based on multi-feature deep fusion learning. Through the multi-layer neural network structure, the deep curve model can automatically extract high-order features in the data, learn complex nonlinear relationships, and fit the growth curve function model parameters, thereby solving the problem of insufficient parameter fitting. The combination of the growth curve model and the deep learning network model is particularly suitable for improving the accuracy and timeliness of livestock and poultry weight estimation.

[0005] The first aspect of the present invention provides a method for estimating the average weight of a flock of chickens based on multi-feature deep fusion learning, comprising:

[0006] Collect the average weight data of the chickens to be tested and related information of the chickens, pre-process the collected data, and store the pre-processed data in the server database;

[0007] Building a flock weight estimation model based on historical average weight data and flock related information in a database, and performing low-order feature learning and high-order feature learning in the flock weight estimation model;

[0008] The relevant information of the flock of chickens is used as prior knowledge, and the parameters fitted by high-order feature learning and the prior knowledge are used to construct a growth curve function, and the predicted value of the average weight of the flock of chickens is output through the growth curve function;

[0009] The output average weight prediction value of the chicken flock is returned to the server, and the average weight prediction value of the chicken flock is visualized on the server.

[0010] In this scheme, the average weight data of the chicken flock to be tested and the relevant information of the chicken flock are collected, the collected data are preprocessed, and the preprocessed data are stored in the server database, specifically:

[0011] Use the fully automatic whole chicken weighing equipment to accurately weigh the chickens to be tested, and automatically calculate the average weight of the chickens. At the same time, collect relevant information of the chickens, including their original age;

[0012] The collected data are subjected to outlier and duplicate value elimination and missing value filling operations to construct a preprocessed historical average weight data set and a chicken flock related information data set, and the preprocessed historical average weight data set and the chicken flock related information data set are stored in a server database.

[0013] In this solution, the chicken flock average weight estimation model includes a low-order feature learning module, a high-order feature learning module and a prediction module;

[0014] Importing the historical average weight data in the database into the chicken flock average weight estimation model, in the low-order feature learning module, using an embedding layer composed of a fully connected layer to perform feature embedding, and performing Dropout processing on the fully connected layer to prevent overfitting;

[0015] The fully connected layer is used to reduce the dimension of the historical average weight data, and the original high-dimensional sparse vector is mapped to a low-dimensional dense vector to obtain the embedding vector ν of the chicken growth information feature. x ={x1v1,x2v2,…,x n ,v n}, where ν i ∈R is the embedding vector of the feature vector corresponding to the i-th historical weighted data, x i ∈R represents the feature vector corresponding to the i-th historical average weight data, i = 1, 2…, n, n represents the total number of feature vectors;

[0016] The embedding layer obtains a low-dimensional dense vector ν x Import the second-order cross layer in the low-order feature learning module as the input vector, perform a pairwise cross product operation on the input vector, and then add the operation results to obtain a low-dimensional cross dense vector f BI (V x ).

[0017] In this solution, in the chicken flock average weight estimation model, the low-dimensional cross-dense vector learned by the low-order feature learning module is input into the high-order feature learning module, and the high-order feature learning module includes an MLP neural network and a SoftPlus function;

[0018] The MLP neural network is used to learn the low-dimensional cross-dense vector, capture the high-order complex intrinsic correlation between feature vectors, fit the growth curve prediction function parameters, obtain the key growth parameters of the output of the MLP neural network, and use the SoftPlus function for further parameter fitting.

[0019] In this scheme, the dimension of the MLP neural network is set to [512, 128, 64, 32, 8], and the network parameters of the MLP neural network are updated and optimized using stochastic gradient descent. The model formula is expressed as:

[0020] z1=σ1(W1f BI (v x )+b1)

[0021] z2=σ2(W2z1+b2)

[0022] z3=σ3(W3z2+b3)

[0023] z4=σ4(W4z3+b4)

[0024] z5=σ5(W5z3+b5)

[0025] y=z5={slope,height,offset}

[0026] Among them, Z l represents the lth hidden layer, l=1,2,3,4,5,W l ,b l and σ represent the weight, bias vector and activation function of the lth layer respectively, f BI (v x ) represents a low-dimensional cross dense vector, slope represents the growth rate of the chicken flock, y represents the key growth parameter, height represents the growth rate of the chicken flock, and offset represents the offset.

[0027] In this solution, the prediction module of the chicken flock average weight estimation model constructs a growth curve function based on the key growth parameters fitted by the high-order feature learning module and the original age characteristics, uses the growth curve function to predict the chicken flock average weight, and outputs the chicken flock average weight prediction value;

[0028] The growth curve function f(slope, height, offset, age) is expressed as:

[0029]

[0030] Among them, slope represents the growth rate of the chicken flock, height represents the growth rate of the chicken flock, offset represents the offset, and age represents the original age characteristics.

[0031] In this solution, the predicted average weight of the chicken flock output by the chicken flock average weight estimation model is visualized and the predicted average weight of the chicken flock is converted into a graphical or chart form.

[0032] The second aspect of the present invention provides a chicken flock average weight estimation system based on multi-feature deep fusion learning, the system comprising a data acquisition module, a chicken flock average weight estimation module and a data visualization module;

[0033] The data acquisition module is responsible for weighing the chickens to be tested through the fully automatic whole chicken weighing equipment, collecting the average weight data of the chickens and related information of the chickens, pre-processing the collected data, and storing the processed data in the database server;

[0034] The flock average weight prediction module is responsible for building a flock average weight estimation model based on historical average weight data and flock related information in the database, using the flock average weight estimation model to predict the flock average weight of a specified time length, and storing the flock average weight prediction value in the database of the server;

[0035] The data visualization module is responsible for visually displaying the predicted average weight of the chicken flock, taking out the predicted average weight of the chicken flock from the database, and displaying the data in the form of graphics and charts.

[0036] Compared with the prior art, the present invention has the following beneficial effects:

[0037] (1) The present invention designs a new growth curve prediction function, which has the characteristics of efficient and accurate prediction of livestock and poultry weight compared with the previous growth curve function. The growth curve model and the deep learning network model are combined, which is particularly suitable for improving the accuracy and timeliness of livestock and poultry weight estimation.

[0038] (2) The present invention constructs a flock weight estimation model based on prior knowledge and multi-feature deep fusion learning. The model combines a low-order feature learning module, a high-order feature learning module and a prediction module to effectively improve the accuracy of livestock and poultry weight prediction. Through a multi-layer neural network structure, the flock weight estimation model can automatically extract high-order features from historical flock weight data, learn complex nonlinear relationships, and fit the growth curve function model parameters, thereby solving the problem of insufficient parameter fitting. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments or exemplary embodiments of the present invention, the drawings required for use in the embodiments or exemplary descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained according to the drawings without paying creative work.

[0040] Figure 1 A flowchart of a chicken flock average weight estimation method based on multi-feature deep fusion learning is shown;

[0041] Figure 2 A framework diagram of a chicken flock average weight estimation model in an embodiment is shown;

[0042] Figure 3 The framework diagram of the chicken flock average weight estimation system based on multi-feature deep fusion learning is shown. DETAILED DESCRIPTION

[0043] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0044] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0045] Figure 1 A flowchart of a chicken flock average weight estimation method based on multi-feature deep fusion learning is shown.

[0046] like Figure 1 As shown, this embodiment provides a chicken flock average weight estimation method based on multi-feature deep fusion learning, including:

[0047] S102, collecting the average weight data of the chicken group to be tested and related information of the chicken group, preprocessing the collected data, and storing the preprocessed data in the server database;

[0048] S104, constructing a flock weight estimation model based on historical average weight data and flock related information in a database, and performing low-order feature learning and high-order feature learning in the flock weight estimation model;

[0049] S106, using the relevant information of the flock of chickens as prior knowledge, using the parameters fitted by high-order feature learning and the prior knowledge to construct a growth curve function, and outputting a predicted value of the average weight of the flock of chickens through the growth curve function;

[0050] S108, returning the outputted predicted average weight value of the chicken flock to the server, and visually displaying the predicted average weight value of the chicken flock on the server.

[0051] It should be noted that the chicken flock to be tested is sent to a fully automatic whole chicken weighing device, which includes an electronic scale and a sensor. The fully automatic whole chicken weighing device is used to accurately weigh the chicken flock to be tested, and automatically calculate the average weight of the chicken flock. At the same time, relevant information of the chicken flock is collected, including the original age in days. To ensure the accuracy and completeness of the data, outliers and duplicate values ​​are removed and missing values ​​are filled in the collected data, and a preprocessed historical average weight data set and a chicken flock related information data set are constructed. The preprocessed historical average weight data set and chicken flock related information data set are stored in the server database for subsequent data analysis and breeding management tasks.

[0052] Figure 2 A framework diagram of a flock average weight estimation model in an embodiment is shown.

[0053] According to an embodiment of the present invention, the chicken flock average weight estimation model includes a low-order feature learning module, a high-order feature learning module and a prediction module. Specifically:

[0054] The low-order feature learning module includes two parts: embedding learning and second-order feature learning. The historical average weight data in the database is imported into the chicken flock average weight estimation model. In the low-order feature learning module, the embedding layer composed of the fully connected layer is used for feature embedding, and the fully connected layer is subjected to Dropout processing to prevent overfitting; the historical average weight data is subjected to feature dimension reduction using the fully connected layer, and the original high-dimensional sparse vector is mapped into a low-dimensional dense vector to solve the challenges brought by data sparsity, and the embedding vector ν of the chicken flock growth information feature is obtained. x ={x1v1,x2v2,…,x n ,v n}, where v i ∈R is the embedding vector of the feature vector corresponding to the i-th historical weighted data, x i ∈R represents the feature vector corresponding to the i-th historical average weight data, i = 1, 2…, n, n represents the total number of feature vectors;

[0055] The embedding layer obtains a low-dimensional dense vector ν x The second-order cross layer in the low-order feature learning module is imported as the input vector, the input vector is cross-producted pairwise, and then the operation results are added. The second-order feature interaction pooling layer outputs a low-dimensional cross dense vector f BI (V x ), the features are interacted with each other so that the model can learn more implicit relationships between features. The formula for calculating the vector of the second-order feature interaction pooling layer is expressed as:

[0056]

[0057] Among them, x i 、x j They are represented as the feature vectors corresponding to the i-th and j-th historical average weight data, v i 、v j They represent the embedding vectors of the feature vectors corresponding to the i-th and j-th historical average weight data respectively, and ⊙ represents the product between the feature vectors.

[0058] In the chicken flock average weight estimation model, the low-dimensional cross-dense vector learned by the low-order feature learning module is input into the high-order feature learning module, and the high-order feature learning module includes an MLP neural network and a SoftPlus function; the low-dimensional cross-dense vector is learned by the MLP neural network to capture the high-order and complex intrinsic correlation between feature vectors, and further alleviate the problems caused by data sparsity, high dynamics and diversity. The MLP neural network is used to fit the growth curve prediction function parameters, output key growth parameters, and use the SoftPlus function for further parameter fitting. The SoftPlus function is responsible for further fitting the key growth parameters learned by the MLP neural network to avoid negative numbers, outliers, etc., thereby improving the prediction accuracy and robustness of the model.

[0059] In the MLP neural network learning, the low-dimensional cross-dense vector learned in the low-order feature learning module is input into the MLP neural network to capture the complex intrinsic correlation between the cross features. The dimension of the MLP neural network is set to [512, 128, 64, 32, 8], and the model formula is expressed as:

[0060] z1=σ1(W1f BI (v x )+b1)

[0061] z2=σ2(W2z1+b2)

[0062] z3=σ3(W3z2+b3)

[0063] z4=σ4(W4z3+b4)

[0064] z5=σ5(W5z3+b5)

[0065] y=z5={slope,height,offset}

[0066] Among them, Z l represents the lth hidden layer, l=1,2,3,4,5,W l ,b land σ represent the weight, bias vector and activation function of the lth layer respectively. Preferably, the activation function is a nonlinear activation function Relu, f BI (v x ) represents a low-dimensional cross dense vector, slope represents the growth rate of the chicken flock, y represents the key growth parameter, height represents the growth rate of the chicken flock, and offset represents the offset.

[0067] The loss function of the MLP neural network is generated according to the mean square error, and the network parameters of the MLP neural network are updated and optimized using stochastic gradient descent. A training sample x is randomly selected, and each network parameter is updated in the direction of its negative gradient. The update process of the network parameters is expressed as:

[0068]

[0069] Among them, θ represents the trainable network parameters, η represents the learning rate that controls the gradient descent step size, y(x) represents the predicted value of training sample x.

[0070] Before the flock weight estimation model receives data, each feature vector is normalized by BN to reduce the internal covariate offset of each layer in the MLP neural network, thereby accelerating the convergence process of the model. i ∈R, all input vectors of each batch are B = x i , the formula of BN normalization is as follows Where μ represents the mini-batch mean, σ represents the mini-batch variance, γ and β represent trainable parameters used for scaling to improve the representation ability of the model, and ∈ represents a constant used to prevent the variance from being zero.

[0071] The parameters slope and height fitted by the MLP neural network model are further fitted by SoftPlus fitting. The SoftPlus function is defined as follows: x represents the function parameter. By fitting the key growth parameters slope and height in the chicken growth curve function, the generalization and accuracy of the model are improved, thereby effectively alleviating the problems of sparse chicken growth information, high dynamics and diversity.

[0072] The prediction module of the chicken flock average weight estimation model constructs a growth curve function based on the key growth parameters fitted by the high-order feature learning module and the original age characteristics, uses the growth curve function to predict the chicken flock average weight, and outputs the chicken flock average weight prediction value;

[0073] The growth curve function f(slope, height, offset, age) is expressed as:

[0074]

[0075] Among them, slope represents the growth rate of the chicken flock, height represents the growth rate of the chicken flock, offset represents the offset, and age represents the original age characteristics.

[0076] The predicted average weight of the flock output by the flock average weight estimation model is visualized, the predicted average weight of the flock is converted into graphics and charts, and the data information is converted into graphics and charts that are easy to analyze, so as to gain a clearer insight into the trends and patterns of the data and provide strong support for decision-making.

[0077] Figure 3 The framework diagram of the chicken flock average weight estimation system based on multi-feature deep fusion learning is shown.

[0078] The second embodiment of the present invention provides a chicken flock average weight estimation system based on multi-feature deep fusion learning, the system comprising a data acquisition module, a chicken flock average weight estimation module and a data visualization module;

[0079] The data acquisition module is responsible for weighing the chickens to be tested through the fully automatic whole chicken weighing equipment, collecting the average weight data of the chickens and related information of the chickens, pre-processing the collected data, and storing the processed data in the database server;

[0080] The data collection module collects data at a daily granularity, collects flock data in batches through intelligent data equipment, and stores the data in the server database. The data preprocessing program retrieves flock data from the server, performs preprocessing on the data such as outliers, duplicate values ​​removal, and missing value completion, and stores the processed data set in the server database.

[0081] The flock average weight prediction module is responsible for building a flock average weight estimation model based on the historical average weight data and flock related information in the database, and in the flock average weight estimation model, according to the fitted flock growth rate, flock growth rate, offset and other parameters, and integrating them with prior knowledge such as age and days into the proposed growth curve function, using the flock average weight estimation model to predict the flock average weight of a specified time length, and storing the flock average weight prediction value in the database of the server;

[0082] The data visualization module is responsible for visually displaying the predicted average weight of the flock, taking out the predicted average weight of the flock from the database, and displaying the data in the form of graphs and charts; by converting data information into easy-to-analyze graphs and charts, it is possible to gain a clearer insight into the trends and patterns of the data, thus providing strong support for decision-making.

[0083] The third embodiment of the present invention provides a computer-readable storage medium, which includes a program for a method for estimating the average weight of a chicken flock based on multi-feature deep fusion learning. When the program for estimating the average weight of a chicken flock based on multi-feature deep fusion learning is executed by a processor, the steps of the method for estimating the average weight of a chicken flock based on multi-feature deep fusion learning are implemented.

[0084] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored, or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0085] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0086] Those skilled in the art can understand that: all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiments; and the aforementioned storage medium includes: mobile storage devices, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), disks or optical disks, and other media that can store program codes.

[0087] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0088] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A chicken flock average weight estimation method based on multi-feature deep fusion learning, characterized in that: The following steps are involved: Collect the average weight data of the chickens to be tested and related information of the chickens, pre-process the collected data, and store the pre-processed data in the server database; Building a flock weight estimation model based on historical average weight data and flock related information in a database, and performing low-order feature learning and high-order feature learning in the flock weight estimation model; The relevant information of the flock of chickens is used as prior knowledge, and the parameters fitted by high-order feature learning and the prior knowledge are used to construct a growth curve function, and the predicted value of the average weight of the flock of chickens is output through the growth curve function; The output average weight prediction value of the chicken flock is returned to the server, and the average weight prediction value of the chicken flock is visualized on the server.

2. The chicken group average weight estimation method based on multi-feature deep fusion learning according to claim 1 is characterized in that: Collect the average weight data of the chickens to be tested and related information of the chickens, pre-process the collected data, and store the pre-processed data in the server database, specifically: Use the fully automatic whole chicken weighing equipment to accurately weigh the chickens to be tested, and automatically calculate the average weight of the chickens. At the same time, collect relevant information of the chickens, including their original age; The collected data are subjected to outlier and duplicate value elimination and missing value filling operations to construct a preprocessed historical average weight data set and a chicken flock related information data set, and the preprocessed historical average weight data set and the chicken flock related information data set are stored in a server database.

3. The chicken group average weight estimation method based on multi-feature deep fusion learning according to claim 1 is characterized in that: The chicken flock average weight estimation model includes a low-order feature learning module, a high-order feature learning module and a prediction module; Importing the historical average weight data in the database into the chicken flock average weight estimation model, in the low-order feature learning module, using an embedding layer composed of a fully connected layer to perform feature embedding, and performing Dropout processing on the fully connected layer to prevent overfitting; The fully connected layer is used to reduce the dimension of the historical average weight data, and the original high-dimensional sparse vector is mapped to a low-dimensional dense vector to obtain the embedding vector ν of the chicken growth information feature. x ={x1v1,x2v2,…,x n ,v n }, where ν i ∈R is the embedding vector of the feature vector corresponding to the i-th historical weighted data, x i ∈R represents the feature vector corresponding to the i-th historical average weight data, i = 1, 2…, n, n represents the total number of feature vectors; The embedding layer obtains a low-dimensional dense vector ν x Import the second-order cross layer in the low-order feature learning module as the input vector, perform a pairwise cross product operation on the input vector, and then add the operation results to obtain a low-dimensional cross dense vector f BI (V x ).

4. The chicken group average weight estimation method based on multi-feature deep fusion learning according to claim 3 is characterized in that: In the chicken flock average weight estimation model, the low-dimensional cross-dense vector learned by the low-order feature learning module is input into the high-order feature learning module, and the high-order feature learning module includes an MLP neural network and a SoftPlus function; The MLP neural network is used to learn the low-dimensional cross-dense vector, capture the high-order complex intrinsic correlation between feature vectors, fit the growth curve prediction function parameters, obtain the key growth parameters of the output of the MLP neural network, and use the SoftPlus function for further parameter fitting.

5. The chicken group average weight estimation method based on multi-feature deep fusion learning according to claim 4 is characterized in that: The dimensions of the MLP neural network are set to [512, 128, 64, 32, 8]. Stochastic gradient descent is used to update and optimize the network parameters of the MLP neural network. The model formula is expressed as: z1=σ1(W1f BI (in x )+b1) z2=σ2(W2z1+b2) z3=σ3(W3z2+b3) z4=σ4(W4z3+b4) z5=σ5(W5z3+b5) y=z5={slope,height,offset} Among them, Z l represents the lth hidden layer, l=1,2,3,4,5,W l ,b l and σ represent the weight, bias vector and activation function of the lth layer respectively, f BI (v x ) represents a low-dimensional cross dense vector, slope represents the growth rate of the chicken flock, y represents the key growth parameter, height represents the growth rate of the chicken flock, and offset represents the offset.

6. The chicken group average weight estimation method based on multi-feature deep fusion learning according to claim 3 is characterized in that: The prediction module of the chicken flock average weight estimation model constructs a growth curve function based on the key growth parameters fitted by the high-order feature learning module and the original age characteristics, uses the growth curve function to predict the chicken flock average weight, and outputs the chicken flock average weight prediction value; The growth curve function f(slope, height, offset, age) is expressed as: Among them, slope represents the growth rate of the chicken flock, height represents the growth rate of the chicken flock, offset represents the offset, and age represents the original age characteristics.

7. The chicken group average weight estimation method based on multi-feature deep fusion learning according to claim 1 is characterized in that: The predicted average weight of the chicken flock output by the chicken flock average weight estimation model is visualized and the predicted average weight of the chicken flock is converted into a graphic or chart form.

8. A chicken flock average weight estimation system based on multi-feature deep fusion learning, characterized in that: Implementing the chicken flock average weight estimation method based on multi-feature deep fusion learning as described in any one of claims 1 to 7, the system includes a data acquisition module, a chicken flock average weight estimation module and a data visualization module; The data acquisition module is responsible for weighing the chickens to be tested through the fully automatic whole chicken weighing equipment, collecting the average weight data of the chickens and related information of the chickens, pre-processing the collected data, and storing the processed data in the database server; The flock average weight prediction module is responsible for building a flock average weight estimation model based on historical average weight data and flock related information in the database, using the flock average weight estimation model to predict the flock average weight of a specified time length, and storing the flock average weight prediction value in the database of the server; The data visualization module is responsible for visually displaying the predicted average weight of the chicken flock, taking out the predicted average weight of the chicken flock from the database, and displaying the data in the form of graphics and charts.

Images