Meteorological element-based dairy cow daily milk yield prediction method and system

Abstract

The invention relates to a dairy cow daily milk yield prediction method and system based on meteorological elements. The method comprises the steps that meteorological data, dairy cow physiological data and milk yield data are collected and processed; based on the processed data, determining a high-temperature heat stress threshold value of the dairy cow, and establishing a high-temperature heat stress model of the dairy cow; performing model training by using a data analysis method in combination with the processed historical data to obtain a prediction model; and performing parameter adjustment on the prediction model to obtain a target prediction model for daily milk yield prediction. Existing meteorological data and dairy cow related data are collected, a dairy cow high-temperature heat stress threshold value is found, a dairy cow high-temperature heat stress model is constructed, model correction is carried out through a data analysis method and verification evaluation, a target prediction model is obtained, dairy cow high-temperature heat stress coping comprehensive technologies are summarized, and then precise and fine dairy cow meteorological services are carried out. Harm of high-temperature heat damage to dairy cows is effectively reduced, and the quality of milk products is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of milk production prediction, and in particular to a method and system for predicting daily milk production of dairy cows based on meteorological elements. Background Art

[0002] my country's dairy cattle animal husbandry is currently in a stable period of industry. Dairy cattle breeding has formed large-scale production, and large and medium-sized farms have become the main operators. In terms of production technology, it is also constantly developing in the direction of modernization. The management of dairy cattle breeding has gradually become stricter, causing small farmers to basically withdraw from the breeding industry, resulting in a sharp decline in the number of dairy cows. According to research: dairy cow milk production is mainly affected by genetics, breeding environment, climate and other factors, and high temperature in climate factors is the main factor affecting milk production. The suitable milk production temperature for dairy cows is 5.0-25.0℃. When the temperature is higher than the suitable temperature, dairy cows will have a heat stress response, resulting in a decrease in food intake, weight and milk production.

[0003] In recent years, with global warming, high temperature heat stress events in dairy cows have occurred frequently, causing great economic losses to ranches, and there is an urgent need for high temperature weather warning services for dairy cows. Therefore, it is of great practical application value to define a precise meteorological threshold range for high temperature heat stress in dairy cows, build a high temperature disaster loss model, and a model for predicting the daily milk production of dairy cows based on meteorological factors. However, the traditional meteorological threshold range for high temperature heat stress in dairy cows is obtained through manual measurement and observation, or through manual experience, and has the problem of low accuracy. Summary of the invention

[0004] Based on this, in order to solve the above technical problems, a method and system for predicting daily milk production of dairy cows based on meteorological elements is provided, which can summarize the comprehensive technology of coping with high temperature heat stress of dairy cows, and then carry out accurate and detailed direct meteorological services for dairy cows, effectively reduce the harm of high temperature heat stress to dairy cows, and improve the quality of dairy products.

[0005] A method for predicting daily milk production of dairy cows based on meteorological factors, the method comprising:

[0006] Collect meteorological data through a regional automatic weather station; obtain physiological data and milk production data of dairy cows; process the meteorological data, physiological data and milk production data respectively to obtain processed data;

[0007] Determine historical meteorological data, historical physiological data of dairy cows, and historical milk production data, and respectively process the historical meteorological data, historical physiological data of dairy cows, and historical milk production data to obtain processed historical data;

[0008] Based on the processed data, determining a high temperature heat stress threshold for dairy cows, and establishing a high temperature heat stress model for dairy cows according to the high temperature heat stress threshold for dairy cows;

[0009] Based on the high temperature heat stress model for dairy cows, a data analysis method is used in combination with the processed historical data to perform model training to obtain a prediction model;

[0010] Performing verification and evaluation on the prediction model, and adjusting parameters of the prediction model according to the verification and evaluation results to obtain a target prediction model;

[0011] The target prediction model is used to predict daily milk production, obtain prediction results, and generate a management report based on the prediction results.

[0012] In one embodiment, the meteorological data, the physiological data of the dairy cow, and the milk production data are processed respectively to obtain the processed data, including:

[0013] The meteorological data, the physiological data of dairy cows and the milk production data are respectively processed by removing abnormal values, filling missing values ​​and normalizing the data to obtain processed data.

[0014] In one embodiment, the method further comprises:

[0015] Determining environmental data and daily total production data from the processed data, and using the environmental data and daily total production data as indicators affecting milk production;

[0016] An independent variable matrix is ​​constructed based on the environmental data, and a dependent variable matrix is ​​constructed based on the daily total output.

[0017] In one embodiment, the method further comprises:

[0018] Taking the environmental data and the daily total production data as input, and inputting them into a principal component regression analysis algorithm;

[0019] Perform sample standardization processing by using the principal component regression analysis algorithm, calculate the covariance matrix and diagonalize the covariance matrix, and perform eigendecomposition processing;

[0020] Selecting a maximum eigenvector based on the decomposed features, and mapping an orthogonal matrix according to the maximum eigenvector to obtain a mapping matrix;

[0021] A principal component regression equation is output based on the mapping matrix.

[0022] In one embodiment, a data analysis method is used to perform model training in combination with the processed historical data to obtain a prediction model, including:

[0023] Using a support vector regression algorithm, taking the processed historical data as input, and introducing a slack variable and a kernel function into the support vector regression algorithm;

[0024] In the support vector regression algorithm, input data is mapped to a high-dimensional space through nonlinear mapping, linear regression is performed in the high-dimensional space, and a prediction model is obtained after gradient optimization.

[0025] In one embodiment, the prediction model is verified and evaluated, and parameters of the prediction model are adjusted according to the verification and evaluation results to obtain a target prediction model, including:

[0026] Use the leave-one-out method or the K-fold cross-validation method to verify the stability and accuracy of the prediction model to obtain a verification result;

[0027] Determine a performance evaluation indicator, and use the performance evaluation indicator to perform a performance evaluation on the prediction model to obtain an evaluation result;

[0028] The parameters of the prediction model are adjusted according to the verification result and the evaluation result to obtain a target prediction model.

[0029] In one embodiment, after using the target prediction model to predict daily milk production and obtaining a prediction result, the method further includes:

[0030] Performing accuracy analysis on the prediction result to obtain an accuracy analysis result;

[0031] According to the accuracy analysis result, determine whether to adjust the parameters of the target prediction model.

[0032] In one embodiment, the method further comprises:

[0033] Regularly evaluating the prediction results output by the target prediction model to determine whether to adjust the parameters of the target prediction model;

[0034] The target prediction model is regularly updated based on the latest collected meteorological data, cow physiological data, and milk production data.

[0035] In one embodiment, the meteorological data includes temperature, humidity, wind speed, wind direction, and sunshine duration; the cow physiological data includes the cow's breed, age, weight, feeding status, and breeding history; and the milk production data includes daily milk production data of each cow.

[0036] A system for predicting daily milk production of dairy cows based on meteorological elements, the system comprising:

[0037] The data acquisition module is used to collect meteorological data through the regional automatic weather station; obtain the physiological data and milk production data of dairy cows; and process the meteorological data, the physiological data and the milk production data respectively to obtain the processed data;

[0038] A data processing module, used to determine historical meteorological data, historical physiological data of dairy cows, and historical milk production data, and to process the historical meteorological data, historical physiological data of dairy cows, and historical milk production data, respectively, to obtain processed historical data;

[0039] A high temperature heat stress model establishment module, used to determine a high temperature heat stress threshold for dairy cows based on the processed data, and establish a high temperature heat stress model for dairy cows according to the high temperature heat stress threshold for dairy cows;

[0040] A model training module, used to perform model training based on the high temperature heat stress model of dairy cows using a data analysis method in combination with the processed historical data to obtain a prediction model;

[0041] A verification and evaluation module, used to verify and evaluate the prediction model, and adjust the parameters of the prediction model according to the verification and evaluation results to obtain a target prediction model;

[0042] The prediction module is used to use the target prediction model to predict the daily milk production, obtain the prediction result, and generate a management report according to the prediction result.

[0043] The above-mentioned method and system for predicting daily milk production of dairy cows based on meteorological elements collects existing meteorological data and dairy cow-related data, finds the high temperature heat stress threshold of dairy cows and constructs a high temperature heat stress model for dairy cows, corrects the model through data analysis methods and verification evaluation, obtains the target prediction model, summarizes the comprehensive technology for coping with high temperature heat stress of dairy cows, and then carries out accurate and meticulous meteorological services for dairy cows, effectively reducing the harm of high temperature heat damage to dairy cows and improving the quality of dairy products. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 This is an application environment diagram of a method for predicting daily milk production of dairy cows based on meteorological elements in one embodiment;

[0045] Figure 2 The present invention is a flow chart of a method for predicting daily milk production of dairy cows based on meteorological factors in one embodiment;

[0046] Figure 3 A schematic diagram of a flow chart of a method for predicting daily milk production of dairy cows based on meteorological factors in another embodiment;

[0047] Figure 4 The present invention is a structural block diagram of a system for predicting daily milk production of dairy cows based on meteorological factors in one embodiment;

[0048] Figure 5 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0050] The method for predicting daily milk production of dairy cows based on meteorological factors provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown. Figure 1 As shown, the application environment includes a computer device 110. The computer device 110 can collect meteorological data through a regional automatic weather station; obtain physiological data and milk production data of dairy cows; process the meteorological data, the physiological data of dairy cows, and the milk production data respectively to obtain processed data; the computer device 110 can determine historical meteorological data, historical physiological data of dairy cows, and historical milk production data, and process the historical meteorological data, historical physiological data of dairy cows, and historical milk production data respectively to obtain processed historical data; the computer device 110 can determine the high temperature heat stress threshold of dairy cows based on the processed data, and establish a high temperature heat stress model for dairy cows based on the high temperature heat stress threshold of dairy cows; the computer device 110 can use a data analysis method based on the high temperature heat stress model of dairy cows to perform model training in combination with processed historical data to obtain a prediction model; the computer device 110 can verify and evaluate the prediction model, and adjust the parameters of the prediction model according to the verification and evaluation results to obtain a target prediction model; the computer device 110 can use the target prediction model to predict daily milk production, obtain a prediction result, and generate a management report based on the prediction result. The computer device 110 may be, but is not limited to, various personal computers, notebook computers, and other devices.

[0051] In one embodiment, Figure 2 As shown, a method for predicting daily milk production of dairy cows based on meteorological elements is provided, comprising the following steps:

[0052] Step 202, collecting meteorological data through a regional automatic weather station; obtaining physiological data and milk production data of dairy cows; and processing the meteorological data, the physiological data and the milk production data respectively to obtain processed data.

[0053] First, the computer equipment can collect meteorological element values ​​such as average temperature, maximum temperature, relative humidity, wind direction, wind speed, sunshine duration, etc. through the regional automatic weather station, especially high temperature related data, including high temperature intensity, occurrence time, duration, etc. At the same time, the computer equipment can collect data such as cow weight, lactation days, food intake, milk production, etc. since the ranch has data records. The staff can separate the proportion of the impact of temperature on the weight, food intake, and milk production of dairy cows by consulting literature and field research. In this embodiment, the data set consists of milk production and meteorological data of 20 dairy cows for a total of 3 lactation cycles. The data has a long time span, high integrity, strong availability, and wide applicability, and has high scientific research applicability value.

[0054] Specifically, in one embodiment, a method for predicting daily milk production of dairy cows based on meteorological elements can also include a data processing process, and the specific process includes: removing outliers, filling missing values, and normalizing the meteorological data, dairy cow physiological data, and milk production data, respectively, to obtain processed data.

[0055] The computer equipment can clean and process the collected data, including removing outliers, filling missing values, and normalizing the data. In this embodiment, the collected data set D∈R n×d , which consists of meteorological data and milk production of 20 dairy cows in 3 lactation cycles, including 10 characteristic values: lactation days, average temperature, maximum temperature, time of maximum temperature, minimum temperature, time of minimum temperature, average relative humidity, minimum relative humidity, time of minimum relative humidity, and daily production. Since milk production data are manually reported and there are errors, in order to remove anomalies and noise in the data, reduce uncertainty in the data set, and improve the reliability and accuracy of the result analysis, the original data was cleaned and preprocessed. For example, cubic spline interpolation was used to replace outliers that deviated from the original data mean by several times the standard deviation.

[0056] In one embodiment, meteorological data includes temperature, humidity, wind speed, wind direction, and sunshine duration; physiological data of dairy cows includes breed, age, weight, feeding status, and breeding history of dairy cows; and milk production data includes daily milk production data of each dairy cow. Computer equipment can obtain data from regional automatic weather stations, collect meteorological parameters such as temperature, humidity, wind speed, and sunshine duration; record information such as breed, age, weight, feeding status, and breeding history of dairy cows; collect daily milk production data of each dairy cow, preferably using a modern automatic milk production measuring device to obtain accurate data.

[0057] Step 204, determining historical meteorological data, historical physiological data of dairy cows, and historical milk production data, and performing data processing on the historical meteorological data, historical physiological data of dairy cows, and historical milk production data, respectively, to obtain processed historical data.

[0058] Step 206: determine the high temperature heat stress threshold of the dairy cow based on the processed data, and establish a high temperature heat stress model for the dairy cow according to the high temperature heat stress threshold of the dairy cow.

[0059] Specifically, based on the graded change data of dairy cow weight, food intake, and milk production, combined with the values ​​of meteorological elements in the same period, the high temperature heat stress threshold of dairy cows can be formulated. Then, according to the high temperature heat stress threshold, grade and monitoring results of dairy cows, combined with actual production, the principal component regression and support vector regression methods are used to establish the high temperature heat stress model of dairy cows, compare the model performance, and realize the prediction and evaluation of the impact of high temperature.

[0060] In one embodiment, a method for predicting daily milk production of dairy cows based on meteorological elements may also include a variable selection process, the specific process including: determining environmental data and daily total production data from processed data, and using the environmental data and daily total production data as indicators affecting milk production; constructing an independent variable matrix based on the environmental data, and constructing a dependent variable matrix based on the daily total production.

[0061] Specifically, we can select key meteorological indicators that potentially affect milk production based on statistics and the biological principles of milk production. We can extract environmental data that are strongly correlated with total daily production and use them as regression variables to convert the data set into Split into an independent variable matrix X∈R consisting of environmental data n×d-1 and the dependent variable Y∈R consisting of the total daily output n×1 , and each indicator has its own dimension and variation, so the original data needs to be preprocessed.

[0062] Step 208, based on the high temperature heat stress model of dairy cows, a data analysis method is used in combination with the processed historical data to perform model training to obtain a prediction model.

[0063] Among them, data analysis methods may include linear regression, support vector machine, neural network and other methods to train prediction models with historical data.

[0064] In one embodiment, a method for predicting daily milk production of dairy cows based on meteorological elements can also include a process of model training using a principal component regression analysis algorithm, and the specific process includes: inputting environmental data and daily total production data into the principal component regression analysis algorithm; performing sample standardization processing through the principal component regression analysis algorithm, calculating the covariance matrix and diagonalizing the covariance matrix, and performing feature decomposition processing; selecting the maximum eigenvector based on the features after the decomposition processing, mapping the orthogonal matrix according to the maximum eigenvector, and obtaining a mapping matrix; outputting the principal component regression equation based on the mapping matrix.

[0065] The specific process includes: First, you can enter the environmental factors Total daily output Y∈R n×1 , information completeness (precision) ε; then perform sample standardization: Compute the covariance matrix: Diagonalize the covariance matrix, perform eigendecomposition, and then select the largest k largest eigenvectors: Mapping orthogonal matrix A = (ω 1 ,ω 2 ,…,ω k ), use XA to obtain the mapping matrix Z∈R n×k ; The output principal component regression equation is: in

[0066] In one embodiment, a method for predicting daily milk production of dairy cows based on meteorological elements is provided, which may also include a process of model training using an SVR algorithm, the specific process including: using a support vector regression algorithm, taking processed historical data as input, introducing slack variables and kernel functions into the support vector regression algorithm; mapping the input data to a high-dimensional space through nonlinear mapping in the support vector regression algorithm, performing linear regression in the high-dimensional space, and obtaining a prediction model after gradient optimization.

[0067] Among them, the support vector regression (SVR) used is a regression method based on support vector machine (SVM). Its principle is similar to linear regression, but it has higher accuracy and generalization ability. SVR models the regression process by finding an optimal hyperplane in high-dimensional space. Since this optimal hyperplane only considers the points on the edge of the training set, the model effectively avoids overfitting of data points. At the same time, according to the complexity control parameter of the reprojection error as the penalty term, the flexibility of the regression model can be well adjusted.

[0068] In this embodiment, compared with the traditional regression problem, the loss is 0 only when f(x)=w·x+b is exactly the same as y. SVR can tolerate a maximum deviation of ε between f(x) and y, that is, the loss is calculated only when |f(x)-y|>ε. The model is optimized by maximizing the width of the interval band and minimizing the total loss. The SVR problem can be formalized as follows:

[0069] However, in real tasks, it is difficult to directly determine the appropriate ε to ensure that most of the data are in the interval band, and the SVR goal is that all training data are in the interval band, so Introducing the slack variable ξ i and for: Since the experimental data D∈Rn×d (n>>d) The data dimension is much lower than the number of samples, and it is difficult to find the optimal hyperplane. Therefore, it is necessary to introduce a kernel function to transform the data set X∈R through a nonlinear mapping Φ. n ×d-1 Map to high-dimensional space F and perform linear regression in high-dimensional space, that is: f(x)=(w·Φ(x))+b,Φ:R d →F,w∈F; the RBF kernel function is used in this experiment: Nonlinear kernel functions can reduce training complexity and obtain smoother estimates. i and In this experiment, gradient optimization, cross-validation and wide-area search were used to select the specific variables with the best performance to ensure that the experimental model has the best performance.

[0070] Step 210, verify and evaluate the prediction model, and adjust the parameters of the prediction model according to the verification and evaluation results to obtain a target prediction model.

[0071] Specifically, in one embodiment, a method for predicting daily milk production of dairy cows based on meteorological elements can also include a process of verifying and evaluating the prediction model, and the specific process includes: using the leave-one-out method or the K-fold cross-validation method to verify the stability and accuracy of the prediction model to obtain a verification result; determining a performance evaluation indicator, and using the performance evaluation indicator to perform a performance evaluation on the prediction model to obtain an evaluation result; adjusting the parameters of the prediction model according to the verification result and the evaluation result to obtain a target prediction model.

[0072] The stability and accuracy of the prediction model can be verified by using methods such as leave-one-out or K-fold cross validation. 2 , root mean square error (RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE) and other indicators can be used to evaluate the performance of the prediction model, so as to adjust the model parameters to improve the prediction accuracy.

[0073] Specifically, in this embodiment, all samples of the data set are divided into a training set and a test set, and the ratio of the number of samples in the training set to the number of samples in the test set is 7:3. In order to improve the operation efficiency, the data set is normalized. The experimental results show that the SVR model is trained using the training set, and the relative error MPAE is only 0.38478% and the absolute error is 0.11582 when the training set is used to verify the model. The test set is used to test the support vector regression prediction effect. The relative error between the regression prediction value and the true value is 0.49908%, and the absolute error is 0.10699. Since the number of data set samples is much larger than the number of features (n>>r), the accuracy of solving the hyperplane using the kernel function high-dimensional mapping is high. At the same time, the experimental results in the support vector regression show a strong generalization ability and a high regression accuracy. The regression prediction results show that the SVR model has excellent applicability and versatility in the data set compared to the PCR model. In actual production, by inputting environmental parameters such as lactation days, temperature, humidity and other data in the data set and substituting them into the support vector regression function, the daily milk production of dairy cows can be predicted, and the average relative error of the prediction is only 5%, and the absolute error is only ±0.05, which can well guide the treatment of high temperature heat stress in dairy cows.

[0074] Step 212, use the target prediction model to predict daily milk production, obtain prediction results, and generate a management report based on the prediction results.

[0075] Among them, the management report can be a daily management report, and the computer equipment can also give environmental adjustment suggestions based on the prediction results, thereby assisting farmers in feeding management and operation adjustments.

[0076] In one embodiment, a method for predicting daily milk production of dairy cows based on meteorological elements may also include a process of adjusting prediction model parameters according to prediction results, and the specific process includes: performing accuracy analysis on the prediction results to obtain accuracy analysis results; and determining whether to adjust parameters of the target prediction model according to the accuracy analysis results.

[0077] Specifically, in this embodiment, the trained prediction model can be integrated into the existing dairy farm management system to ensure that meteorological data and milk production data can be automatically obtained to achieve automatic prediction of daily milk production; specifically, based on real-time meteorological data and cow information, the model is run to generate a predicted value of the cow's daily milk production; then the prediction results are analyzed to determine whether the prediction model needs to be adjusted or trained with longer data.

[0078] In one embodiment, a method for predicting daily milk production of dairy cows based on meteorological elements is provided, which may also include a process of regularly maintaining and updating the prediction model. The specific process includes: regularly evaluating the prediction results output by the target prediction model to determine whether to adjust the parameters of the target prediction model; regularly updating the target prediction model based on the latest collected meteorological data, dairy cow physiological data, and milk production data.

[0079] Computer equipment can regularly evaluate the model's prediction results to ensure long-term stability; conduct real-time data training and efficiently iterate the regression model to improve applicability; and regularly update the model with new data to reflect the impact of the latest environmental changes on milk production.

[0080] Since animal diseases, feed quality, management measures, etc. may also affect milk production, these factors can be incorporated into the prediction model to improve the prediction accuracy. Before using the method for predicting daily milk production of dairy cows based on meteorological elements provided in this application, dairy farm staff can be trained on the use and maintenance of the model.

[0081] In one embodiment, Figure 3 As shown, a method for predicting daily milk production of dairy cows based on meteorological factors is provided, which specifically includes:

[0082] Collect data on dairy cows on the farm, including lactation days, weight, milk production, etc.;

[0083] At the same time, data collected by automatic weather stations in the surrounding areas are collected, including temperature, humidity, wind direction, wind speed, etc.

[0084] Based on the collected data, conduct field research, consult literature, and separate the degree of influence of high temperature and humidity;

[0085] Then correlation analysis and polynomial curve fitting were performed;

[0086] Determine the heat stress threshold of dairy cows; establish a heat stress model for dairy cows;

[0087] Principal component regression analysis, support vector regression analysis;

[0088] Establish a prediction model for dairy cow milk production.

[0089] In this embodiment, the trained prediction model can be used to accurately and quickly predict the daily milk production of dairy cows, and the scientificity and accuracy of pasture management can be improved by timely adjusting environmental conditions; it helps to optimize resource allocation, such as feed distribution, staff arrangements and health monitoring measures, to improve the overall production efficiency and economic benefits of the pasture.

[0090] Specifically, in this embodiment, based on the existing meteorological data and cow-related data, the temperature and humidity at different thresholds, the response differences of cow weight, food intake and milk production are found, the meteorological indicators of high temperature heat stress in dairy cows are determined, and a high temperature heat stress model for dairy cows is constructed. The model is then corrected through project period data, and a prediction model for daily milk production of dairy cows based on meteorological elements is established. The comprehensive technology for coping with high temperature heat stress in dairy cows is summarized, and precise and detailed direct meteorological services for dairy cows are carried out to effectively reduce the harm of high temperature heat damage to dairy cows, improve the quality of dairy products, and ensure local economic development.

[0091] It should be understood that, although the steps in the above-mentioned flowcharts are shown in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in the above-mentioned flowcharts may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these sub-steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0092] In one embodiment, Figure 4 As shown, a system for predicting daily milk production of dairy cows based on meteorological elements is provided, comprising: a data acquisition module 410, a data processing module 420, a high temperature heat stress model establishment module 430, a model training module 440, a verification and evaluation module 450 and a prediction module 460, wherein:

[0093] The data acquisition module 410 is used to collect meteorological data through a regional automatic weather station; obtain physiological data and milk production data of dairy cows; and process the meteorological data, physiological data and milk production data respectively to obtain processed data;

[0094] The data processing module 420 is used to determine the historical meteorological data, the historical physiological data of dairy cows, and the historical milk production data, and respectively process the historical meteorological data, the historical physiological data of dairy cows, and the historical milk production data to obtain the processed historical data;

[0095] A high temperature heat stress model establishment module 430 is used to determine a high temperature heat stress threshold for dairy cows based on the processed data, and to establish a high temperature heat stress model for dairy cows according to the high temperature heat stress threshold for dairy cows;

[0096] A model training module 440 is used to perform model training based on a high temperature heat stress model for dairy cows using a data analysis method combined with processed historical data to obtain a prediction model;

[0097] The verification and evaluation module 450 is used to verify and evaluate the prediction model, and adjust the parameters of the prediction model according to the verification and evaluation results to obtain a target prediction model;

[0098] The prediction module 460 is used to use the target prediction model to predict the daily milk production, obtain the prediction result, and generate a management report according to the prediction result.

[0099] In one embodiment, the data processing module 420 is also used to remove abnormal values, fill missing values, and perform data normalization processing on the meteorological data, the physiological data of dairy cows, and the milk production data to obtain processed data.

[0100] In one embodiment, the model training module 440 is also used to determine environmental data and daily total production data from the processed data, and use the environmental data and daily total production data as indicators affecting milk production; construct an independent variable matrix based on the environmental data, and construct a dependent variable matrix based on the daily total production.

[0101] In one embodiment, the model training module 440 is also used to input environmental data and daily total production data as inputs to the principal component regression analysis algorithm; perform sample standardization processing through the principal component regression analysis algorithm, calculate the covariance matrix and diagonalize the covariance matrix, and perform feature decomposition processing; select the maximum eigenvector based on the features after the decomposition processing, map the orthogonal matrix according to the maximum eigenvector, and obtain the mapping matrix; output the principal component regression equation based on the mapping matrix.

[0102] In one embodiment, the model training module 440 is also used to use a support vector regression algorithm, take the processed historical data as input, introduce slack variables and kernel functions into the support vector regression algorithm; in the support vector regression algorithm, the input data is mapped to a high-dimensional space through nonlinear mapping, linear regression is performed in the high-dimensional space, and a prediction model is obtained after gradient optimization.

[0103] In one embodiment, the verification and evaluation module 450 is also used to verify the stability and accuracy of the prediction model using the leave-one-out method or the K-fold cross-validation method to obtain a verification result; determine a performance evaluation indicator, use the performance evaluation indicator to perform a performance evaluation on the prediction model to obtain an evaluation result; adjust the parameters of the prediction model according to the verification result and the evaluation result to obtain a target prediction model.

[0104] In one embodiment, the prediction module 460 is further used to perform accuracy analysis on the prediction results to obtain accuracy analysis results; and determine whether to adjust parameters of the target prediction model based on the accuracy analysis results.

[0105] In one embodiment, the model training module 440 is also used to regularly evaluate the prediction results output by the target prediction model to determine whether to adjust the parameters of the target prediction model; and regularly update the target prediction model based on the latest collected meteorological data, cow physiological data, and milk production data.

[0106] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 5 As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for predicting the daily milk production of dairy cows based on meteorological elements is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covered on the display screen, or a key, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.

[0107] Those skilled in the art will understand that Figure 5 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0108] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the steps of a method for predicting daily milk production of dairy cows based on meteorological elements are implemented.

[0109] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of a method for predicting daily milk production of dairy cows based on meteorological elements are implemented.

[0110] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0111] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0112] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A method for predicting daily milk production of dairy cows based on meteorological factors, characterized in that: The method comprises: Collect meteorological data through a regional automatic weather station; obtain physiological data and milk production data of dairy cows; process the meteorological data, physiological data and milk production data respectively to obtain processed data; Determine historical meteorological data, historical physiological data of dairy cows, and historical milk production data, and respectively process the historical meteorological data, historical physiological data of dairy cows, and historical milk production data to obtain processed historical data; Based on the processed data, determining a high temperature heat stress threshold for dairy cows, and establishing a high temperature heat stress model for dairy cows according to the high temperature heat stress threshold for dairy cows; Based on the high temperature heat stress model for dairy cows, a data analysis method is used in combination with the processed historical data to perform model training to obtain a prediction model; Performing verification and evaluation on the prediction model, and adjusting parameters of the prediction model according to the verification and evaluation results to obtain a target prediction model; The target prediction model is used to predict daily milk production, obtain prediction results, and generate a management report based on the prediction results.

2. The method for predicting daily milk production of dairy cows based on meteorological elements according to claim 1, characterized in that: The meteorological data, the physiological data of the dairy cow, and the milk production data are processed respectively to obtain the processed data, including: The meteorological data, the physiological data of dairy cows and the milk production data are respectively processed by removing abnormal values, filling missing values ​​and normalizing the data to obtain processed data.

3. The method for predicting daily milk production of dairy cows based on meteorological elements according to claim 1, characterized in that: The method further comprises: Determining environmental data and daily total production data from the processed data, and using the environmental data and daily total production data as indicators affecting milk production; An independent variable matrix is ​​constructed based on the environmental data, and a dependent variable matrix is ​​constructed based on the daily total output.

4. The method for predicting daily milk production of dairy cows based on meteorological elements according to claim 3, characterized in that: The method further comprises: Taking the environmental data and the daily total production data as input, and inputting them into a principal component regression analysis algorithm; Perform sample standardization processing by using the principal component regression analysis algorithm, calculate the covariance matrix and diagonalize the covariance matrix, and perform eigendecomposition processing; Selecting a maximum eigenvector based on the decomposed features, and mapping an orthogonal matrix according to the maximum eigenvector to obtain a mapping matrix; A principal component regression equation is output based on the mapping matrix.

5. The method for predicting daily milk production of dairy cows based on meteorological elements according to claim 1, characterized in that: The data analysis method is used in combination with the processed historical data to perform model training to obtain a prediction model, including: Using a support vector regression algorithm, taking the processed historical data as input, and introducing a slack variable and a kernel function into the support vector regression algorithm; In the support vector regression algorithm, input data is mapped to a high-dimensional space through nonlinear mapping, linear regression is performed in the high-dimensional space, and a prediction model is obtained after gradient optimization.

6. The method for predicting daily milk production of dairy cows based on meteorological elements according to claim 1, characterized in that: The prediction model is verified and evaluated, and parameters of the prediction model are adjusted according to the verification and evaluation results to obtain a target prediction model, including: Use the leave-one-out method or the K-fold cross-validation method to verify the stability and accuracy of the prediction model to obtain a verification result; Determine a performance evaluation indicator, and use the performance evaluation indicator to perform a performance evaluation on the prediction model to obtain an evaluation result; The parameters of the prediction model are adjusted according to the verification result and the evaluation result to obtain a target prediction model.

7. The method for predicting daily milk production of dairy cows based on meteorological elements according to claim 1, characterized in that: After using the target prediction model to predict daily milk production and obtaining a prediction result, the method further includes: Performing accuracy analysis on the prediction result to obtain an accuracy analysis result; According to the accuracy analysis result, determine whether to adjust the parameters of the target prediction model.

8. The method for predicting daily milk production of dairy cows based on meteorological elements according to claim 1, characterized in that: The method further comprises: Regularly evaluating the prediction results output by the target prediction model to determine whether to adjust the parameters of the target prediction model; The target prediction model is updated regularly based on the latest collected meteorological data, cow physiological data, and milk production data.

9. The method for predicting daily milk production of dairy cows based on meteorological elements according to claim 1, characterized in that: The meteorological data include temperature, humidity, wind speed, wind direction, and sunshine duration; the physiological data of dairy cows include the breed, age, weight, feeding status, and breeding history of dairy cows; and the milk production data include daily milk production data of each dairy cow.

10. A system for predicting daily milk production of dairy cows based on meteorological factors, characterized in that: The system comprises: The data acquisition module is used to collect meteorological data through the regional automatic weather station; obtain the physiological data and milk production data of dairy cows; and process the meteorological data, the physiological data and the milk production data respectively to obtain the processed data; A data processing module, used to determine historical meteorological data, historical physiological data of dairy cows, and historical milk production data, and to process the historical meteorological data, historical physiological data of dairy cows, and historical milk production data, respectively, to obtain processed historical data; A high temperature heat stress model establishment module determines a high temperature heat stress threshold for dairy cows based on the processed data, and establishes a high temperature heat stress model for dairy cows according to the high temperature heat stress threshold for dairy cows; A model training module, based on the high temperature heat stress model of dairy cows, uses a data analysis method in combination with the processed historical data to perform model training to obtain a prediction model; A verification and evaluation module, used to verify and evaluate the prediction model, and adjust the parameters of the prediction model according to the verification and evaluation results to obtain a target prediction model; The prediction module is used to use the target prediction model to predict the daily milk production, obtain the prediction result, and generate a management report according to the prediction result.

Images