Calf island capable of sensing growth environment and intelligently adjusting calf island

Abstract

The invention discloses a growth environment perception and intelligent regulation calf island, and the intelligent calf island can achieve the precise control of the growth environment of calves through the comprehensive application of an advanced automatic control technology and an environment regulation system. The island body is made of polyethylene materials and has the good functions of keeping out wind and rain and keeping cool in summer and warm in winter, and it is guaranteed that the calves can obtain the best living environment in different seasons. The ventilation system and the heating device are automatically adjusted according to real-time environment data, it is ensured that the temperature and the gas concentration in the island are appropriate, the adverse effect of the environment on calf health is reduced, meanwhile, the feeding window is provided with an identification device and an intelligent control system, personalized feeding of the calves is achieved, and waste is avoided. The automatic and intelligent management of the whole system greatly improves the breeding efficiency, reduces the manual intervention, reduces the cost, and improves the health level and growth speed of the calves.

Description

Technical Field

[0001] The present invention belongs to the field of calf islands, and in particular, relates to a growth environment sensing and intelligently regulating calf island. Background Art

[0002] At present, most traditional calf farming methods rely on single-pen feeding or open farming environments. Although these methods can meet the basic growth needs of calves, they have many shortcomings and challenges in environmental management, feeding efficiency, animal health and growth rate. First of all, traditional farming environments often lack efficient temperature and humidity control systems, making it difficult to cope with seasonal climate changes and extreme weather. Especially in cold winters or hot summers, temperatures that are too high or too low may affect the appetite, activity level and overall health of calves, and even cause stress reactions, affecting their growth rate and immunity. In addition, many traditional farming environments also lack a complete ventilation system, which is prone to poor air circulation, resulting in excessive gas concentrations, especially ammonia and carbon dioxide. These harmful gases can seriously endanger the respiratory health of calves and increase the risk of disease transmission.

[0003] Secondly, most existing feeding systems use artificial feeding or simple automatic feeding devices, which cannot effectively avoid competition between calves, resulting in low feeding efficiency. In this environment, some calves may not be able to eat on demand, especially when some calves have a dominant position when eating, other calves cannot fully obtain nutrition, resulting in uneven weight gain, stunted development and even malnutrition. More importantly, traditional feeding systems often fail to achieve personalized management, ignoring the different growth needs and health conditions of each calf, resulting in waste of resources and inconsistent growth effects, which affects production performance after adulthood.

[0004] In view of this, the present invention is proposed. Summary of the invention

[0005] The technical problem to be solved by the present invention is to overcome the deficiencies of the prior art, provide a calf island for sensing and intelligently adjusting the growth environment, and solve the problems raised in the above-mentioned background technology.

[0006] In order to solve the above technical problems, the basic concept of the technical solution adopted by the present invention is:

[0007] A growth environment sensing and intelligently regulating calf island, comprising: an island body, made of polyethylene material, having the characteristics of wind and rain protection, cool in summer and warm in winter, for accommodating calves, a ventilation window being arranged on one side of the island body, and a closing plate being controlled by a servo motor on one side of the ventilation window;

[0008] A fence is used to enclose each calf, a feeding window is provided at the fence, an identification door is provided at the feeding window, and an identification device is provided at the identification door;

[0009] Ventilation system, which includes temperature sensors and harmful gas detection sensors. When the temperature inside the island is too high or the concentration of harmful gases is too high, the ventilation windows are automatically opened and closed to maintain air circulation inside the island and ensure a suitable environment for calves to move around.

[0010] Automatic disinfection system, which includes infrared sensors, automatically starts the disinfection spray system when no one is in the island to carry out disinfection, and immediately stops disinfection when the calf enters;

[0011] A heating device, the heating device includes a heating rod, the heating rod is located below the island body, and automatically heats the island body when the temperature inside the island body is lower than a set value;

[0012] Awnings: Awnings are used to protect calves' feeding areas and activity areas from rain.

[0013] Optionally, it also includes a heating device and a solar panel, which is installed on the top of the island body and supplies power to various electronic components in the island body through a power storage control component.

[0014] Optionally, a monitoring device is also provided in the island body, and the monitoring device includes a modeling unit, a control unit, a feedback unit, an analysis unit, and a prediction unit;

[0015] The modeling unit is used to collect real-time data transmitted from sensors and their identification devices and to establish a visual mathematical model of the Calf Island environment through data processing and statistical modeling techniques;

[0016] The analysis sheet is used to analyze environmental parameters and the growth data of calves, and store the analysis results in the repository to facilitate the subsequent call of the prediction unit;

[0017] The control unit is connected to the feedback unit and is used to adjust the status of each system in the island body in real time according to the processing results of the modeling unit and the analysis unit;

[0018] The feedback unit is used to monitor and feedback the real-time changes of various environmental data in the island to verify the effect of the control unit after executing the instructions. If the monitored environmental parameters deviate too much from the set values, the feedback unit will provide a warning to the control unit, indicating that the control unit needs to make adjustments;

[0019] Based on historical data and real-time collected information, the prediction unit predicts future trends in temperature and gas concentration changes to adjust the ventilation system and its heating device, analyzes the calf's activity data, predicts their behavior patterns, and adjusts the opening time of the identification device and feed supply based on the prediction results.

[0020] Optionally, the steps of collecting real-time data transmitted from sensors and their identification devices and establishing a mathematical model of the Calf Island environment through data processing and statistical modeling techniques are as follows:

[0021] The modeling unit collects environmental data inside and outside the calf island and calf behavior data through sensors and identification devices. Then, multiple regression analysis is used to model the impact of environmental factors on calf behavior. The expression is: Y = β0 + β1T + β2H + β3C + ∈, where Y is the growth parameter of the calf, T is the temperature, H is the humidity, C is the gas concentration, β0 is the constant term of the model, β1, β2, β3 are the regression coefficients of environmental factors, indicating the degree of influence of temperature, humidity and gas concentration on calf growth, ∈ is the error term, indicating the random error that cannot be explained in the model;

[0022] Based on the regression analysis in the second step, the modeling unit further established the environmental model of the calf island. Then, a weighted regression model was used to consider the non-balanced effects of different environmental factors on calf growth. The expression is: Among them, X i are different environmental factors, i represents the number of the environmental factor, β i is the regression coefficient corresponding to the environmental factor, indicating the specific impact of each factor on calf health, w i is a weighting factor that indicates the relative importance of different environmental factors to calf health.

[0023] Optionally, the analysis sheet is used to analyze the environmental parameters and the growth data of the calves, and store the analysis results in the repository to facilitate the subsequent calling of the prediction unit. The steps are:

[0024] The analysis unit collects real-time environmental data and calf growth data from various sensors and identification devices. Then, the analysis unit uses time series analysis to mine patterns and relationships in the data.

[0025] After the analysis is completed, the analysis unit will store the results in the data repository. Finally, the analysis results stored in the data repository will be connected with the prediction unit. The prediction unit will call the stored analysis results by database query and other means to provide data support for subsequent environmental adjustment and feeding plan formulation.

[0026] Optionally, the prediction unit predicts the future trend of temperature and gas concentration based on historical data and real-time collected information, adjusts the opening and closing of the ventilation system and its heating device, and analyzes the activity data of the calf to predict its behavior pattern. According to the prediction results, the opening time of the feeding window and the feed supply are adjusted in the following steps:

[0027] The prediction unit collects environmental data of the calf island from sensors and data acquisition systems, including temperature T t、Humidity t , gas concentration C t , and calf activity data;

[0028] Based on historical data and real-time data, the prediction unit uses the ARIMA model to predict the future trend of temperature and gas concentration, calculates the temperature and gas concentration changes in the future time period, and determines whether the ventilation system or heating device needs to be adjusted;

[0029] By analyzing the activity data of calves, a sequence model based on deep learning is used to predict the behavior pattern of calves, and its expression H t =LSTM(X t ,H t-1 ), where H t is the hidden state at the current moment, indicating the calf’s current behavior pattern, X t is the input data at the current moment, H t-1 is the hidden state of the previous moment, representing the memory of the previous behavior;

[0030] According to the predicted temperature and gas concentration change trends, the prediction unit generates corresponding control instructions and transmits them to the control unit. For feeding control, the prediction unit predicts the feed supply according to the calf's feeding needs, using formula F feed，t =κF predict，t +μ·sign(F actual,t -F predict，t) Adjust the opening and closing of the identification device and the feed amount, where F feed,t is the control command of the identification device and the feed supply, F predict,t is the predicted feed requirement, F actual,t is the actual feed supply, and κ and μ are adjustment coefficients representing the difference between the forecast and actual demand.

[0031] Optionally, the steps for building and training the ARIMA model are:

[0032] Collect historical time series data. For environmental data such as temperature and gas concentration, use the unit root test to test the stability of the data. If the data is not stable, perform difference processing. After the difference processing, perform the stability test again until the data is stable.

[0033] The sliding average order q is determined by the autocorrelation function, the autoregressive order p is determined by PACF, and then the Akaike information criterion and Bayesian information criterion are used to select the best combination of p, d, q. The smallest AIC or BIC value indicates the optimal model.

[0034] According to the determined order, the ARIMA model is established using the selected time series data, and its expression is: Among them, Y t is the value at the current moment, φ i is the coefficient of the autoregressive part, which represents the relationship between the current value and the previous p lagged values, θ j is the coefficient of the sliding average part, which represents the relationship between the current error and the previous q lag errors, ∈ t is the white noise error term at the current moment, and α is a constant term.

[0035] Optional, the steps for building and training a sequence model based on deep learning are:

[0036] Obtain environmental parameters and calf behavior data and divide the obtained data into a training set and a validation set;

[0037] Select the long short-term memory network as the architecture of the sequence model, and select the mean square error and Adam as the loss function and optimizer respectively;

[0038] Use the training set to train the deep learning model, adjust the model's weights through the back-propagation algorithm and optimizer, then verify the model's performance on the validation set, optimize the model's performance by adjusting the model's hyperparameters, and deploy the model to the actual system after the model is verified on the validation set.

[0039] Optionally, the recognition device is a camera, and the recognition device can automatically open and close according to the recognition results of the recognition system. The window is opened only when a specific calf approaches, and other calves cannot open the window. When the calf approaches the feeding window, the recognition device confirms the identity of the calf by scanning the tag. Once the calf approaches the window, the system will compare the recognition result with a pre-set database, wherein the database stores a unique identifier for each calf. When the identity authentication of the specific calf is passed, the recognition system sends a signal to the control unit to instruct the automatic door to open.

[0040] After adopting the above technical solution, the present invention has the following beneficial effects compared with the prior art. Of course, any product implementing the present invention does not necessarily need to achieve all the advantages described below at the same time:

[0041] The intelligent calf island of the present invention can achieve precise control of the calf's growth environment by comprehensively using advanced automatic control technology and environmental adjustment systems. The island body is made of polyethylene material, which has good wind and rain protection, cool in summer and warm in winter functions, ensuring that calves can get the best living environment in different seasons. The ventilation system and heating device are automatically adjusted according to real-time environmental data to ensure that the temperature and gas concentration in the island are appropriate, reducing the adverse effects of the environment on the health of the calves. At the same time, the feeding window is equipped with an identification device and an intelligent control system to achieve personalized feeding of calves and avoid waste. The automation and intelligent management of the overall system greatly improves breeding efficiency, reduces manual intervention, reduces costs, and also improves the health level and growth rate of calves.

[0042] The specific implementation modes of the present invention are further described in detail below in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The drawings described below are only some embodiments. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0044] In the figure:

[0045] Figure 1 This is one of the schematic diagrams of the three-dimensional structure of Calf Island;

[0046] Figure 2 This is the second schematic diagram of the three-dimensional structure of Calf Island;

[0047] Figure 3 This is the third schematic diagram of the three-dimensional structure of Calf Island;

[0048] Figure 4 This is the fourth schematic diagram of the three-dimensional structure of Calf Island;

[0049] Figure 5 for Figure 1 Schematic diagram of the structure at point A in the middle.

[0050] In the accompanying drawings, the components represented by the reference numerals are listed as follows:

[0051] 1. Island body; 2. Closing plate; 3. Fence; 4. Identification door; 5. Identification device; 6. Water storage mechanism; 7. Spray mechanism; 8. Water tank; 9. Water pump; 10. Water outlet pipe; 11. Heating rod; 12. Awning; 13. Column; 14. Wheels; 15. Solar panel.

[0052] It should be noted that these drawings and textual descriptions are not intended to limit the conceptual scope of the present invention in any way, but are intended to illustrate the concept of the present invention for those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0053] The present invention will now be described in further detail with reference to the accompanying drawings.

[0054] See also Figure 1-5 As shown, in this embodiment, a growth environment sensing and intelligently regulating calf island is provided, including an island body 1, which is made of polyethylene material, has the characteristics of sheltering from wind and rain, cool in summer and warm in winter, and is used to accommodate calves. A ventilation window is provided on one side of the island body 1, and a closing plate 2 is controlled by a servo motor on one side of the ventilation window;

[0055] A fence 3, the fence 3 is used to enclose each calf, a feeding window is arranged at the fence 3, an identification gate 4 is arranged at the feeding window, and an identification device 5 is arranged at the identification gate 4;

[0056] Ventilation system, which includes temperature sensors and harmful gas detection sensors. When the temperature in the island body 1 is too high or the concentration of harmful gases is too high, the ventilation windows are automatically controlled to open and close, so as to maintain air circulation in the island and ensure a suitable activity environment for the calves;

[0057] The automatic disinfection system includes an infrared sensor. When no one is in the island, the disinfection spray system is automatically started to disinfect. When the calf enters, the disinfection is stopped immediately. The automatic disinfection spray system consists of a water storage mechanism 6 and a spray mechanism 7. The water storage mechanism 6 includes a water tank 8 and a water pump 9 connected to the water outlet of the water tank 8. The spray mechanism 7 includes a water outlet pipe 10 connected to the water outlet of the water pump 9 and located in the island body 1. A plurality of atomizing nozzles are arranged below the water outlet pipe 10. An infrared sensor is installed in each calf island to monitor whether there is a calf entering the island. When the sensor detects that no one is in the island, the system automatically starts the disinfection spray system for disinfection. The spray system consists of a water storage mechanism 6, a water tank 8 and a spray mechanism 7 nozzle. When the disinfection mode is started, the water pump 9 draws water from the water tank 8 and flows to the plurality of atomizing nozzles through the water outlet pipe 10 to perform atomization disinfection of the air and the environment. Once the system detects that the calf has entered, the infrared sensor is triggered and the disinfection system stops automatically to prevent the disinfectant from affecting the calf.

[0058] A heating device, the heating device comprises a heating rod 11, the heating rod 11 is located below the island body 1, and automatically heats the island body 1 when the temperature inside the island body 1 is lower than a set value;

[0059] The canopy 12 is used to protect the feeding area and activity area of ​​the calves from rain, wherein the canopy 12 includes two column assemblies, and the column assembly includes two columns 13, wherein wheels 14 driven by a motor are arranged below the two columns 13, and the other two columns 13 are fixed to the ground.

[0060] The present embodiment also includes a heating device and a solar panel 15. The solar panel 15 is installed on the top of the island body 1 and supplies power to various electronic components in the island body 1 through a power storage control component.

[0061] In this embodiment, a monitoring device is also provided in the island body 1, and the monitoring device includes a modeling unit, a control unit, a feedback unit, an analysis unit, and a prediction unit;

[0062] The modeling unit is used to collect real-time data transmitted from the sensor and its identification device 5 and establish a visual mathematical model of the Calf Island environment through data processing and statistical modeling technology;

[0063] The analysis sheet is used to analyze environmental parameters and the growth data of calves, and store the analysis results in the repository to facilitate the subsequent call of the prediction unit;

[0064] The control unit is connected to the feedback unit and is used to adjust the status of each system in the island body 1 in real time according to the processing results of the modeling unit and the analysis unit;

[0065] The feedback unit is used to monitor and feedback the real-time changes of various environmental data in the island to verify the effect of the control unit after executing the instructions. If the monitored environmental parameters deviate too much from the set values, the feedback unit will provide a warning to the control unit, indicating that the control unit needs to make adjustments;

[0066] Based on historical data and real-time information collected, the prediction unit predicts the future trend of temperature and gas concentration changes to adjust the opening and closing of the ventilation system and its heating device, and analyzes the activity data of the calves, such as feeding, resting, and activities, to predict their behavior patterns. According to the prediction results, the opening time of the identification device 5 and the amount of feed supply are adjusted. The prediction unit uses the ARIMA model to predict the trend of temperature and gas concentration changes in the next few hours or days, and adjusts the ventilation windows and heating devices based on these predictions.

[0067] For example, if the ARIMA model predicts that the temperature will rise in the next three hours, the system will automatically open the ventilation windows to increase air circulation to avoid the impact of high temperatures on calf growth. Once the temperature is predicted to be too high or the gas concentration is too high, the system will automatically adjust the ventilation system and heating device. The control unit starts the servo motor through the signal to open and close the ventilation windows to ensure the appropriate temperature and air quality in the island.

[0068] In this embodiment, the steps of collecting real-time data transmitted from the sensor and its identification device 5 and establishing a mathematical model of the Calf Island environment through data processing and statistical modeling technology are as follows:

[0069] The modeling unit collects environmental data inside and outside the calf island and calf behavior data through sensors and identification devices 5, and then uses multiple regression analysis to model the impact of environmental factors on calf behavior, and its expression is: Y = β0 + β1T + β2H + β3C + ∈, where Y is the growth parameter of the calf, T is the temperature, H is the humidity%, and C is the gas concentration such as ammonia concentration, in ppm. β0 is the constant term intercept of the model. β1, β2, β3 are the regression coefficients of environmental factors, indicating the degree of influence of temperature, humidity and gas concentration on calf growth. is the error term, indicating the random error that cannot be explained in the model;

[0070] Based on the regression analysis in the second step, the modeling unit further established the environmental model of the calf island. Then, a weighted regression model was used to consider the non-balanced effects of different environmental factors on calf growth. The expression is: Among them, X i are different environmental factors, and i represents the number of the environmental factor. i is the regression coefficient corresponding to the environmental factor, indicating the specific impact of each factor on calf health, w i is a weighting factor that indicates the relative importance of different environmental factors to calf health.

[0071] In this embodiment, the analysis sheet is used to analyze the environmental parameters and the growth data of the calves, and store the analysis results in the storage repository to facilitate the subsequent calling of the prediction unit. The steps are:

[0072] The analysis unit collects real-time environmental data such as temperature, humidity, gas concentration, etc. and calf growth data from various sensors and identification devices 5. Then, the analysis unit uses time series analysis to mine the patterns and relationships in the data; the analysis focuses on:

[0073] The relationship between environmental data and calf growth: for example, analyzing the effects of environmental parameters such as temperature and humidity on calf growth, and evaluating the effects of environmental factors on calf health and daily weight gain.

[0074] Calf growth pattern: Analyze the growth pattern of calves under different environmental conditions and find key nodes such as growth acceleration period and stagnation period.

[0075] Anomaly detection: Identifying unusual patterns in the data, such as when the environment changes too quickly or when calves’ growth deviates from the normal trajectory;

[0076] After the analysis is completed, the analysis unit stores the results such as model parameters, analysis reports, and forecast trends in the data repository. Finally, the analysis results stored in the data repository are connected to the forecast unit. The forecast unit calls the stored analysis results by means of database query, etc., to provide data support for subsequent environmental adjustment and feeding plan formulation.

[0077] In this embodiment, the prediction unit predicts the future temperature and gas concentration change trends based on historical data and real-time collected information, adjusts the ventilation system and its heating device, and analyzes the calf's activity data such as feeding, resting, and activities to predict its behavior pattern. According to the prediction results, the opening time of the feeding window and the feed supply are adjusted in the following steps:

[0078] The prediction unit collects environmental data of the calf island from sensors and data acquisition systems, including temperature T t 、Humidity t , gas concentration C t , and calf activity data;

[0079] Based on historical data and real-time data, the prediction unit uses the ARIMA model to predict the future trend of temperature and gas concentration, calculates the temperature and gas concentration changes in the future time period, and determines whether the ventilation system or heating device needs to be adjusted;

[0080] By analyzing the activity data of calves, a sequence model based on deep learning is used to predict the behavior pattern of calves, and its expression H t =LSTM(X t ,H t-1 ), where H t is the hidden state at the current moment, indicating the calf’s current behavior pattern. t is the input data at the current moment, such as environmental variables such as temperature, humidity, and gas concentration, or the calf's behavioral data, such as food intake and activity level, etc. t-1 is the hidden state of the previous moment, representing the memory of the previous behavior;

[0081] According to the predicted temperature and gas concentration change trends, the prediction unit generates corresponding control instructions and transmits them to the control unit. For feeding control, the prediction unit predicts the feed supply according to the calf's feeding needs, using formula F feed,t =κF predict,t +μ·sign(F actual,t -F predict,t ) adjust the opening and closing of the identification device 5 and the amount of feed, wherein F feed,t is the control instruction of the identification device 5 and the feed supply, F predict,t is the predicted feed requirement, F actual,t is the actual feed supply, κ and μ are adjustment coefficients, which represent the difference between the predicted and actual demand. If the temperature is predicted to be too high or the gas concentration is too high, the system will automatically adjust the ventilation system switch; if the temperature is too low, the system will activate the heating device.

[0082] In this embodiment, the steps of establishing and training the ARIMA model are:

[0083] Collect historical time series data. For environmental data such as temperature and gas concentration, use the unit root test to test the stability of the data. If the data is not stable, perform differential processing. After differential processing, perform the stability test again until the data is stable;

[0084] The sliding average order q is determined by the autocorrelation function, the autoregressive order p is determined by PACF, and then the Akaike information criterion and Bayesian information criterion are used to select the best combination of p, d, q. The smallest AIC or BIC value indicates the optimal model.

[0085] According to the determined order p, d, q, the ARIMA model is established using the selected time series data, and its expression is: Among them, Y t is the value at the current moment, φ i is the coefficient of the autoregressive part, which represents the relationship between the current value and the previous p lagged values, θ j is the coefficient of the sliding average part, which represents the relationship between the current error and the previous q lag errors, ∈ t is the white noise error term at the current moment, and α is a constant term.

[0086] In this embodiment, the steps of establishing and training the sequence model based on deep learning are:

[0087] Obtain environmental parameters such as temperature, humidity, gas concentration, and calf behavior data and divide the acquired data into training sets and validation sets;

[0088] Select the long short-term memory network as the architecture of the sequence model, and select the mean square error and Adam as the loss function and optimizer respectively;

[0089] Use the training set to train the deep learning model, adjust the model's weights through the back-propagation algorithm and optimizer, then verify the model's performance on the validation set, optimize the model's performance by adjusting the model's hyperparameters, and deploy the model to the actual system after the model is verified on the validation set.

[0090] In this embodiment, the identification device 5 is a camera. The identification device 5 can automatically open and close according to the identification result of the identification system during identification. The window is opened only when a specific calf approaches, and other calves cannot open the window. When the calf approaches the feeding window, the identification device 5 confirms the identity of the calf through the tag scanning. Once the calf approaches the window, the system will compare the identification result with the pre-set database, wherein the database stores the unique identifier of each calf. When the identity authentication of the specific calf is passed, the identification system sends a signal to the control unit to instruct the automatic door to open. Each calf is equipped with an RFID tag. When the calf approaches the feeding window, the system identifies the calf's identity through a camera or an RFID scanner and confirms its unique identifier. The system compares the calf information stored in the database with the real-time identification result. For example, assuming that the calf number is 12345, when the calf approaches the feeding window, the identification device 5 will confirm the number 12345 and allow the calf to enter through the identification door 4. After the identification system confirms the identity, it sends a signal to the control unit to instruct the identification door 4 to open. Only verified calves can enter the feeding area to avoid interference from other calves. If the system detects other calves approaching and fails to pass identity verification, the door will not open. According to the calf's feeding needs, the system adjusts the feed supply through data analysis. For example, when it is predicted that the calf will enter the peak feeding period, the system will automatically increase the feed amount to ensure that each calf can eat enough.

[0091] The present invention is not limited to the above-mentioned embodiments. Anyone should be aware that any structural changes made under the enlightenment of the present invention, and any technical solutions that are the same or similar to the present invention, fall within the protection scope of the present invention. The technology, shape, and structural parts not described in detail in the present invention are all well-known technologies.

Claims

1. A calf island for sensing and intelligently adjusting the growth environment, characterized in that: include: The island body (1) is made of polyethylene material, has the characteristics of shielding from wind and rain, keeping cool in summer and warm in winter, and is used to accommodate calves. A ventilation window is arranged on one side of the island body (1), and a closing plate (2) is controlled by a servo motor on one side of the ventilation window; A fence (3), the fence (3) is used to enclose each calf, a feeding window is arranged at the fence (3), an identification door (4) is arranged at the feeding window, and an identification device (5) is arranged at the identification door (4); A ventilation system, the ventilation system includes a temperature sensor and a harmful gas detection sensor. When the temperature in the island body (1) is too high or the concentration of harmful gases is too high, the ventilation windows are automatically controlled to open and close, so as to maintain air circulation in the island and ensure that the calves have a suitable activity environment; Automatic disinfection system, which includes infrared sensors, automatically starts the disinfection spray system when no one is in the island to carry out disinfection, and immediately stops disinfection when the calf enters; A heating device, the heating device comprising a heating rod (11), the heating rod (11) being located below the island body (1) and automatically heating the island body (1) when the temperature inside the island body (1) is lower than a set value; A rain shelter (12) is used to protect the feeding area and activity area of ​​calves from rain.

2. A growth environment sensing and intelligent adjustment calf island according to claim 1, characterized in that: It also includes a heating device and a solar panel (15). The solar panel (15) is installed on the top of the island body (1) and supplies power to various electronic components in the island body (1) through a power storage control component.

3. The calf island for growth environment perception and intelligent regulation according to claim 1, characterized in that: A monitoring device is also provided in the island body (1), and the monitoring device comprises a modeling unit, a control unit, a feedback unit, an analysis unit, and a prediction unit; The modeling unit is used to collect real-time data transmitted from the sensor and its identification device (5) and establish a visual mathematical model of the Calf Island environment through data processing and statistical modeling technology; The analysis sheet is used to analyze environmental parameters and the growth data of calves, and store the analysis results in the repository to facilitate the subsequent call of the prediction unit; The control unit is connected to the feedback unit and is used to adjust the status of each system in the island body (1) in real time according to the processing results of the modeling unit and the analysis unit; The feedback unit is used to monitor and feedback the real-time changes of various environmental data in the island to verify the effect of the control unit after executing the instructions. If the monitored environmental parameters deviate too much from the set values, the feedback unit will provide a warning to the control unit, indicating that the control unit needs to make adjustments; The prediction unit predicts the future trend of temperature and gas concentration changes based on historical data and real-time information collected, adjusts the opening and closing of the ventilation system and its heating device, and analyzes the activity data of the calves to predict their behavior patterns. According to the prediction results, the opening time of the identification device (5) and the feed supply are adjusted.

4. The calf island for growth environment perception and intelligent regulation according to claim 1, characterized in that: The steps of collecting real-time data transmitted from the sensor and its identification device (5) and establishing a mathematical model of the Calf Island environment through data processing and statistical modeling techniques are as follows: The modeling unit collects environmental data inside and outside the calf island and calf behavior data through sensors and identification devices (5), and then uses multiple regression analysis to model the impact of environmental factors on calf behavior. The expression is: Y = β0 + β1T + β2H + β3C + ∈, where Y is the growth parameter of the calf, T is the temperature, H is the humidity, C is the gas concentration, β0 is the constant term of the model, β1, β2, β3 are the regression coefficients of environmental factors, indicating the degree of influence of temperature, humidity and gas concentration on calf growth, and is the error term, indicating the random error that cannot be explained in the model; Based on the regression analysis in the second step, the modeling unit further established the environmental model of the calf island. Then, a weighted regression model was used to consider the non-balanced effects of different environmental factors on calf growth. The expression is: Among them, X i are different environmental factors, i represents the number of the environmental factor, β i is the regression coefficient corresponding to the environmental factor, indicating the specific impact of each factor on calf health, w i is a weighting factor that indicates the relative importance of different environmental factors to calf health.

5. The calf island for growth environment perception and intelligent regulation according to claim 1, characterized in that: The analysis sheet is used to analyze environmental parameters and calf growth data, and store the analysis results in the repository to facilitate the subsequent prediction unit call steps: The analysis unit collects real-time environmental data and calf growth data from various sensors and identification devices (5), and then uses time series analysis to mine patterns and relationships in the data; After the analysis is completed, the analysis unit will store the results in the data repository. Finally, the analysis results stored in the data repository will be connected with the prediction unit. The prediction unit will call the stored analysis results by database query and other means to provide data support for subsequent environmental adjustment and feeding plan formulation.

6. The growth environment sensing and intelligent adjustment calf island according to claim 1 is characterized in that: Based on historical data and real-time information collected, the prediction unit predicts the future trend of temperature and gas concentration changes, adjusts the ventilation system and its heating device, and analyzes the calf's activity data (such as eating, resting, and activities) to predict its behavior pattern. According to the prediction results, the opening time of the feeding window and the feed supply are adjusted as follows: The prediction unit collects environmental data of the calf island from sensors and data acquisition systems, including temperature T t 、Humidity t , gas concentration C t , and calf activity data; Based on historical data and real-time data, the prediction unit uses the ARIMA model to predict the future trend of temperature and gas concentration, calculates the temperature and gas concentration changes in the future time period, and determines whether the ventilation system or heating device needs to be adjusted; By analyzing the activity data of calves, a sequence model based on deep learning is used to predict the behavior pattern of calves, and its expression H t =LSTM(X t ,H t-1 ), where H t is the hidden state at the current moment, indicating the calf’s current behavior pattern, X t is the input data at the current moment, H t-1 is the hidden state of the previous moment, representing the memory of the previous behavior; According to the predicted temperature and gas concentration change trends, the prediction unit generates corresponding control instructions and transmits them to the control unit. For feeding control, the prediction unit predicts the feed supply according to the calf's feeding needs, using the formula F feed,t =kF predict,t +μ·sign(F actual,t -F predict，t ) adjusts the opening and closing of the identification device (5) and the amount of feed, wherein F feed,t is the control instruction of the identification device (5) and the feed supply, F predict,t is the predicted feed requirement, F actual,t is the actual feed supply, k and μ are adjustment coefficients representing the difference between the forecast and actual demand.

7. The growth environment sensing and intelligent adjustment calf island according to claim 1 is characterized in that: The steps for establishing and training the ARIMA model are: Collect historical time series data. For environmental data such as temperature and gas concentration, use the unit root test to test the stability of the data. If the data is not stable, perform difference processing. After the difference processing, perform the stability test again until the data is stable. The sliding average order q is determined by the autocorrelation function, the autoregressive order p is determined by PACF, and then the Akaike information criterion and Bayesian information criterion are used to select the best combination of p, d, q. The smallest AIC or BIC value indicates the optimal model. According to the determined order (p, d, q), the ARIMA model is established using the selected time series data, and its expression is: Among them, Y t is the value at the current moment, φ i is the coefficient of the autoregressive part, which represents the relationship between the current value and the previous p lagged values, θ j is the coefficient of the sliding average part, which represents the relationship between the current error and the previous q lag errors, ∈ t is the white noise error term at the current moment, and α is a constant term.

8. The growth environment sensing and intelligent adjustment calf island according to claim 1, characterized in that: The steps for establishing and training a sequence model based on deep learning are: Obtain environmental parameters and calf behavior data and divide the obtained data into a training set and a validation set; Select the long short-term memory network as the architecture of the sequence model, and select the mean square error and Adam as the loss function and optimizer respectively; Use the training set to train the deep learning model, adjust the model's weights through the back-propagation algorithm and optimizer, then verify the model's performance on the validation set, optimize the model's performance by adjusting the model's hyperparameters, and deploy the model to the actual system after the model is verified on the validation set.

9. The growth environment sensing and intelligent adjustment calf island according to claim 1, characterized in that: The identification device (5) is a camera. The identification device (5) can automatically open and close according to the identification result of the identification system during identification. The window is opened only when a specific calf approaches, and other calves cannot open the window. When the calf approaches the feeding window, the identification device (5) confirms the identity of the calf by scanning the tag. Once the calf approaches the window, the system will compare the identification result with a pre-set database, wherein the database stores a unique identifier of each calf. When the identity authentication of the specific calf is passed, the identification system sends a signal to the control unit to instruct the automatic door to open.

Images