Live pig breeding production management method and system based on artificial intelligence

Abstract

The invention discloses a pig breeding production management method and system based on artificial intelligence, and relates to the technical field of breeding management, and the method comprises the following steps: S1, determining piglet growth influence factors; s2, determining a feed ratio and an environment threshold value; s3, obtaining future environment data; s4, dynamically adjusting the pig house environment; and S5, accurately controlling the feed amount. By establishing the relationship among the feed ratio of the piglets, the living environment of the piglets and the growth curve of the piglets, important parameters of the feed ratio, the feed amount and the living environment which influence the growth of the piglets are obtained, so that corresponding equipment is controlled to dynamically adjust the environment of a pig house and determine the feed consumption and the feed ratio of the piglets every day; in addition, the accurate feed amount is provided for the piglets through the fuzzy control technology on the feeding equipment, the feed cost can be well saved through the fuzzy control technology, and an effective way is provided for cost management of live pig breeding.

Description

Technical Field

[0001] The present invention relates to the technical field of breeding management, and specifically to a pig breeding production management method and system based on artificial intelligence. Background Technique

[0002] With the increase in market demand and the progress of breeding technology, the pig breeding industry in China has gradually developed towards large-scale and professional directions. The scale of operation shows a rapid upward trend, and the proportion of large-scale pig breeding production in the country's total output has been continuously increasing. However, behind the rapid development of the breeding scale, many problems such as low breeding management level, backward management technology, and low automation level have continuously emerged, resulting in high pig breeding production costs, high labor intensity, and poor economic benefits. In the Chinese patent with the application number 202410875872.6, there is disclosed "a pig breeding intelligent decision-making management system based on multi-modal information monitoring, which relates to the technical field of pig breeding management. The intelligent decision-making management system includes: an information monitoring module, an abnormal warning module, an analysis and decision-making module, a management and regulation module, and an intelligent collaboration module. Among them, the information monitoring module is used to divide the breeding pig house into several captive areas through the captive fence, and use the monitoring equipment configured in each captive area to continuously monitor the multi-modal information of the pig group during the growth process in the captive area. The present invention realizes the intelligent management of the whole process of pig breeding by comprehensively applying the information monitoring, abnormal warning, analysis and decision-making, management and regulation, and intelligent collaboration modules."

[0003] The above document quickly identifies and warns of abnormal situations by comprehensively grasping the growth status of the pig group. However, the process of pig breeding is relatively complex, especially when raising piglets, strict management is required. There are many factors affecting the growth of piglets, such as old and worn-out equipment and low automation level, lack of standardization in feed management, inaccurate feed ratio, and obvious feed waste. If only the growth status of pigs is grasped for abnormal identification without timely and accurately controlling the growth environment and providing more accurate feed quantity and feed ratio for piglets, the breeding quality of piglets cannot be guaranteed, and the breeding cost will also increase. The intelligent effect and operation effect of the entire management process are relatively poor, and there are still great problems in intelligent breeding. Summary of the Invention

[0004] The purpose of the present invention is to provide a pig breeding production management method and system based on artificial intelligence to solve the problems raised in the above background technique.

[0005] To achieve the above purpose, the present invention provides the following technical solution: A pig breeding production management method based on artificial intelligence, including the following steps:

[0006] S1. Determine the influencing factors of piglet growth: Input the collected pigsty environmental data, piglet growth data, and piglet feeding data into the SVM model for training, so as to establish the relationship between the feed ratio of piglets, the living environment of piglets, and the growth curve of piglets;

[0007] S2. Determine the feed ratio and environmental thresholds: Calculate the daily required nutrient values of piglets according to the SVM model parameters, determine the feed amount and feed ratio based on the nutrients contained in the feed and the amount of each nutrient, and determine the temperature threshold, humidity threshold, and gas concentration threshold of the pigsty environment according to the piglet growth data;

[0008] S3. Obtain future environmental data: Input the collected pigsty environmental data into the GRU model to predict the temperature change curve T(t), humidity change curve H(t), carbon dioxide concentration change curve C(t), and ammonia concentration change curve N(t) in the next 24 hours. If environmental anomalies occur, the controller formulates strategies to operate the equipment;

[0009] S4. Dynamically adjust the pigsty environment: Monitor the environmental changes in the pigsty in real time. After obtaining the environmental prediction data of the temperature, humidity, and harmful gas concentration in the pigsty, dynamically adjust the output power of relevant equipment;

[0010] S5. Precisely control the feed amount: Use fuzzy control technology on the feeding equipment to provide precise feed amounts for piglets. Select the percentage error E of the falling material amount and the error change rate EC as the input quantities of the fuzzy controller, and select the falling time amount U of the feed feeding equipment as the output quantity of the fuzzy controller.

[0011] Preferably, in step S1, the RFID technology is used to collect various indicators of piglets. The various indicators include the pig number, body temperature, body weight, and backfat thickness of the pig. The body weight of the piglet is collected when the piglet calms down. When the piglet is completely relaxed when it has eaten half of its food, the body temperature and backfat thickness of the piglet are collected. When the piglet finishes eating, the remaining feed in the feeding trough is weighed, and then the remaining amount is compared with the previously configured feed amount to obtain the feed amount per meal of the piglet. The environmental data of the temperature, humidity, and harmful gas concentration in the pigsty are collected through temperature sensors, humidity sensors, and gas sensors.

[0012] Preferably, in step S2, the basic information of each pigsty is recorded. The basic information of the pigsty includes the pigsty number, the size of the pigsty space, the pigsty environmental data, and the maximum capacity of piglets that the pigsty can accommodate. The size of the piglet activity space is proportional to the growth of piglets, while the size of the piglet activity space is inversely proportional to the cost. Therefore, the maximum capacity of the pigsty is determined according to the growth status of piglets and the cost to ensure the breeding efficiency.

[0013] Preferably, step S3 includes the following steps:

[0014] S301. Compare the temperature change curve T(t) with the set high temperature threshold Th and low temperature threshold Tl. If Tl < T(t) < Th, it indicates that the temperature is normal. If T(t) ≤ Tl, it indicates a low temperature anomaly, and the controller starts to formulate a heating strategy. If T(t) ≥ Th, it indicates a high temperature anomaly, and the controller starts to formulate a cooling strategy.

[0015] S302. Compare the humidity change curve H(t) with the set high humidity threshold Hh and dry threshold Hl. If Hl < H(t) < Hh, it indicates that the humidity is normal. If H(t) ≤ Hl, it indicates a low humidity anomaly, and the controller starts to formulate a humidifying strategy. If H(t) ≥ Hh, it indicates a high humidity anomaly, and the controller starts to formulate a dehumidifying strategy.

[0016] S303. Compare the carbon dioxide concentration change curve C(t) and ammonia concentration change curve N(t) with the respectively set gas concentration thresholds. If only carbon dioxide exceeds the gas concentration threshold, formulate a ventilation strategy according to the concentration of carbon dioxide. If only ammonia exceeds the gas concentration threshold, formulate a ventilation strategy according to the concentration of ammonia. If both carbon dioxide and ammonia exceed the gas concentration threshold, calculate the anomaly coefficients of carbon dioxide and ammonia, and formulate a ventilation strategy according to the gas with the larger anomaly coefficient, where the anomaly coefficient is used to measure the anomaly degree of harmful gases.

[0017] Preferably, step S4 includes the following steps:

[0018] S401. Execute the strategies of heating, cooling, humidifying, dehumidifying and ventilation according to the instructions of the controller, and control the output power of each strategy during the equipment execution stage.

[0019] S402. When adjusting the temperature, humidity and harmful gases in the pig house, considering the situation that the adjustment processes of various environmental parameters affect each other, set the priority of the adjustment strategies from high to low as harmful gases, temperature and humidity.

[0020] Preferably, step S5 includes the following steps:

[0021] S501. The fuzzy control technology is as follows: First, determine the initial input quantity of the fuzzy controller, then determine the basic domain of the initial input quantity, determine the quantization factor of fuzzification through the basic domain, and finally obtain the final input quantity of the fuzzy controller through the conversion formula to complete the fuzzification operation of the input quantity, and then accurately control the output quantity.

[0022] S502. The initial input quantity is the value of the real-time material falling quantity obtained by using a weighing sensor. The automatically set material falling quantity is compared with the material falling quantity obtained from the weighing sensor. After the comparison, the material falling quantity error percentage E and the error change rate EC are obtained.

[0023] Preferably, step S5 further includes the following steps:

[0024] S503. Determine that the continuous value range of the material falling quantity error percentage E is [eL, eH], and the basic domain is defined as {-a, -a + 1, ···, -1, 0, 1, ···, a - 1, a}. The quantization factor ke of the material falling quantity error percentage E is:

[0025]

[0026] where a, eL, and eH are all constant values;

[0027] S504. Determine that the continuous value range of the error change rate EC is [ecL, ecH], and the basic domain is defined as {-b, -b + 1, ···, -1, 0, 1, ···, b - 1, b}. The quantization factor kec of the error change rate EC is:

[0028]

[0029] where b, ecL, and ecH are all constant values. Let the time for obtaining the material falling quantity in one cycle be m seconds, then

[0030] S505. Determine that the continuous value range of the material falling time quantity U is [uL, uH], and the basic domain is defined as {-c, -c + 1, ···, -1, 0, 1, ···, c - 1, c}. The quantization factor ku of the material falling time quantity U is:

[0031]

[0032] uH - uL = km

[0033] where c, uL, and uH are all constant values, and k is the cycle coefficient of the material falling quantity;

[0034] S506. Substitute the quantization factor ke, the quantization factor kec, and the quantization factor ku into the conversion formula respectively to complete the fuzzy calculation of the error percentage E, the error change rate EC, and the material falling time quantity U. The conversion formula is:

[0035]

[0036] where E * is the error percentage of the fuzzification operation, and EC *is the error change rate of the fuzzification operation, U * is the blanking time amount after precise control.

[0037] An artificial intelligence-based production management system for pig breeding includes the following units:

[0038] Growth correlation unit: The growth correlation unit inputs the collected pigsty environmental data, piglet growth data, and piglet feeding data into the SVM model for training, thereby establishing the relationship between the feed ratio of piglets, the living environment of piglets, and the growth curve of piglets. Determine the feed amount and feed ratio according to the SVM model parameters, and determine the temperature threshold, humidity threshold, and gas concentration threshold of the pigsty environment according to the piglet growth data;

[0039] Equipment startup unit: The equipment startup unit inputs the environmental data of the temperature, humidity, and harmful gas concentration collected by the temperature sensor, humidity sensor, and gas sensor in the pigsty into the GRU model to predict the environmental data for the next 24 hours, thereby controlling the operation of the equipment;

[0040] Environmental regulation unit: The environmental regulation unit monitors the environmental changes in the pigsty in real time. After obtaining the environmental prediction data of the temperature, humidity, and harmful gas concentration in the pigsty, it dynamically adjusts the output power of relevant equipment;

[0041] Feed control unit: The feed control unit uses fuzzy control technology on the feeding equipment to provide precise feed amounts for piglets. Select the percentage of blanking amount error and error change rate as the input quantities of the fuzzy controller, and select the blanking time amount of the feed feeding equipment as the output quantity of the fuzzy controller.

[0042] Compared with the prior art, the beneficial effects of the present invention are:

[0043] 1. By establishing the relationship between the feed ratio of piglets, the living environment of piglets, and the growth curve of piglets, the present invention obtains important parameters of the feed ratio, feed amount, and living environment that affect the growth of piglets, provides decision-making information for breeding management, thereby controlling the corresponding equipment to dynamically adjust the environment of the pigsty and determine the daily feed consumption and feed ratio of piglets, ensuring the breeding quality of live pigs. In addition, by using fuzzy control technology on the feeding equipment to provide precise feed amounts for piglets and applying it to the feed ratio of piglets, more accurate feed amounts are provided for piglets. The fuzzy control technology can well save feed costs and better save the breeding costs of the entire pig farm, providing an effective way for the cost management of pig breeding;

[0044] 2. The present invention formulates strategies based on the predicted future environmental change curve, and combines the environmental parameters of the pigsty and the equipment parameters for regulating the environment to dynamically adjust the output power of relevant equipment. It not only restores the environmental parameters to the target values, but also reduces the occurrence of abnormal environments, keeping the living environment of piglets within the normal range. Since there are mutual influences during the adjustment process of each environmental parameter when regulating the temperature, humidity, and harmful gases in the pigsty, the harmful gases that are more harmful to piglets are given priority treatment. The entire management process has an obvious intelligent effect, operates stably, and has high application value in the breeding field. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a schematic diagram of the overall method flow provided by an embodiment of the present invention;

[0046] Figure 2 It is a schematic diagram of the overall system structure provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0047] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0048] Please refer to Figures 1 to 2 , the present invention provides a technical solution: a pig breeding production management method based on artificial intelligence, including the following steps:

[0049] S1. Determine the influencing factors for piglet growth: Input the collected environmental data of the pigsty, the growth data of piglets, and the feeding data of piglets into the SVM model for training, so as to establish the relationship between the feed ratio of piglets, the living environment of piglets, and the growth curve of piglets;

[0050] S2. Determine the feed ratio and environmental thresholds: Calculate the daily required nutritional value of piglets according to the SVM model parameters, determine the feed amount and feed ratio according to the nutritional components contained in the feed and the amount of each nutritional component, and determine the temperature threshold, humidity threshold, and gas concentration threshold of the pigsty environment according to the growth data of piglets;

[0051] S3. Obtain future environmental data: Input the collected environmental data of the pigsty into the GRU model to predict the temperature change curve T(t), humidity change curve H(t), carbon dioxide concentration change curve C(t), and ammonia concentration change curve N(t) in the next 24 hours. If there is an environmental abnormality, the controller formulates strategies to operate the equipment;

[0052] S4. Dynamically adjust the pigsty environment: Monitor the environmental changes in the pigsty in real time. After obtaining the environmental prediction data of the temperature, humidity, and harmful gas concentration in the pigsty, dynamically adjust the output power of relevant equipment;

[0053] S5. Precisely control the feed quantity: Use fuzzy control technology on the feeding equipment to provide precise feed quantity for piglets. Select the percentage error E of the material dropping quantity and the error change rate EC as the input quantities of the fuzzy controller, and select the material dropping time quantity U of the feed feeding equipment as the output quantity of the fuzzy controller.

[0054] In step S1, use RFID technology to collect various indicators of piglets. The various indicators include the pig's number, body temperature, weight, and backfat thickness. When the piglet calms down, collect the weight of the piglet. When the piglet is completely relaxed when it has eaten half of its food, collect the body temperature and backfat thickness of the piglet. When the piglet finishes eating, weigh the remaining feed in the feeding trough, and then compare the remaining amount with the previously configured feed amount to obtain the feed amount per meal of the piglet. Collect the environmental data of the temperature, humidity, and harmful gas concentration in the pigsty through temperature sensors, humidity sensors, and gas sensors;

[0055] The temperature sensor and humidity sensor are specifically SHT30 type temperature and humidity sensors. The gas sensor for detecting carbon dioxide is PTG - 100 - 01 type, and the gas sensor for detecting ammonia is NH3 / CR - 200 type;

[0056] In step S2, record the basic information of each pigsty. The basic information of the pigsty includes the pigsty number, pigsty space size, pigsty environmental data, and the maximum capacity of piglets that the pigsty can accommodate. The size of the piglet activity space is proportional to the growth of piglets, while the size of the piglet activity space is inversely proportional to the cost. Therefore, the maximum capacity of the pigsty is determined according to the growth status of piglets and the cost to ensure the breeding efficiency;

[0057] Step S3 includes the following steps:

[0058] S301. Compare the temperature change curve T(t) with the set high - temperature threshold Th and low - temperature threshold Tl. If Tl < T(t) < Th, it means that the temperature is normal. If T(t) ≤ Tl, it means that a low - temperature anomaly has occurred, and the controller starts to formulate a heating strategy. If T(t) ≥ Th, it means that a high - temperature anomaly has occurred, and the controller starts to formulate a cooling strategy;

[0059] S302. Compare the humidity change curve H(t) with the set high humidity threshold Th and low humidity threshold Tl. If Hl < H(t) < Hh, it indicates that the humidity is normal. If H(t) ≤ Hl, it indicates a low humidity anomaly, and the controller starts to formulate a humidification strategy. If H(t) ≥ Hh, it indicates a high humidity anomaly, and the controller starts to formulate a dehumidification strategy;

[0060] S303. Compare the carbon dioxide concentration change curve C(t) and ammonia concentration change curve N(t) with the respectively set gas concentration thresholds. If only carbon dioxide exceeds the gas concentration threshold, formulate a ventilation strategy based on the concentration of carbon dioxide. If only ammonia exceeds the gas concentration threshold, formulate a ventilation strategy based on the concentration of ammonia. If both carbon dioxide and ammonia exceed the gas concentration threshold, calculate the anomaly coefficients of carbon dioxide and ammonia, and formulate a ventilation strategy based on the gas with the larger anomaly coefficient. Among them, the anomaly coefficient is used to measure the anomaly degree of harmful gases;

[0061]

[0062] Among them, A(t) represents the anomaly coefficient of the gas at time t, C(t) represents the predicted value of the gas at time t, and C target represents the target value of the gas;

[0063] Step S4 includes the following steps:

[0064] S401. Execute the strategies of heating, cooling, humidifying, dehumidifying and ventilation according to the instructions of the controller, and control the output power of each strategy during the equipment execution stage;

[0065] The controller for controlling the operation of the equipment is specifically a SIMATIC S7-200 CN CPU 226 model controller;

[0066] If a heating strategy is formulated, the output power calculation formula of the heating fan in the time period (t, t+dur) is:

[0067]

[0068] Among them, P1(t, t+dur) represents the output power of the heating fan in the time period (t, t+dur), C represents the specific heat capacity of air, ρ represents the density of air, V represents the volume of the pigsty, A(t) and A target respectively represent the values measured by the temperature sensor and the target value, dur represents the operation duration of the equipment, η represents the efficiency of the heating equipment, t represents the time, and |·| represents the absolute value;

[0069] The heating fan is specifically a BGO-70A-19-F model warm air blower;

[0070] If a cooling strategy is formulated, the calculation formula for the output power of the cooling fan in the time period (t, t+dur) is as follows:

[0071]

[0072] where P2(t, t+dur) represents the output power of the cooling fan in the time period (t, t+dur), A(t) and A target respectively represent the values measured by the temperature sensor and the target value, V represents the volume of the pigsty, S represents the cross-sectional area of the air inlet, k represents the cooling efficiency of the wet curtain. When the water pump is not turned on, k is 0, dur represents the operating duration of the equipment, p represents the increase in the output power of the fan corresponding to each 1 m / s increase in the air inlet wind speed, t represents the moment, and |·| represents the absolute value;

[0073] The specific model of the cooling fan is the wet curtain cooling fan of the Run Dong Fang model;

[0074] If a humidification strategy is formulated, the calculation formula for the output power of the humidifier in the time period (t, t+dur) is as follows:

[0075] P3(t, t+dur) = p×V×ρ×[A target -A(t)]×ts×C×dur / 1000

[0076] where P3(t, t+dur) represents the output power of the humidifier in the time period (t, t+dur), p represents the energy consumption of the humidifier for delivering a unit volume of moisture to the air, V represents the volume of the pigsty, ρ represents the density of the air, ts represents the ventilation rate, C represents the loss coefficient, dur represents the operating duration of the equipment, A(t) and A target respectively represent the values measured by the humidity sensor and the target value, and t represents the moment;

[0077] The specific humidifier is the humidifier of the Philips HU4803 / 00 model;

[0078] If a dehumidification strategy is formulated, the calculation formula for the output power of the dehumidifier in the time period (t, t+dur) is as follows:

[0079] P4(t, t+dur) = p×V1×V2×[A(t)-A target ×C×dur

[0080] where P4(t, t+dur) represents the output power of the dehumidifier in the time period (t, t+dur), p represents the energy consumption of the dehumidifier for removing a unit volume of moisture from the air, V1 represents the volume of the pigsty, V2 represents the volume of fresh air, A(t) and A target respectively represent the values measured by the humidity sensor and the target value, C represents the loss coefficient, and dur represents the operating duration of the equipment;

[0081] The model of the dehumidifier is specifically the HJ-858H dehumidifier;

[0082] If a ventilation strategy is formulated, the calculation formula for the output power of the ventilation fan in the time period (t, t+dur) is:

[0083]

[0084] Among them, P5(t, t+dur) represents the output power of the ventilation fan in the time period (t, t+dur), C r and C target respectively represent the actual value and the target value of the gas, C in and C out respectively represent the gas concentrations inside and outside the pig house, S represents the cross-sectional area of the air inlet, V represents the volume of the breeding house, dur represents the operation duration of the equipment, and p represents the increase in the output power of the fan corresponding to a 1 m / s increase in the air inlet wind speed;

[0085] The model of the ventilation fan is specifically the 1100-type negative pressure fan;

[0086] S402. When adjusting the temperature, humidity and harmful gases in the pig house, in view of the situation that the adjustment processes of various environmental parameters affect each other, the adjustment strategy is set with the priority from high to low as harmful gases, temperature and humidity;

[0087] Step S5 includes the following steps:

[0088] S501. The fuzzy control technology is as follows: First, determine the initial input quantity of the fuzzy controller, then determine the basic domain of the initial input quantity, determine the quantization factor of fuzzification through the basic domain, and finally obtain the final input quantity of the fuzzy controller through the conversion formula to complete the fuzzification operation of the input quantity, and then accurately control the output quantity;

[0089] The fuzzy controller is specifically a controller of the Mikro intelligent fuzzy PID controller model MIK2300;

[0090] S502. The initial input quantity is to obtain the real-time value of the material falling amount by using a weighing sensor, compare the automatically set material falling amount with the material falling amount obtained from the weighing sensor, and obtain the percentage of material falling amount error E and the error change rate EC after the comparison;

[0091] The model of the weighing sensor is specifically the American SunCells HSX-SS-200KG weighing sensor;

[0092] Step S5 also includes the following steps:

[0093] S503. Determine that the continuous value range of the blanking quantity error percentage E is [eL, eH], and the basic domain is defined as {-a, -a + 1, ···, -1, 0, 1, ···, a - 1, a}. The quantization factor ke of the blanking quantity error percentage E is:

[0094]

[0095] where a, eL, and eH are all constant values;

[0096] S504. Determine that the continuous value range of the error change rate EC is [ecL, ecH], and the basic domain is defined as {-b, -b + 1, ···, -1, 0, 1, ···, b - 1, b}. The quantization factor kec of the error change rate EC is:

[0097]

[0098] where b, ecL, and ecH are all constant values. Let the time for obtaining the blanking quantity in one cycle be m seconds, then

[0099] S505. Determine that the continuous value range of the blanking time quantity U is [uL, uH], and the basic domain is defined as {-c, -c + 1, ···, -1, 0, 1, ···, c - 1, c}. The quantization factor ku of the blanking time quantity U is:

[0100]

[0101] uH - uL = km

[0102] where c, uL, and uH are all constant values, and k is the cycle coefficient of the blanking quantity;

[0103] S506. Substitute the quantization factor ke, the quantization factor kec, and the quantization factor ku into the conversion formula respectively to complete the fuzzy calculation of the error percentage E, the error change rate EC, and the blanking time quantity U. The conversion formula is:

[0104]

[0105]

[0106] where E * is the error percentage of the fuzzy operation, EC * is the error change rate of the fuzzy operation, and U * is the blanking time quantity after precise control;

[0107] The pig breeding production management system based on artificial intelligence includes the following units:

[0108] Growth correlation unit, which inputs the collected pigsty environmental data, piglet growth data, and piglet feeding data into the SVM model for training, thereby establishing the relationship between the feed ratio of piglets, the living environment of piglets, and the growth curve of piglets, determining the feed amount and feed ratio according to the SVM model parameters, and determining the temperature threshold, humidity threshold, and gas concentration threshold of the pigsty environment according to the piglet growth data;

[0109] Device startup unit, which inputs the environmental data of the temperature, humidity, and harmful gas concentration collected by the temperature sensor, humidity sensor, and gas sensor in the pigsty into the GRU model to predict the environmental data for the next 24 hours, thereby controlling the operation of the device;

[0110] Environmental regulation unit, which monitors the environmental changes in the pigsty in real time, and dynamically adjusts the output power of relevant devices when obtaining the environmental prediction data of the temperature, humidity, and harmful gas concentration in the pigsty;

[0111] Feed control unit, which uses fuzzy control technology on the feeding equipment to provide precise feed amount for piglets, selects the percentage of the error in the falling material amount and the change rate of the error as the input of the fuzzy controller, and selects the falling time amount of the feed feeding equipment as the output of the fuzzy controller.

[0112] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0113] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for pig farming production management based on artificial intelligence, characterized in that, The method includes: S1. Determine the influencing factors of piglet growth: Input the collected pigsty environmental data, piglet growth data, and piglet feeding data into the SVM model for training, so as to establish the relationship between the feed ratio of piglets, the living environment of piglets, and the growth curve of piglets; S2. Determine the feed ratio and environmental thresholds: Calculate the daily nutritional value required by piglets according to the SVM model parameters, determine the feed amount and feed ratio according to the nutritional components contained in the feed and the amount of each nutritional component, and determine the temperature threshold, humidity threshold, and gas concentration threshold of the pigsty environment according to the piglet growth data; S3. Obtain future environmental data: Input the collected pigsty environmental data into the GRU model to predict the temperature change curve T(t), humidity change curve H(t), carbon dioxide concentration change curve C(t), and ammonia concentration change curve N(t) in the next 24 hours. If environmental abnormalities occur, the controller formulates strategies to operate the equipment; S4. Dynamically adjust the pigsty environment: Monitor the environmental changes in the pigsty in real time. After obtaining the environmental prediction data of the temperature, humidity, and harmful gas concentration in the pigsty, dynamically adjust the output power of relevant equipment; S5. Precisely control the feed amount: Use fuzzy control technology on the feeding equipment to provide precise feed amount for piglets. Select the percentage error E of the material dropping amount and the error change rate EC as the input of the fuzzy controller, and select the material dropping time amount U of the feed feeding equipment as the output of the fuzzy controller.

2. The method for pig farming production management based on artificial intelligence according to claim 1, characterized in that: In step S1, the RFID technology is used to collect various indicators of piglets. The indicators include the pig number, body temperature, body weight, and backfat thickness of the pig. The body weight of the piglet is collected when the piglet is quiet. When the piglet is half-way through eating and completely relaxed, the body temperature and backfat thickness of the piglet are collected. When the piglet finishes eating, the remaining feed in the feeding trough is weighed, and then the remaining amount is compared with the previously configured feed amount to obtain the feed amount per meal of the piglet. The environmental data of the temperature, humidity, and harmful gas concentration in the pigsty are collected through temperature sensors, humidity sensors, and gas sensors.

3. The method for pig farming production management based on artificial intelligence according to claim 1, wherein: In step S2, record the basic information of each pigsty. The basic information of the pigsty includes the pigsty number, pigsty space size, pigsty environmental data, and the maximum capacity of piglets that the pigsty can accommodate. The size of the piglet activity space is proportional to the growth of piglets, and the size of the piglet activity space is inversely proportional to the cost. Therefore, the maximum capacity of the pigsty is determined according to the growth status of piglets and the cost to ensure the breeding efficiency.

4. The method for pig breeding production management based on artificial intelligence according to claim 1, characterized in that: Step S3 includes the following steps: S301. Compare the temperature change curve T(t) with the set high temperature threshold Th and low temperature threshold Tl. If Tl < T(t) < Th, it means that the temperature is normal. If T(t) ≤ Tl, it means that a low temperature abnormality has occurred, and the controller starts to formulate a heating strategy. If T(t) ≥ Th, it means that a high temperature abnormality has occurred, and the controller starts to formulate a cooling strategy; S302. Compare the humidity change curve H(t) with the set high humidity threshold Th and the drying threshold Tl. If Hl < H(t) < Hh, it indicates that the humidity is normal. If H(t) ≤ Hl, it indicates a low humidity anomaly, and the controller starts to formulate a humidification strategy. If H(t) ≥ Hh, it indicates a high humidity anomaly, and the controller starts to formulate a dehumidification strategy. S303. Compare the carbon dioxide concentration change curve C(t) and the ammonia concentration change curve N(t) with the respectively set gas concentration thresholds. If only carbon dioxide exceeds the gas concentration threshold, formulate a ventilation strategy based on the concentration of carbon dioxide. If only ammonia exceeds the gas concentration threshold, formulate a ventilation strategy based on the concentration of ammonia. If both carbon dioxide and ammonia exceed the gas concentration threshold, calculate the anomaly coefficients of carbon dioxide and ammonia, and formulate a ventilation strategy based on the gas with the larger anomaly coefficient. Here, the anomaly coefficient is used to measure the anomaly degree of harmful gases.

5. The method for pig breeding production management based on artificial intelligence according to claim 1, characterized in that: The step S4 includes the following steps: S401. Execute the strategies of heating, cooling, humidifying, dehumidifying, and ventilating according to the instructions of the controller, and control the output power of each strategy during the equipment execution stage. S402. When adjusting the temperature, humidity, and harmful gases in the pig house, considering the mutual influence of the adjustment processes of various environmental parameters, set the priority of the adjustment strategies from high to low as harmful gases, temperature, and humidity.

6. The method for pig breeding production management based on artificial intelligence according to claim 1, characterized in that: The step S5 includes the following steps: S501. The fuzzy control technology is as follows: First, determine the initial input quantity of the fuzzy controller, then determine the basic domain of the initial input quantity, determine the quantization factor of fuzzification through the basic domain, and finally obtain the final input quantity of the fuzzy controller through the conversion formula to complete the fuzzification operation of the input quantity, and then accurately control the output quantity. S502. The initial input quantity is the value of the real-time material falling amount obtained by using a weighing sensor. Compare the automatically set material falling amount with the material falling amount obtained from the weighing sensor. After the comparison, obtain the percentage of the material falling amount error E and the error change rate EC.

7. The method for pig breeding production management based on artificial intelligence according to claim 6, characterized in that: The step S5 also includes the following steps: S503. Determine that the continuous value range of the percentage of the material falling amount error E is [eL, eH], and the basic domain is defined as {-a, -a + 1, ···, -1, 0, 1, ···, a - 1, a}. The quantization factor ke of the percentage of the material falling amount error E is: where a, eL, and eH are all constant values; S504. Determine that the continuous value range of the error change rate EC is [ecL, ecH], and the basic domain is defined as {-b, -b + 1, ···, -1, 0, 1, ···, b - 1, b}. The quantization factor kec of the error change rate EC is: where b, ecL, and ecH are all constant values. Let the time for obtaining the blanking amount in one cycle be m seconds, then S505. Determine that the continuous value range of the material falling time quantity U is [uL, uH], and the basic domain is defined as {-c, -c + 1, ···, -1, 0, 1, ···, c - 1, c}. The quantization factor ku of the material falling time quantity U is: uH - uL = km; where c, uL, and uH are all constant values, and k is the cycle coefficient of the material falling amount. S506. Substitute the quantization factors ke, kec, and ku into the conversion formula respectively to complete the fuzzy calculation of the error percentage E, the error change rate EC, and the blanking time quantity U. The conversion formula is as follows: Among them, E * is the percentage error of the fuzzification operation, EC * is the rate of change of the error of the fuzzification operation, and U * is the blanking time amount after precise control.

8. The pig breeding production management system based on artificial intelligence is characterized in that The pig breeding production management system is applicable to the artificial intelligence-based pig breeding production management method described in claims 1-7, and includes the following units: Growth correlation unit. The growth correlation unit inputs the collected pigsty environmental data, piglet growth data, and piglet feeding data into the SVM model for training, so as to establish the relationship between the feed ratio of piglets, the living environment of piglets, and the growth curve of piglets. Determine the feed quantity and feed ratio according to the SVM model parameters, and determine the temperature threshold, humidity threshold, and gas concentration threshold of the pigsty environment according to the piglet growth data; Equipment startup unit. The equipment startup unit inputs the environmental data of the pigsty temperature, humidity, and harmful gas concentration collected by the temperature sensor, humidity sensor, and gas sensor into the GRU model to predict the environmental data for the next 24 hours, so as to control the operation of the equipment; Environmental regulation unit. The environmental regulation unit monitors the environmental changes in the pigsty in real time. When obtaining the environmental prediction data of the temperature, humidity, and harmful gas concentration in the pigsty, it dynamically adjusts the output power of the relevant equipment; Feed control unit. The feed control unit uses fuzzy control technology on the feeding equipment to provide accurate feed quantity for piglets. Select the blanking quantity error percentage and the error change rate as the input quantities of the fuzzy controller, and select the blanking time quantity of the feed feeding equipment as the output quantity of the fuzzy controller.

Images