JP2021190114A - Livestock information management system, livestock information management method and livestock information management program - Google Patents

Abstract

To perform breeding management toward a target productivity by sequentially acquiring current productivity of livestock for breeding.SOLUTION: A livestock information management system 1 uses breeding livestock basic information to be associated with breeding livestock and including at least breed information of the breeding livestock, rearing environment information including information relating to at least an ambient temperature of a farm acquired with a measuring device, rearing history information relating to a rearing history of the breeding livestock including at least history information of litter sizes of the breeding livestock, information relating to the at least ambient temperature of the farm included in the rearing environment information, the breed information included in the breeding livestock basic information and the history information of the liter sizes of the breeding livestock included in the rearing history information, as learning data to learn them into a mathematical model, and estimates the litter sizes of the breeding livestock at any future timing by using the breeding mathematical model.SELECTED DRAWING: Figure 1

Description

This disclosure relates to a livestock information management system, a livestock information management method, and a livestock information management program.

In recent years, livestock information, which is the result of livestock and its products that contribute to farm management, is managed from paper ledger management and electronic information management on general-purpose PCs to management on a Web server that communicates via the Internet line, so-called cloud management. We are entering the era.

For example, in the pig farming industry, there is known a livestock management system that can reduce the input work of livestock data for breeding management and ensure the breeding environment of livestock and traceability from birth to shipment from accumulated data (Patent Document 1).

Patent No. 5464619

However, the technique described in Patent Document 1 merely digitizes livestock information, and only suggests the optimization of livestock management. Therefore, the farmers themselves must consider specific measures to improve productivity using the livestock information.

In particular, breeding livestock breeding measures play an important role in livestock, and in order to bring out the potential productivity of breeding livestock, the current productivity of breeding livestock should be grasped one by one and the target should be. It is necessary to manage the breeding for productivity.

The present disclosure has been made to solve such problems, and the purpose of the present disclosure is to sequentially grasp the current productivity of breeding livestock by using the breeding environment information obtained from various measuring devices. The purpose is to provide an automatic livestock management system, an automatic livestock management server, an automatic livestock management method, and an automatic livestock management program that can manage breeding toward the target productivity.

In order to achieve the above-mentioned object, the livestock information management system according to the present disclosure includes a measuring device for acquiring breeding environment information regarding the breeding environment of the breeding livestock installed on the farm where the breeding livestock is raised, and livestock. Basic information on breeding livestock including at least breeding livestock information linked to breeding livestock stored in the information storage unit, and breeding environment information including at least information on the temperature of the farm acquired by the measuring device. Breeding history information regarding the breeding history of breeding livestock, including at least history information on the number of offspring of breeding livestock stored in the livestock information storage unit, at least information on the temperature of the farm included in the breeding environment information, and breeding livestock. By learning from the mathematical model using the breed information included in the basic information and the history information of the number of offspring of breeding livestock included in the breeding history information as learning data, the number of offspring by breeding livestock can be obtained. Using a mathematical model construction unit that generates a breeding mathematical model for estimation, a mathematical model storage unit that stores the breeding mathematical model generated by the mathematical model construction unit, and a breeding mathematical model stored in the mathematical model storage unit, It is equipped with an offspring number estimation unit that estimates the number of offspring of breeding livestock at any timing in the future, and the offspring number estimation unit estimates the number of offspring of breeding livestock using a breeding actuarial model.

Further, in order to achieve the above-mentioned object, the livestock information management method according to the present disclosure includes a measuring device for acquiring breeding environment information regarding the breeding environment of the breeding livestock installed on the farm where the breeding livestock is raised. , Breeding livestock basic information including at least breeding livestock breeding information linked to breeding livestock stored in the livestock information storage unit, and breeding environment information including at least farm temperature information acquired by the measuring device. , The mathematical model construction unit is included in at least the breeding environment information using the breeding history information regarding the breeding history of the breeding livestock, including at least the history information of the number of offspring of the breeding livestock stored in the livestock information storage unit. For breeding by learning from a mathematical model using the temperature information of the farm, the breed information included in the basic information on breeding livestock, and the history information on the number of offspring of breeding livestock included in the breeding history information as learning data. A mathematical model construction step to generate a breeding actuarial model for estimating the number of offspring by birth history of livestock, and a future arbitrary by using the breeding actuarial model generated by the breeding actuarial model construction department by the offspring number estimation unit. A step of estimating the number of offspring for estimating the number of offspring of breeding livestock at the timing of is provided.

In addition, in order to achieve the above-mentioned objectives, the livestock information management program according to the present disclosure is a measuring device for acquiring breeding environment information regarding the breeding environment of breeding livestock installed on the farm where breeding livestock are raised. , Breeding livestock basic information including at least breeding livestock breeding information linked to breeding livestock stored in the livestock information storage unit, and breeding environment information including at least farm temperature information acquired by the measuring device. The Mathematical Model Construction Department includes at least the breeding environment information using the breeding history information regarding the breeding history of the breeding livestock, including at least the history information of the number of offspring of the breeding livestock stored in the livestock information storage unit. Breeding by learning from a mathematical model using the temperature information of the farm, the breed information included in the basic information on breeding livestock, and the history information on the number of offspring of breeding livestock included in the breeding history information as learning data. A mathematical model construction step to generate a breeding actuarial model for estimating the number of offspring by birth history of domestic livestock, and a breeding actuarial model generated by the breeding actuarial model construction department by the offspring number estimation unit in the future. Have the computer perform a step of estimating the number of offspring to estimate the number of offspring of breeding livestock at any time.

According to the livestock information management system, livestock information management server, livestock information management method, and livestock information management program of the present invention using the above means, the current production of breeding livestock using the breeding environment information obtained from the measuring device. It is possible to grasp the sex one by one and manage the breeding toward the target productivity.

It is a system block diagram which shows the livestock information management system which concerns on this embodiment. It is a figure for demonstrating the breeding cycle. It is a figure explaining the process performed by the mathematical model construction part which concerns on this embodiment. It is a schematic diagram which shows an example of a breeding performance model. It is a figure explaining the process performed by the child-bearing number estimation part which concerns on this embodiment. It is a figure explaining the process performed by the configuration optimization unit which concerns on this embodiment. It is a figure explaining the process performed by the candidate livestock number calculation part which concerns on this embodiment. It is a conceptual diagram explaining an example of cooperation of each configuration. It shows an example of a screen displaying the optimum livestock composition information, abandoned livestock candidates, maintenance livestock candidates, and the number of candidate livestock required as newly introduced breeding livestock. It is a flowchart explaining the flow of the process of mathematical model generation. It is a figure which shows an example of the estimation of the number of births. It is a flowchart explaining the flow of the alert issuance processing. It is a figure which shows an example of the alert issuance management screen. It is a flowchart explaining the flow of process of the product history composition optimization. It is a schematic block diagram which shows the structure of a computer.

Hereinafter, each embodiment will be described with reference to the drawings. It should be noted that it is a domestic pig in the embodiment of the present disclosure described below. In embodiments, pigs, which are livestock products, are bred on farms. There are multiple barns on the farm where livestock products are bred. The barn is also called a pig barn, especially in the case of pigs. A barn, which is a unit of a partitioned section in a barn, is also called a pig bunch, especially in the case of pigs. When the pig is a female, the breeding livestock is also referred to as a mother pig (or a breeding mother pig), and when the pig is a male, it is also referred to as a sow (or a breeding sow), a breeding pig, or a breeding sow.

In addition, a farmer is an individual or a corporation that is a farmer who raises livestock, and is also referred to as a livestock farmer or a livestock farmer. Basically, he is a pig farmer and is also called a pig farmer. These farmers and farmers are users of the livestock information management system.

First, the outline and usefulness of productivity management of sows as breeding livestock, which are carried out using the embodiments of the present disclosure, will be described. In the pig farming industry, mother pigs play an important role in producing piglets that become meat pigs. Overseas, breeding is repeated, and a large number of piglets are produced in one calving, and efficient calving is carried out in one year to improve the fertility. However, in Japan, the breeding performance is low despite the introduction of improved breeding pigs overseas and the same production potential as in Europe and the United States. Therefore, it is required to bring out the fertility that the breed of mother pig originally possesses.

In addition, after a certain number of deliveries, the sow is renewed (it is disused after finishing its role as a sow and replaced with a new sow). If this renewal is not carried out properly, high-yielding sows tend to have fewer offspring than low-yielding sows, leading to poor farmer profits. However, since the depreciation period stipulated by the tax law is set for sows, if the period is not exhausted and is abolished, the period that can be deducted will be shortened, which will worsen the actual profit of the farmer. It leads to.

In the embodiment of the present disclosure, the breeding environment information and the breeding history information are also obtained by using a measuring device or the like, and these are learned as training data by a mathematical model represented by a statistical model or a machine learning model. Predict the number of offspring that a mother pig can produce. In addition, as an intermediate index indicating the state of the breeding process such as the weight and physical condition of the mother pig, the combination of these intermediate indexes that improve the reproductive performance is estimated by artificial intelligence, machine learning, or the like. That is, providing this intermediate index to farmers leads to control of the intermediate index and improvement of the final index, reproductive performance, specifically, for example, the number of offspring.

Furthermore, the number of future births and the number of rotations of the breeding cycle are estimated for each mother pig using the number of birth history of the mother pig and the past breeding performance information, and the number of future births and the mother pig for each piggery and farm. Predict the birth history composition of. In addition, the number of offspring of standard sows in the birth history is estimated from the past reproductive performance information and used for comparison.

Using this, we will identify the mother pigs whose number of offspring is lower than the standard range in the number of births for each individual, and at the same time, necessary breeding of candidate pigs to be newly introduced as alternative breeding livestock. It is possible to estimate the number, estimate the birth history composition that optimizes the number of offspring and profits in the future, and provide the information to the farmers. As a result, the mother pig can be appropriately renewed and the management efficiency of the farmer can be improved.

<Structure>
FIG. 1 is a system configuration diagram showing a livestock information management system 1 including a livestock information management server 101 according to an embodiment of the present disclosure. As shown in FIG. 1, in the livestock information management system 1 according to the embodiment of the present disclosure, the livestock information management server 101 and the terminal device 201 used by the user are connected to each other so as to be communicable via the network NW. .. The network NW is, for example, a network such as the Internet or a VPN (Virtual Private Network). For the sake of simplification of the explanation, FIG. 1 shows only the terminal device 201 assuming one user, but the livestock information management server 101 includes two or more terminal devices and two or more users via the network NW. It is possible to connect. Further, one terminal device may be used by a plurality of users.

The livestock information management system 1 is a system mainly for supporting the management of livestock information by farmers. The operator of the livestock information management system 1 provides a service for supporting the management of livestock information to the farmer using the system. The operator can provide services to multiple farmers. The livestock information management server 101 is managed by the operator of the livestock information management system 1, and the farmer uses the terminal device 201. The livestock information management server 101 may be managed by the farmer himself.

The livestock information management system 1 is a system mainly for supporting the management of livestock information by farmers. The operator of the livestock information management system 1 provides a service for supporting the management of livestock information to the farmer using the system. The operator can provide services to multiple farmers. The livestock information management server 101 is managed by the operator of the livestock information management system 1, and the farmer uses the terminal device 201. The livestock information management server 101 may be managed by the farmer himself.

The configuration of the livestock information management system 1 is roughly divided into four systems. Sensor system for obtaining livestock information, information generated by the sensor, and data system that is a storage unit thereof, processing system that processes the obtained and stored information to generate new information, processing It is a terminal system for farmers and agricultural workers to confirm and input the output of the system. It should be noted that these four lines are presented for the sake of understanding of the present disclosure, and not all configurations belong to any of these four lines.

(Terminal system)
The terminal device 201 is a device such as a PC, a smartphone, a tablet PC, and a mobile phone. The terminal device 201 may access the livestock information management server 101 by the dedicated application software installed in the terminal. Further, the livestock information management server 101 may be accessed by using the operating environment (API (application programming interface), platform, etc.) provided by the livestock information management server 101.

The input unit 211 is a device capable of inputting and selecting information by being operated by a user such as a keyboard, a mouse, and a touch pad. The input and selection of information includes, for example, input by inputting numerical values, characters, and symbols, selection of items, and input for determining processing. Further, a touch panel integrated with a display unit 212 such as a liquid crystal display or an organic EL display in a smartphone, a tablet, or a PC may be used. Further, the input unit 211 may be a voice input device.

The display unit 212 is a display device or the like that displays information or the like to the user. It may be a display device independent of the terminal device 201, or it may be a display device such as a liquid crystal display or an organic EL display in a smartphone or tablet.

The terminal device 201 may be configured integrally with the livestock information management server 101 without going through the network NW.

(Sensor system)
The barn 301 is provided with a measuring device 310 for obtaining information on the breeding environment in the barn, a feeding device 312 for feeding livestock, and the like, and these are used to raise various breeding environment information through the network NW. It is transmitted to the information management server 101.

Mother pigs are generally bred in multiple stalls that are constructed in a barn, which is a building on the farm. The stall is an environment and place for raising one mother pig.

The barn 301 is, for example, a pig barn, and a feeding device 320 is provided for each stall. In the case of a free stall piggery, one or more feeding devices 320 are provided at known locations within the piggery. The feeding device 320 is a device for feeding and watering pigs. The feeding device 320 can automatically supply a predetermined amount of food and water at a predetermined time for each stall. It should be noted that the cumulative amount of salary and the like can be measured not only in the case of restricted feeding but also in the case of constant feeding. Further, the feeding device 320 supplies the type and amount of food (feeding amount: weight (g) or volume (l)), the amount of water (water supply amount: (l)), and the supply time (year) supplied for each stall. It is possible to generate feeding and water supply information including information such as date, time), etc. The generated feeding and water supply information is transmitted to the livestock information management server 101 through various breeding environment information through the network NW.

The image capturing unit 311 is a kind of measuring device 310, and is, for example, a camera capable of obtaining an image in a barn. The camera is, for example, a visible light camera, and has a function of detecting light reflected from a subject and generating image (still image or moving image) information. The camera may be a combination of an infrared camera (thermal camera) and an infrared light in addition to the visible light camera so that nighttime imaging and daytime body surface temperature measurement can be performed.

As the other measuring device 310, an environmental information acquirer such as a microphone, an odor sensor, a temperature / humidity sensor, an air flow rate sensor, a pressure pressure sensor, and an illuminance sensor can be used.

The microphone has a function of collecting sound generated in the vicinity of the installation location, converting it into an electric signal, and generating sound information. The microphone is installed at a position (height of the mother pig's daily life) where the sound emitted in the space where the pig is located can be detected. Sounds detected by the microphone include pig sounds, breath sounds, coughing, sneezing, and moving sounds. The sound can be measured in decibel (dB) units, and the number of times / cumulative time / continuous time when the sound becomes equal to or higher than a predetermined voice waveform pattern or a predetermined dB value can be totaled.

The odor sensor detects ammonia, methyl mercaptan, etc. that cause odor and generates odor information. The odor detected by the odor sensor includes the odor emitted from the pig itself, the odor of pig excrement, the odor of the barn, and the like. The odor can be measured quantitatively (ppm), for example, by the sensor reading value of the odor sensor.

The temperature / humidity sensor has a function of detecting the temperature and humidity of the place where the sensor is installed and generating temperature / humidity information. The temperature / humidity sensor may be a temperature sensor and a humidity sensor capable of measuring temperature and humidity, respectively. The temperature sensor can measure the temperature of the place where it is installed in degrees Celsius or degrees Fahrenheit. The humidity sensor can measure, for example, relative humidity as (%) or (% RH).

The air flow rate sensor has a function of detecting the air flow rate entering and exiting the installed space and generating air flow rate information. For example, it is an anemometer, which can be installed indoors or outdoors in a barn and can measure an air volume (m / s).

The barometric pressure sensor has a function of detecting the barometric pressure in the installed space and generating barometric pressure information. The barometric pressure sensor can measure, for example, the barometric pressure (Pa) around the barn.

The illuminance sensor measures a predetermined place in the barn, for example, the barn where breeding livestock live, the brightness of the stall, the brightness (cd / m2), and the illuminance (W / cm2, lux, lm / m2). be able to. It is also possible to measure and store these average values and cumulative values.

There is one or more barns 301 in the farm, and for the sake of simplicity, only one barn 301 is shown in FIG. 1, but the livestock information management server 101 has two or more barns via the network NW. It is possible to connect to two or more farms.

The livestock information management server 101 includes a child birth number estimation unit 111, a livestock state estimation unit 112, an alert generation unit 113, a mathematical model construction unit 114, a mathematical model management unit 115, a management accounting unit 116, a configuration optimization unit 117, and a candidate livestock. It includes a number calculation unit 118, an information acquisition unit 119, and an information storage unit 120.

(Data system)
The information storage unit 120 is a storage device that stores information used for executing management by the livestock information management system 1. Although it is described integrally with the livestock information management server 101 in FIG. 1, it may be a storage device physically independent of the livestock information management server 101. Hereinafter, the information stored in the information storage unit 120 will be described.

The information storage unit 120 mainly includes a livestock information storage unit 121, a mathematical model storage unit 122, a financial basic information storage unit 123, and an alert condition storage unit 124. The information storage unit 120 also stores other information that does not belong to these storage units.

The livestock information storage unit 121 stores basic livestock information for breeding, breeding environment information, and breeding history information.

The basic information on livestock for breeding is basic information on sows and boars (or semen) that are breeding livestock raised on farms. For example, the identification ID of the breeding livestock, which is an electronic identification for identifying each individual, and the information on the breed of the breeding livestock, the source of the breeding livestock, and the information related to the identification ID of the breeding livestock. Information on the age of breeding livestock (which may be age or age), information on the shipment performance of offspring, and the like. In addition, each mother pig is like a physical mother pig identifier such as an ear tag, an IC tag or a tattoo, an identifier given to a bred stall, and the above-mentioned identification ID stored in the information storage unit 120. It is managed on the premise that it corresponds to various electronic identifiers. These breeding livestock basic information is input by a farmer who is a user using, for example, a terminal device 201, and is stored in the livestock information storage unit 121.

Explain the basics of the technical significance of breeding livestock information. The breeding livestock identification ID is used as data for identifying and distinguishing individual breeding livestock when various information on the breeding livestock is used as learning data when constructing various analyzes and mathematical models. .. Information on breeding livestock varieties, which will be described later, is stored for learning as a type of genetic factor that affects the number of offspring because the characteristics of each breed differ. The day age (age / month age) of breeding livestock is memorized for learning as one of the factors that affect the number of offspring because the fertility changes with aging.

The information on the shipment performance of the offspring may include information such as the number of offspring shipped for meat in the past, so-called meat pigs, the shipping time, and the selling price.

A breed is a breed or pedigree of livestock, and corresponds to a subordinate concept rather than a breed of livestock such as cattle, pigs, and sheep. Specifically, in addition to varieties such as Landrace, Duroc, Large White, Berkshire, and Hampshire, hybrids obtained by multiplying these pure species are included. For example, a hybrid of Landrace (L), Large White (W), and Duroc (D) is included. Regarding hybrids, fattening pigs such as ternary hybrids such as LWD, so-called ternary pigs, are known as offspring, but as mother pigs, they are often binary pigs (three offspring). If you are a former pig). As the breed information of the hybrid species, the pedigree ratio (%) or the pedigree table may be used. Further, the crossed species may be breed information such as LW and WL that can identify the sow and the crossed generation. For example, there are "Landrace pigs" that have excellent fertility and fast growth, and "Dual pig" mother pigs that were born by crossing large white-haired breeds of "Large White pig" that have a good balance of lean meat and fat. It is conceivable to multiply such a mother pig with a boar of the red pig "Duroc breed", which tends to have sashimi (marbled meat) as a meat quality.

The breeding environment information is, for example, temperature, humidity, air volume, light intensity, sound, odor, and the like. It is also a farm and a barn. Quantitative values of temperature, humidity, air volume, light volume, sound, and odor can be obtained from the above-mentioned various measuring devices 310. These acquisition frequencies are determined by the acquisition frequency information stored in the information storage unit 120, and are periodically acquired by the information acquisition unit 119, which will be described later. The breeding environment information is basically automatically acquired and stored, but may be input and stored by a farmer who is a user using, for example, the terminal device 201. Farms and barns can be obtained as farm IDs and barn IDs.

Explain the basic technical significance of breeding environment information. Air temperature, specifically air temperature, is a typical example that can represent the breeding environment of livestock for breeding. If the temperature inside the barn deviates from a predetermined range, it will ultimately affect the number of offspring. For example, when the temperature is above a predetermined level, heat stress may occur for the sow. Pigs have poorly developed sweat glands and are covered with thick fat, so they have a poor ability to diffuse metabolic heat outside the body and tend to be vulnerable to heat. Heat stress affects productivity and physiology, and phenomena include decreased feed intake, weight loss during pregnancy, milk loss during lactation, delayed estrus recurrence, and decreased conception rate. Etc. will eventually affect the number of offspring. In addition, even when the temperature falls below a predetermined amount, feed intake increases, water consumption decreases, and excessive tension occurs in order to maintain metabolic energy. Drinking water also has an effect on miscarriage and abnormal delivery.

Humidity, like air temperature, is a factor in heat stress. The above-mentioned metabolic heat is controlled by the body temperature center in the pituitary gland in relation to the outside air temperature and humidity. The relative humidity of the air also affects the release of heat by the heat of vaporization.

Air volume can be one of the factors that cause a decrease in blood flow. For example, when a cold, dry wind hits a sleeping mother pig directly, it causes a cold stomach, a decrease in drinking water, a decrease in urination volume, a decrease in circulating blood volume, and nutritional support for the fetus and other offspring. It contributes to the shortage and affects abnormal delivery and weight loss of the offspring.

The amount of light has an effect on the endocrine function of luteinizing hormones such as progesterone, and affects the conception rate and the like. As for sound, the load of noise has an effect on the endocrine function as well as the amount of light.

The odor is generated by the respiration of pigs, radiator heat, water vapor generated from manure, ammonia gas, etc., leading to disease morbidity and affecting reproductive performance. Specifically, air unsanitary environmental substances such as ammonia irritate the mucous membranes and cause respiratory diseases (pig breeding / respiratory disorder syndrome: PRRS, etc.). Therefore, since air pollution in the piggery causes the occurrence of respiratory diseases, it is important to measure the odor value to quantitatively obtain whether or not proper ventilation is provided.

Regarding the breeding environment information that can be obtained from such farm measuring equipment, temperature (air temperature) is important, but the accuracy of prediction can be improved by adding only one element from humidity, air volume, light volume, sound, and odor. improves.

Breeding history information includes disease morbidity history of breeding livestock (for example, date and disease name), medication history for breeding livestock (for example, date and medication type, and medication type used for treatment or prevention of disease). Or, including those used to promote reproduction such as gonad stimulating hormones), history of vaccine administration to breeding livestock (for example, date and vaccine type), information on the number of births of breeding livestock, birth of breeding livestock. Historical information on numbers (eg, number of offspring, number of surviving offspring, average weight / dispersion of the most recent offspring, number of weaning offspring, and their cumulative total), history of normal or abnormal birth of breeding livestock (eg) Abnormal delivery of shirako, kuroko, mummy fetal, frail child), estrus, recurrence rate, delivery rate, seeding date for breeding livestock, seeded semen, semen properties (eg sperm vitality, sperm concentration, semen volume) , Total sperm count, etc.) information, weight of breeding livestock, information on physical condition of breeding livestock (for example, backfat thickness (cm), body color (RGB), and body condition score information of breeding livestock), breeding Information on the type and / or amount of feed (feed feed (g)) for livestock for domestic use, information on the amount of water supply (l) for livestock for breeding, and the like. These pieces of information are stored in association with the timing at which the information was generated. Timing is time and date.

The history of disease morbidity of breeding livestock and the history of vaccination to breeding livestock relate to the health condition of sows and have various influences such as the number of offspring and the conception rate.

Regarding dosing, for example, administration of a hormone preparation can be used as a means for reducing reproductive performance due to heat stress. In addition, depending on the type of feeding, there are some that claim the same effects as the medication, and it is considered that the administration and the presence or absence of feeding affect the number of offspring.

Information on the number of births of breeding livestock has an impact on the number of offspring, as mentioned above. Historical information on the number of offspring of breeding livestock (for example, the number of offspring, the number of surviving offspring, the average weight and dispersion of offspring following the latest birth, the number of weaning offspring, and the cumulative total of these) is used as past actual values. , It is important in predicting the number of offspring in the future. Regarding the normal or abnormal history of normal or abnormal calving of breeding livestock (for example, abnormal calving of shirako, kuroko, mummy fetal, frail child, etc.) It will be helpful about the possibilities.

The seeding date for breeding livestock is important in consideration of the time when the sow was bred and the environment during gestation. Information on the seeded semen and semen properties (eg, sperm vitality, sperm concentration, sperm volume, sperm count, etc.) is important as a factor of fertility on the boar side, not the mother pig.

Information on the weight of breeding livestock and the physical condition of breeding livestock is important information for considering the health condition of sows.

Information on the type and / or amount of feed for breeding livestock and information on the amount of water supplied to breeding livestock affect the weight and health of sows and are important information.

In addition, information regarding the breeding cycle of breeding livestock may be stored. The information on the breeding cycle of breeding livestock is information including the period required for repeated breeding events such as seeding, pregnancy, parturition, and weaning, and the history of the timing when the events occur. For example, if a breeding livestock of a certain breed has 114 days from calving to calving, 21 days from calving to weaning, and 5 days from weaning to the next seeding, timing information such as the period and date. including. Information about the breeding cycle varies greatly not only with individual differences, but also with the season, weight, physical condition, and breed at that time.

FIG. 2 is a schematic for schematically explaining the breeding cycle of sows. In this figure, the horizontal axis represents the time interval by taking the number of days. For example, in the mother pig 1, the gestation period from calving to calving, the lactation period from calving to weaning of the offspring, and the empty period from weaning to the next seeding are shown. Then, the breeding cycle is one cycle from the seeding to the next seeding in estrus. The predicted number of offspring p11 in the first cycle can be predicted by the number of offspring prediction unit. Similarly, the predicted number of offspring p12 in the second cycle can also be predicted. Here, in the first cycle and the second cycle, each period can be corrected so as to fluctuate depending on the variety, production order / history, aging, and season (year / month / day). Then, by using the period during which the sow can be active as a breeding livestock and these breeding cycles (for example, the number of days), future and comprehensive reproductive performance estimates such as lifetime (total) number of offspring can be estimated. Can be obtained. On the other hand, it is assumed that the breeding cycle of the mother pig 2 is different from that of the mother pig 1 due to, for example, different breeds (including combinations), different farms, and the like. Among them, the predicted number of offspring p21 in the first cycle and the predicted number of offspring p22 in the second cycle, which are different from those of the mother pig 1, are produced, and it is possible to obtain an estimate of the reproductive performance such as the lifetime number of offspring. Will be. In this figure, each period is schematically drawn for the sake of clarity, and does not necessarily match the actual number of days.

The weight of breeding livestock is the weight of breeding livestock obtained by various measuring devices. Generally, the body weight (kg) obtained by directly measuring the weight with a weight scale can be used. In addition, information on the body weight estimated by the livestock state estimation unit 112 can also be used by using the image information of the breeding livestock obtained by the image capturing unit 311. Weight information can be stored in association with the timing of measurement. For each breeding livestock, the daily weight gain and the date may be associated and memorized.

The weight information and the information on the physical condition of the breeding livestock can be used as the final index of the breeding livestock, for example, the feeding intermediate index information for the number of offspring.

The information on the physical condition of the breeding livestock is, for example, the body color and backfat thickness of the breeding livestock that can be generated by the livestock state estimation unit 112 by image analysis using the image information obtained by the image capturing unit 311. , The condition of a specific part (ear, abdominal dent, breast tension, swelling around the breast, shape of the genital area, etc.), body condition score, etc. be. These may be used as qualitative data together with the name. In addition, information such as sound and odor information obtained from the breeding environment information by the measuring device may be combined with the shape change of the sow that can be estimated from the above image as a physical condition parameter for estimating the health condition. The information generated by the livestock state estimation unit 112 may be stored in the livestock information storage unit 121 as breeding history information.

The mathematical model storage unit 122 stores mathematical models such as a standard breeding mathematical model, a correction coefficient, and an individual breeding mathematical model constructed by the mathematical model building unit 114. In addition, for these mathematical models, we store mathematical models according to the breed of livestock. The details of these mathematical models will be described later.

The basic financial information storage unit 123 stores basic financial information.

Basic financial information is basic financial information related to farm management. For example, the cost from the candidate pig to the start of breeding activity as a mother pig, the cost incurred for each individual related to disease morbidity measures such as vaccination, the cost for equipment installation, utility costs, labor costs, etc. Includes expenses such as distribution costs proportionally divided by the number of breeding livestock, and those that link information on depreciation to breeding livestock individuals. Therefore, for example, the feed amount, the water supply amount, the shipping weight, the feed conversion ratio, the feed unit price, etc. may be stored as basic financial information. Simply put, this basic financial information includes information for calculating the costs required for breeding used in performing the income and expenditure calculation of individual breeding livestock. Expenses are also called losses in tax accounting.

The information on depreciation of breeding livestock is information including the acquisition price of each breeding livestock, the useful life according to the use and details, the depreciation amount, the depreciation cost, the timing when depreciation is started, and the like.

In addition, the basic financial information storage unit 123 stores information such as farm facility and equipment information. This information can also be used as a constraint on the number of livestock that can be raised on the farm. For example, the number of barns, the area of each barn, the type, size, number of stalls of mother pigs, the number of pig bunches, the area, and the like. It is also possible to calculate the upper limit of the number of livestock that can be used by using these equipment information.

Further, the basic financial information storage unit 123 may store the number of heads shipped, the shipping amount, the shipping time, and the selling price as a sales record of meat livestock, so-called meat pigs, which are shipped and sold as meat among livestock. As the selling price, for example, in addition to the actual value, the average selling price in a unit time such as a day, a week, or a month may be stored. The selling price may be an index expressed by a price per unit weight, for example, (yen / kg). The average selling price per head may be stored so that the selling price per head for meat can be estimated. You may remember the profit amount after deducting the cost spent on raising meat and livestock. The average profit amount per head may be stored so that the profit amount per head for meat can be estimated. Such information on the selling price and the profit amount can be used to calculate the profit of the farmer obtained by selling the meat and livestock.

The alert condition storage unit 124 stores alert condition information, which is a condition when an alert should be issued for a breeding intermediate index and / or a final index for breeding livestock. Specifically, the alert condition information based on the breeding intermediate index is, for example, weight information and / or information on physical condition at a predetermined timing of breeding livestock. An example of information about physical condition is a strong factor that directs the number of offspring in the negative direction, such as abdominal dents and coughs. Further, the alert condition information based on the final index is, for example, information on the number of offspring estimated by a mathematical model described later. That is, the criteria for comparing the breeding intermediate index and / or the final index with these alert conditions are stored. Further, the alert condition storage unit 124 of the present embodiment also stores a re-alert condition for issuing an alert again to the target for which the alert has been issued once. The content of the re-alert condition may be the same as the alert condition, may be stricter than the alert condition, or may be relaxed.

(Processing system)
The information acquisition unit 119 has a function of acquiring information necessary for the following processing from the information storage unit 120 and the storage units belonging to these. Further, the measuring device 310 has a function of acquiring the breeding environment information at a frequency specified in the measuring device master information stored in the information storage unit 120.

The mathematical model construction unit 114 is a breeding mathematics for estimating the number of offspring of breeding livestock by processing information such as breeding livestock basic information, breeding environment information, and breeding history information as learning data in a mathematical model. It has the function of generating a model. The breeding mathematical model is a standard breeding mathematical model for estimating a standard breeding number based on information on multiple breeding livestock, for example, past breeding performance, and a mathematical model for factors that affect the number of breeding. It includes a correction factor generated so that it can be applied to the standard breeding model, and an individual breeding actuarial model for estimating the number of offspring of individual breeding livestock by applying individual factors such as the correction factor to the standard breeding actuarial model. These reproductive mathematical models can be generated using, for example, arbitrary methods and algorithms such as decision trees, regression analysis, and principal component analysis. Further, a trained model generated by machine learning or the like (DNN: deep neural network, RNN: recurrent neural network, SVM: support vector machine, random forest, etc.) can also be used.

As described above, there are various factors that affect the number of offspring of sows, and each one does not necessarily have an independent effect, but functions interactively. Therefore, it is more accurate to make predictions that take into account many factors rather than simply making predictions based on past reproductive performance. However, if the number of explanatory elements is increased, problems such as multicollinearity may occur. Therefore, it is useful to construct a trained mathematical model by selecting appropriate parameters from breeding livestock basic information, breeding environment information, breeding history information, etc. and performing machine learning.

As an example, but not a limitation, one example of combination of learning data for constructing a model in addition to the number of sows and their offspring is farm, breed, birth history, and temperature. For farms, the content of the breeding cycle (lactation period, estrus recursion period, feed composition during pregnancy, hormonal administration, etc.) will be incorporated into the mathematical model as different artificial variables. In addition, breeds and birth history will be incorporated into the mathematical model as internal reproductive variables of the sow derived from the genetic fertility inherently possessed by the individual breeding livestock (mother pig). Temperature will be incorporated into mathematical models as an environment variable that affects energy intake and metabolism. In addition, each of these variables does not act individually, but is linked to the reproductive performance as learning data in a form that includes the result of the interaction.

Further preferable examples of combinations of learning data are farm, breed, birth history, temperature, and humidity, in addition to the number of sows and their offspring. As for humidity, by combining it with air temperature, it is possible to incorporate the ease of heat release from the body into a mathematical model as an environment variable, such as the so-called heat index. It will be easier to express.

Also, by combining the mother pig and its calving number, breed, temperature, as well as weight or physical condition as learning data, is the condition sufficient to share the nutrients and minerals required from the mother to the foetation? In addition, since it can be incorporated into a mathematical model as an internal reproductive variable of the mother pig to grasp whether it is in a state where normal fetal growth and delivery such as an increase in the risk of late miscarriage and the risk of premature birth due to changes in body length can be performed, the mother. It becomes easier to express the condition of the pig.

In addition, by combining the origin or age of the sow as learning data in addition to the number of offspring, varieties, and temperature of the sow, the genetic properties and constitution inherently possessed by the sow, and Mathematical as an internal reproductive variable of the sow regarding the difficulty of pregnancy due to the aging of the sow, the increase in the abortion rate, and the deterioration of the reproductive performance due to aging including various abnormalities (motor function, digestive disorders, genital disorders, etc.) Since it can be incorporated into the model, it becomes easier to express the condition of the sow.

In addition to the number, breed, and temperature of the sow and its offspring, the type of feeding to the sow, the amount of feeding, and the medication history (used to promote reproduction) are combined as learning data to conceive the piglet. Are you ready to deliver as a normal offspring, fully recover after parturition and enter the next breeding cycle, and have sufficient energy and nutrients to give birth to a piglet during calving? Since it is possible to incorporate it into a mathematical model using the breeding variables such as whether or not the pigs are bred, it becomes easier to represent the environment in which the sows are bred.

In addition to the number of sows and their offspring, varieties, and temperature, the history of disease morbidity of sows, the history of vaccination for sows, the history of medication for sows (used for the treatment or prevention of diseases), and the birth of sows. By incorporating the normality or abnormality of the mother pig as learning data, it is possible to incorporate into the mathematical model the presence or absence of disease morbidity of the mother pig in the breeding cycle, the presence or absence of measures such as vaccines, etc. as the health condition variables of the mother pig. It is easier to express your health condition.

In addition to the number, breed, and temperature of the sow and its offspring, the breeding date for the sow, the semen that has been sown, and the semen property information are combined as learning data to make a breeding pair of sows. Since it is possible to incorporate the reproductive performance of the pig into a mathematical model as an external reproductive variable of the sow, it becomes easier to express the reproductive situation.

FIG. 3 is a diagram illustrating a process performed by the mathematical model construction unit according to the present embodiment. For example, in the standard breeding actuarial model, in order to estimate a standard breeding number model for a set of breeding livestock, the actual value of the offspring by birth history included in the breeding history information is used as the objective variable, and the breeding environment. Enter indicators other than information, feeding, water supply information, weight, information on physical condition, and breeding history information by birth history as explanatory variables. In this way, the mathematical model construction unit 114 uses the number of offspring as the objective variable, performs operations such as selection, combination, and weighting of the explanatory variables thereof, and outputs a mathematical model that can best explain the objective variable.

As a result of the analysis, the mathematical model construction unit 114 can cut out and generate factors that universally affect the number of offspring and factors that affect the number of offspring in common regardless of individual circumstances as correction coefficients. can. Correction factors include, for example, seasonal factors, past normal and abnormal delivery factors, feeding, water supply factors, etc., and these factors increase or decrease the estimated number of offspring. Correct the coefficient. As a seasonal factor, seeding and delivery time may be roughly a factor for qualitative data such as summer and winter, or may be a factor for quantitative data such as temperature and humidity. Factors due to past normal and abnormal calving are not based solely on the actual value of the number of offspring, but abnormal delivery such as a large number of piglets showing abnormalities (black pups, mummified fetuses, etc.) Numerical values such as rate can also be used. As a factor depending on the feeding and the amount of water supply, data such as how much the breeding livestock ate and drank water at a predetermined time before and after the seeding and the calving time may be a factor. The correction coefficient may be generated by cutting out not only the one that directly corrects the fluctuation (increase, decrease) of the number of offspring, but also the one that directly affects the number of offspring as a result. For example, as related to the breeding cycle such as the delivery time, it is possible to correct the number of days such as the shortening or lengthening of the lactation period or the empty period depending on the seeding time, the feeding amount, and the water supply amount. It should be noted that these correction coefficients can also be treated as being included in machine learning.

Further, in order to cope with the case where a large number of breeds exist in the breeding livestock, a standard breeding mathematical model corresponding to the breed may be generated and a plurality of standard breeding mathematical models may be stored. This is because varieties have a great influence on the number of offspring and the breeding cycle. By preparing a standard breeding actuarial model according to the breed, it is possible to select an appropriate model when predicting the number of offspring. As for the variety, not only the mathematical model corresponding to the variety may be constructed, but also it may be used as one parameter for predicting the number of offspring.

The mathematical model construction unit 114 can generate an individual breeding mathematical model individually applied to each breeding livestock by using these standard breeding mathematical models and correction coefficients. Specifically, for example, first, a standard breeding mathematical model according to the breed of the breeding livestock is selected. Next, by selecting and adding a correction factor applicable to the breeding livestock, individual breeding that can more accurately estimate the number of offspring at a predetermined timing of the breeding livestock than the standard breeding mathematical model. You can build a mathematical model. For the individual breeding actuarial model constructed in this way, data such as breeding history information associated with the breeding livestock, for example, past breeding performance, breeding environment information, and the contents of the feed to be given are obtained. By inputting, the breeding livestock can estimate the number of offspring at any time in the future.

Whether it is a standard breeding mathematical model or an individual breeding mathematical model, it is possible to estimate the number of offspring at any timing in the future. Even if it is the first birth, it is possible to estimate the number of offspring with a prediction interval based on the mathematical model already constructed. As a matter of course, the accuracy of estimation becomes higher as the actual data linked to the number of birth history increases. Arbitrary timing in the future is, for example, if there is data on the actual value of the next birth up to 3 births, the number of births by birth history after 4 births, which is the next birth history, and the cumulative number of births. It can be estimated. The upper limit of the number of births is set arbitrarily.

In addition, in order to obtain an individual breeding mathematical model, constructing a standard breeding mathematical model and then using a correction coefficient is an example of a construction method, and a mathematical model corresponding to the individual breeding mathematical model may be constructed from the beginning. ..

The constructed mathematical model re-acquires actual value data as appropriate, periodically reconstructs the model, and is maintained in a state of being able to respond to the latest data state by the mathematical model management unit 115.

FIG. 4 is a schematic diagram for explaining an example of a standard breeding mathematical model. In this figure, the vertical axis represents the number of offspring and the horizontal axis represents the number of offspring for each birth history. D1, D2, D3, D4, D5, D6, and D7 represent the distribution of the number of offspring in the first, second, third, fourth, fifth, sixth, and seventh births, respectively. The vertical solid line represents the range from the minimum number of offspring to the maximum number of offspring of the sow in the birth history. The rectangle represents the distribution. The distribution may be, for example, a plot of points, a probability density distribution, or a quartile to a third quartile as in a box plot. ..

The curve drawn by the dotted line connects the median of the standard total production values that can be represented by these distributions. Originally, since the birth history is a discrete value, it cannot be displayed on a curve, but it is schematically expressed with the intention of assisting understanding.

As described above, the breeding mathematical model obtained by the mathematical model construction unit 114 is a mathematical model for estimating the number of offspring produced by the breeding livestock according to various conditions such as breed and birth history. This includes a model of average number of offspring based on the birth history of sows and a standard total number of offspring based on the number of birth history as a standard actuarial model.

The livestock number estimation unit 111 includes breeding livestock basic information linked to breeding livestock stored in the livestock information storage unit 121, breeding environment information acquired by various measuring devices 310, and livestock information storage unit 121. It has a function of estimating the number of offspring of breeding livestock at a predetermined timing by the breeding history information regarding the breeding history of the breeding livestock stored in and the breeding mathematical model stored in the mathematical model storage unit 122. ..

FIG. 5 is a diagram illustrating a process performed by the child-bearing number estimation unit according to the present embodiment. That is, as described above, the offspring number estimation unit 111 uses the breeding actuarial model to input the breeding livestock basic information, the breeding environment information, and the breeding history information as input information, so that the breeding livestock It outputs the number of offspring as an estimated value.

In addition, the number of livestock estimation unit 111 includes history information including at least history information of the temperature of the breeding environment included in the breeding environment information, and history information showing the transition of weight and / or physical condition information included in the breeding history information. It has a function to estimate the number of offspring of breeding livestock at a predetermined timing by a breeding actuarial model using the history information of the number of offspring.

That is, the birth number estimation unit 111 adopts at least the history information of the temperature of the breeding environment as the breeding environment information necessary for estimating the number of births, and the weight and / or physical condition as the breeding history information. It means to use the history information showing the transition of the information about and the history information of the past number of offspring. For example, heat stress due to an increase in temperature (air temperature) is considered to have an effect on a decrease in conception rate, so this should be taken into consideration. In addition, changes in body weight information and information on physical condition are also considered to affect the number of offspring, so these are taken into consideration. In addition, since it is considered that there is a possibility that the number of babies born after that will have the same tendency in the past birth record, these are taken into consideration.

In addition, the number of offspring estimation unit 111 has a function of estimating the number of future offspring for each number of births, which is the number of births of livestock for breeding.

That is, as shown in the above-mentioned mathematical model of breeding, the number of offspring is influenced by the number of birth history, so this is taken into consideration. The birth calendar indicates the birth calendar including abnormalities such as miscarriage after conception, and is a number counted at the time of past seeding. On the other hand, the birth order is the number counted as the actual result when the baby was delivered normally in the past. Therefore, strictly speaking, the birth history and the birth order are not the same, but the birth order can be used instead of the birth history number.

In addition, the breeding number estimation unit 111 estimates each breeding number of the plurality of breeding livestock for the plurality of breeding livestock managed on the farm by a breeding actuarial model applied to each breeding livestock and corrected. By doing so, it has a function to calculate the total number of offspring of the farm.

That is, the number of offspring of breeding livestock that has been estimated individually is estimated for the breeding livestock of the entire farm, and the total number of offspring of the entire farm is calculated. The total number of offspring of this farm as a whole can be used to evaluate the breeding performance of the whole farm and to help optimize the composition of breeding livestock belonging to the farm.

The breeding number estimation unit 111 selects a breeding mathematical model according to the breed included in the livestock basic information for breeding from the mathematical model storage unit 122, and uses the selected breeding mathematical model to produce the number of livestock for breeding. Has a function to estimate.

That is, since the tendency of the number of offspring differs depending on the variety, an appropriate mathematical model of reproductive performance according to the variety is selected and the number of offspring is estimated accurately.

The livestock state estimation unit 112 has a function of estimating the weight information and / or the information on the physical condition of each breeding livestock by using the image of the breeding livestock acquired by the imaging unit 311.

That is, as described above, information on the weight and physical condition of breeding livestock can be obtained, for example, by measuring the body weight with a weight scale. However, since it is burdensome for both farmers and livestock to measure their weight daily through a physical device such as a pig weighing machine, the estimated weight is calculated from the images of breeding livestock acquired by the imaging unit 311. It is to calculate. Although a more specific explanation is omitted here, the weight is estimated by analyzing the acquired image, calculating the dimensions and area of the breeding livestock, and comparing it with the past measurement results by the scale. Can be done.

The alert generation unit 113 has a function of generating alert information for the user when the breeding intermediate index such as the weight information and / or the information on the physical condition of the breeding livestock and / or the final index such as the number of offspring satisfy a predetermined condition. Have.

That is, as described above, the alert generation unit 113 issues an alert to the user when the condition stored in the alert condition storage unit 124 occurs. For example, as an intermediate index for breeding livestock, weight information and / or information on physical condition is always grasped, and when weight information and / or information on physical condition at a predetermined timing matches the alert condition, an alert is given. Is reported. In addition, the number of offspring is estimated using a breeding mathematical model, and an alert is issued when the estimated number of offspring at a predetermined timing matches the alert condition.

More specifically, for example, when a certain breeding livestock weighs a predetermined amount less than a standard weight at a predetermined age after giving birth, or when the breeding is acquired by various measuring devices 310 at that time. Using the environmental information and the breeding history information regarding the breeding history of the breeding livestock stored in the livestock information storage unit 121, the number of offspring predicted by the breeding number estimation unit 111 using the breeding actuarial model is the average number of offspring. When the number is smaller than the standard number of offspring, it is conceivable to issue an alert including basic information on the breeding livestock that is occurring and detailed information that matches the alert conditions. The notification can be a preset method, e-mail depending on the destination, and not limited to this, text notification, link notification such as URL can be performed using various message applications. The alert generation unit 113 issues an alert using at least one of the alert conditions based on weight information, physical condition information, number of offspring, etc., or combines a plurality of alert conditions to determine a predetermined number of alert conditions. If the above conditions are met, an alert may be issued.

Further, the alert generation unit 113 can also include the alert management information regarding the alert cancellation setting and the re-alert setting after the alert cancellation in the information regarding the alert issuance. In the re-alert setting, a plurality of different re-alert conditions set in advance may be presented so that the user can select them.

The mathematical model management unit 115 refers to the breeding actuarial model generated by using the historical information of the number of offspring of a plurality of breeding livestock and the historical information of the breeding environment information of a plurality of breeding livestock, with respect to the breeding livestock. It has a function to update the breeding actuarial model by using the breeding environment information, the breeding history information of the breeding livestock, and the actual value of the number of offspring of the breeding livestock.

That is, since the breeding record, breeding history information, and breeding environment information change day by day, the breeding mathematical model is maintained in the latest state corresponding to these data.

The management accounting department 116 arises from the breeding of the breeding livestock by using the financial basic information about the farm management stored in the financial basic information storage unit 123 and the information of the breeding livestock included in the breeding basic information. It has a function to calculate management accounting information including information on profits.

Information on the benefits that result from breeding livestock is, for example, the costs of breeding livestock tied to breeding livestock, depreciation tied to breeding livestock, and the sale of meat livestock. Information on profits obtained, information on profits obtained from the sale of meat and livestock, etc.

As mentioned above, in order to improve the efficiency of livestock management, it is important to grasp the cost information regarding the breeding livestock and the depreciation information of the breeding livestock held as assets. As a livestock industry, the main profit generated from breeding livestock is the profit obtained from the sale of meat livestock that are offspring. Therefore, it is important not only to raise costs, but also to grasp the profits and profits obtained by raising livestock and selling them, and the Management Accounting Department 116 calculates the profits and profits obtained by selling meat livestock. As a result, it has a function to calculate information on the profits generated from the breeding of livestock in total.

Revenues obtained from the sale of meat livestock can be calculated, for example, by multiplying the unit of past average selling price and average profit amount stored as basic financial information by the number and weight of meat livestock. good. Further, it may be calculated using the actual value of the selling price in the market, that is, the so-called market price. That is, the number of offspring estimated by the number of offspring estimation unit 111 is useful for calculating the expected profit obtained from the sale of meat and livestock. For example, it is possible to estimate the estimated amount of profit obtained from the sale of meat and livestock by multiplying the number of offspring and the average selling price at any time in the future.

The profits arising from the breeding of breeding livestock may be calculated as the profits earned by the livestock industry, such as the sale of these meat livestock, minus costs and are expressed as a general accounting term. It may be gross profit, operating profit, ordinary profit, pre-tax net profit, or net profit. That is, roughly speaking, the management accounting department 116 generates management accounting information from the breeding livestock by subtracting the expenses required for breeding livestock from the profit obtained from the sale of meat livestock. You can calculate the profit.

The configuration optimization unit 117 includes basic information on breeding livestock, management accounting information including information on benefits generated from breeding breeding livestock calculated by the management accounting department 116, and breeding livestock estimated by a breeding actuarial model. Multiple breeding to optimize any one of the total number of offspring on the farm at a given timing and the benefits of breeding livestock using the number of offspring in a given number of births. It has a function to generate the optimum breeding history composition information indicating the combination of breeding history composed of livestock.

FIG. 6 is a diagram illustrating a process performed by the configuration optimization unit 117 according to the present embodiment. That is, the composition optimization unit 117 determines the number of offspring in the future birth history estimated according to the information such as the breed and birth history of each breeding livestock in captivity recorded in the basic information on breeding livestock. Direct and indirect costs and future production costs, information on depreciation in current and future production history, information on profits from selling meat livestock, breeding livestock as a livestock industry Management accounting information including information on profits generated from livestock is used as input information, and if necessary, as a constraint condition, for example, when information on the upper limit of the number of equipment (stalls) that a farmer can breed is input. The optimum birth history composition information of breeding livestock is generated as output information.

In addition, the composition optimization unit 117 is either the basic information on livestock for breeding or the benefit generated from the total number of offspring of the farm or the breeding of livestock for breeding at a predetermined timing according to the optimum birth history composition information. It has a function to generate information on candidate livestock for disuse that proposes candidates for livestock that should be disused in order to optimize.

That is, using basic information on breeding livestock and optimal birth history composition information, at least one of the total number of offspring on the farm or the benefit resulting from breeding livestock is optimized at a predetermined timing in the future. Therefore, it is possible to calculate which breeding livestock should be abolished and individually propose the identification ID of the abandoned candidate livestock. Since the farmer decides whether or not to actually disuse the breeding livestock, the composition optimization unit 117 proposes a candidate.

At the same time, it calculates which breeding livestock should be maintained in order to optimize either the total number of offspring of the farm at a given timing in the future or the benefit generated from breeding livestock, and individual maintenance. It can be extracted as an identification ID of a candidate livestock. Since the farmer decides whether or not to actually maintain the breeding livestock, the composition optimization unit 117 proposes a candidate.

In addition, the candidate livestock number calculation unit 118 newly uses the abandoned livestock selection information selected based on the abolition candidate livestock information and the candidate livestock number information stored in the livestock information storage unit to newly breed livestock. It has a function to calculate the number of preparation candidate livestock that should be prepared as candidates for.

FIG. 7 is a diagram illustrating a process performed by the candidate livestock number calculation unit according to the present embodiment. That is, the candidate livestock number calculation unit 118 has at least abandoned livestock selection information, maintenance livestock selection information, candidate pig selection information (selection rate), and, if necessary, as constraint conditions, for example, breeding livestock being raised. By inputting information such as the number of candidate livestock and facility equipment, the number of livestock to be newly prepared as a candidate for breeding livestock (number of candidate livestock to be prepared) is calculated. The disuse livestock selection information is a display of disuse livestock candidates presented by the configuration optimization unit 117 using the display unit 212 of the terminal device 201, and the farmer himself who is the user inputs the terminal device 201 from the display unit. This is the information selected and input using the part 211 to select which breeding livestock should be disused. Not all candidate livestock for breeding livestock enter the breeding process, and are usually selected to become breeding livestock. The candidate livestock selection information, which is the past selection rate, is included in the livestock information storage unit 121 or the basic financial information and is stored. In order to fill the shortage due to the disuse of breeding livestock, the number of breeding livestock to be newly supplemented is calculated using the disuse livestock candidate information. In order to satisfy the number of breeding livestock to be supplemented, the number of candidate livestock already bred or the number of prepared candidate livestock required for newly introduced breeding livestock is calculated by back calculation from the selection rate.

FIG. 8 is a conceptual diagram showing the cooperation of each configuration described above. To reiterate, to confirm the outline of the role of these configurations, the offspring number estimation unit 111 predicts the number of offspring using a breeding mathematical model. If the number of births can be predicted, the management accounting department 116 can estimate the profit based on the number of births, and can calculate the management accounting information such as the profit of the business together with the known costs. If management accounting information can be obtained, the composition optimization department 117 will optimize the composition of sows raised on the farm in order to maximize profits and total number of offspring. And, it is possible to generate information on candidate livestock for disuse. This makes it possible to weed out sows with reduced productivity and optimize farm operations.

FIG. 9 shows the optimum livestock composition information output by the configuration optimization unit 117, the disuse livestock candidate, the maintenance livestock candidate, and the number of candidate livestock required as the newly introduced breeding livestock output by the candidate livestock number calculation unit 118. Is an example of a screen displayed on the display unit 212 of the terminal device 201.

In this figure, the optimum production history composition information calculated by the composition optimization unit 117 is shown in the upper and middle rows. The graph shown in the upper part is a vertical bar graph showing the ratio of the number of candidate livestock for disuse proposed by the Composition Optimization Department 117 to the number of livestock for breeding in the entire farm by birth history. From this graph, the farmer who is a user can recognize how much breeding livestock is proposed to be abolished in which birth history. Here, in the breeding livestock for which disuse is proposed, for example, the number of offspring at a predetermined timing estimated by the individual breeding mathematical model is lower than the number of offspring estimated by the standard breeding mathematical model, and the cost and depreciation are reduced. It is a breeding livestock that is judged to be underperform overall in consideration of amortization and profit.

The list format shown in the middle of this figure is information about breeding livestock belonging to the farm, and you can enter selection information on whether to use maintenance livestock or disused livestock from the left column. Check box, button to confirm the selection, identification ID to identify individual breeding livestock, information indicating birth history, information indicating delivery rate, information indicating cumulative number of offspring, cumulative number of livestock surviving Information to show, information to show the average weight of the offspring in the most recent birth history and its details, information to show the cumulative number of weaning heads of the offspring and its details, information to show the cumulative cost and its details, depreciation and information to show its details. Is displayed. The farmer who is the user can select which breeding livestock to maintain or disuse while browsing and examining various information shown in this list.

Displayed on the lower left side of this figure is a bar graph showing the total number of offspring, breeding costs, depreciation, and profits of the entire farm, which varies depending on the selection information of disused livestock and maintained livestock, and is shown in the middle. At the same time, how these figures are estimated to fluctuate (updated (predicted)) compared to when no breeding livestock are disposed of (existing) according to the selection information provided by the check boxes. Shows. The farmer, who is the user, can make decisions while seeing how the total number of offspring, feeding costs, depreciation, and profits of the farm are estimated to change depending on the choice of disuse or maintenance.

Displayed on the lower right side of this figure is information on the number of candidate livestock required for newly introduced breeding livestock. The newly introduced breeding livestock are selected from the candidate livestock raised as breeding candidate livestock. Therefore, the number of candidate livestock currently bred and their details, and the number of candidate livestock after 150 days, which is the number of candidate livestock that have reached the age suitable for breeding, and their details, and the candidate livestock are breeding livestock. It shows the selection rate when it is introduced as. The age is not limited to 150 days, and the age can be set as appropriate. Further, based on the disused livestock selection information selected in the list in the middle of the above screen, the required number of candidate livestock can be displayed by back calculation from the selected number of disused livestock and the selection rate. The farmer who is the user can make a decision while referring to this necessary candidate livestock information.

<Processing flow>
FIG. 10 shows a flowchart illustrating a flow chart for constructing a mathematical model executed on the livestock information management server 101 of the livestock information management system 1 according to the embodiment of the present disclosure. The information management method will be explained. The flowchart is an example, and the mathematical model construction is not limited to the processing of the flowchart.

In step S101, the mathematical model construction unit 114 acquires information necessary for the following processing by the information acquisition unit 119. The information is mainly stored in the livestock information storage unit 121, and is breeding livestock basic information, breeding environment information acquired and stored from a measuring device 310, and the like, and breeding history information.

In step S102, the mathematical model building unit 114 generates a standard breeding mathematical model. Specifically, as input information, breeding livestock basic information, breeding environment information, etc. are used, and the actual value of the offspring by birth history of the breeding livestock basic information and its distribution are used as objective variables, and the breeding environment information, Estimate the standard number of offspring of breeding livestock by selecting and combining the best explainable explanatory variables from indicators other than the number of offspring by birth history in feeding, water supply information, and basic information on livestock for breeding. Generate a standard breeding actuarial model that can be. This standard breeding actuarial model may be generated for each breed.

In step S103, the mathematical model construction unit 114 performs a process of extracting a factor that affects the number of offspring as a correction coefficient. The correction factor is based on a standard mathematical model of reproduction, for example, due to seasonal factors (summer, winter, temperature, humidity, etc.), feeding, factors due to water supply, and factors due to normal / abnormal past deliveries. It adds an increase or decrease correction to the number of children.

In step S104, the mathematical model building unit 114 generates an individual breeding mathematical model. The individual breeding actuarial model is based on the standard breeding actuarial model, using a correction coefficient and using the number of offspring of the breeding livestock basic information as the objective variable, and the breeding environment information, feeding, water supply information, and breeding livestock basic information. Using explanatory variables other than the number of offspring by birth history, we generate an individual breeding actuarial model that can more accurately estimate the number of offspring at a given timing of the breeding livestock.

FIG. 11 shows an example of predicting the number of offspring. In this figure, the predicted number of offspring estimated by the number of offspring estimation unit 111 is shown. As the training data, 70% of the 34506 mother pigs in the sample are used, and the attribute data related to the remaining 30% of the mother pigs are input to the model to predict the predicted number of offspring in the next calving, and the number of offspring is calculated. It is expressed as a frequency histogram of the mother pig as a class. Random forest is used as a machine learning model. As input information (explanatory variable), breeding environment information, breed and number of birth history were used. More specifically, the farm and temperature are used as the breeding environment information, but the year and month of the seeding date are used as alternative data for the temperature. Actually, the accuracy is improved by using the temperature obtained from the measuring device, and as a stress load, the cumulative temperature from the seeding date to the birth date, the daily range, the maximum temperature difference, etc. By using synthetic variables, environmental stress factors can be reflected in the model, so accuracy improvement can be expected.

In this figure, the horizontal axis is the predicted number of offspring for the next time, and the vertical axis is the number of sows. All predicted numbers of offspring are calculated by natural numbers. The expected number of offspring can be calculated in this way. This is the total predicted number, and the number of offspring under predetermined conditions, such as for each farm, can also be predicted. This makes it possible to roughly grasp the number of sows that are expected to produce the number of pigs shown on the horizontal axis in the next calving.

FIG. 12 shows a flowchart illustrating the control of the breeding intermediate index executed in the livestock information management server 101 of the livestock information management system 1 according to the embodiment of the present disclosure, specifically, the flow of alert issuance processing. Hereinafter, the livestock information management method will be described according to the same flowchart. The flowchart is an example, and the alert issuance is not limited to the processing of the flowchart.

In step S201, the livestock state estimation unit 112 acquires the information necessary for estimating the following states by the information acquisition unit 119. For example, the image captured by the image capturing unit 311 is acquired.

In step S202, the livestock state estimation unit 112 calculates the breeding intermediate index. Specifically, for example, the estimated body weight of the breeding livestock is calculated based on the acquired image.

In step S203, the calculated breeding intermediate index and the alert condition are compared. For example, the alert generation unit 113 compares the body weight estimated by the livestock state estimation unit 112 with the body weight which is the alert condition stored in the alert condition storage unit 124. The body weight, which is the alert condition here, is a condition of the body weight that the breeding livestock should satisfy at a predetermined timing such as a predetermined postpartum and a predetermined age. The alert condition may be met if the weight is less than or equal to the predetermined weight and is greater than or equal to the predetermined weight, or may be met if the weight deviates from the predetermined range, and is appropriately set. It is remembered. Further, the alert generation unit 113 compares the number of births estimated by the birth number estimation unit 111 with the standard number of births such as the average number of births which is the alert condition stored in the alert condition storage unit 124. conduct. The standard number of offspring, which is the alert condition here, is, for example, the average number of offspring of breeding livestock in a predetermined birth history. The alert condition may be satisfied when the number of offspring is less than or equal to the predetermined number, or the condition is met when the number of offspring deviates from the predetermined range. It may be present, and it is set and stored as appropriate. In addition, based on the distribution of the number of offspring in the birth history, a range of 1σ or 1.5σ was set as a threshold value for the standard number of offspring, and the range was deviated from that range (assuming that the number of offspring was placed in a natural number). In some cases, the conditions may be met. The average number of offspring is an example of the standard number of offspring, and another reference number of offspring may be used as the standard number of offspring.

In this way, the alert condition realizes an appropriate threshold setting by statistically setting whether or not the situation is such that the number of offspring finally obtained is pushed down in the negative direction. Due to continuous breeding of breeding sows, the median number of offspring tends to increase. Therefore, in the case of farmers who are changing varieties, the threshold once decided is not always correct, and even though they have no experience, there is no solution as to what they should have to decide the threshold. become. By processing this statistically, it is possible to set a scientific threshold value according to the situation. For example, if the number of offspring (normal offspring) is significantly different in adjacent pig bunches (stalls) that have the same birth history and breed and are considered to have the same temperature and humidity conditions, the factors that cause differences are limited. It will be. Therefore, we will manage variables to improve the number of offspring by sequentially inputting the limited reoccurrence factors into the prediction model and selecting the one with the smallest number of offspring. Is possible.

In step S204, the alert generation unit 113 returns to the calculation of the breeding intermediate index in step S202 when the condition is not satisfied (No) by comparison with the above alert condition. If the condition is satisfied (Yes), the process proceeds to the next step.

In step S205, an alert is issued for the target livestock that matches the alert condition. As a specific notification means, the alert may be displayed on the screen by using the display unit 212 of the terminal device 201, or may be limited to a preset method, e-mail depending on the destination, and the like. Instead, various message applications can be used to notify texts and links such as URLs.

Here, referring to FIG. 13, a diagram showing an example of the alert issuance management screen is shown. As shown in the figure, when the alert generation unit 113 issues an alert, the display unit 212 of the terminal device 201 displays a list of breeding livestock for which the alert is issued. A second area 402 for selecting the necessity of an alert and a third area 403 for selecting a re-alert condition are displayed.

In the first area 401, an identification ID (mother pig ID) for identifying the breeding livestock to be alerted, a birth order, an average calving rate, a detail button, and an alert cancellation check box are displayed for each identification ID. Will be done. When the user selects and operates the detail button, detailed information on the selected breeding livestock is displayed. If the user checks the clear alert checkbox, the alert to the corresponding breeding livestock will be cleared, otherwise the corresponding breeding livestock will be kept as an alert target and will be listed when the next alert is issued. Will be done.

In the second area 402, a button (so-called radio button) that can select either to set or not to set the re-alert for each breeding livestock to be alerted is displayed. The user selects whether or not to set a re-alert for each breeding livestock in the second area 402.

In the third area 403, a check box and a condition detail button on which at least one of the three condition sets of the first set to the third set can be selected as the re-alert condition after the alert is canceled are displayed. The three condition sets are set to different re-alert conditions, and the user can display the condition contents of each condition set by selecting the condition detail button. For example, the first set is a condition related to body weight, the second set is a condition related to physical condition, the third set is a condition related to the number of offspring, and so on. Alternatively, the threshold values may be set differently under the same type of conditions. Then, when the re-alert condition selected by the user is satisfied, the corresponding breeding livestock is listed again on the alert issuance management screen. Alternatively, although not shown, a management screen for re-alert may be displayed.

FIG. 14 shows a flowchart illustrating a flow of configuration optimization of breeding livestock bred on a farm, which is executed in the livestock information management server 101 of the livestock information management system 1 according to the embodiment of the present disclosure. , Hereinafter, the livestock information management method will be described according to the same flowchart. The flowchart is an example, and the optimization of the composition of livestock for breeding is not limited to the processing of the flowchart.

In step S301, the offspring number estimation unit 111 estimates the number of offspring of individual breeding livestock using the constructed individual breeding mathematical model.

In step S302, the number of offspring estimation unit 111 totals the number of offspring of individual breeding livestock and estimates the number of offspring on the farm, which is the number of offspring of the entire farm.

In step S303, the configuration optimization unit 117 relates to basic information on livestock for breeding, the number of offspring produced on the farm, and the benefits generated from the breeding of livestock associated with each livestock calculated by the Management Accounting Department 116. Consists of multiple breeding livestock to optimize at least the number of offspring produced on the farm or the benefits resulting from the breeding of the breeding livestock at a given time using the information. Generates optimal birth history composition information showing combinations of birth history.

In step S304, the configuration optimization unit 117 determines at least the total number of offspring of the farm or the profit generated from the breeding of the breeding livestock at a predetermined timing according to the breeding livestock basic information and the optimum birth history composition information. We will extract breeding livestock for disuse candidates and maintenance candidates that propose livestock candidates that should be discontinued in order to optimize either of them.

In step S305, the farmer who is the user inputs the disuse livestock selection information for selecting which breeding livestock is to be disused from the extracted disuse candidates. The number of breeding livestock to be supplemented is calculated, and the number of breeding candidate livestock required for the introduction of the breeding livestock is calculated.

In step S306, the candidate livestock number calculation unit 118 calculates the number of candidate livestock to be newly prepared as a candidate for breeding livestock by using the waste livestock selection information and the candidate livestock selection information included in the basic financial information. do.

In step S307, the configuration optimization unit 117 considers constraint conditions (for example, whether or not the required number of candidate livestock can be secured as a result of referring to the number of candidate livestock in use and the selection rate), and as a result, the birth history is determined by the constraint conditions. It is determined whether or not the re-execution of the configuration optimization is necessary, and if the re-execution of the optimization is unnecessary (No), the process is terminated. If it is necessary to re-execute the optimization, the process returns to step S303, and the optimum birth history composition information is generated again in consideration of the constraint conditions.

As described above, in the livestock information management system 1 according to the embodiment of the present disclosure, the measuring device 310 for acquiring the breeding environment information regarding the breeding environment of the breeding livestock installed on the farm where the breeding livestock is reared. And the basic information on breeding livestock including at least breeding livestock information associated with the breeding livestock stored in the livestock information storage unit 121, and at least information on the temperature of the farm acquired by the measuring device 310. Regarding the breeding environment information, the breeding history information regarding the breeding history including at least the history information of the number of offspring of the breeding livestock stored in the livestock information storage unit 121, and at least the temperature of the farm included in the breeding environment information. By learning the information, the breed information included in the basic information on breeding livestock, and the historical information on the number of offspring of breeding livestock included in the breeding history information as learning data in a mathematical model, the production of breeding livestock In the mathematical model construction unit 114 that generates a breeding mathematical model for estimating the number of offspring by history, the mathematical model storage unit 122 that stores the breeding mathematical model generated by the mathematical model construction unit 114, and the mathematical model storage unit 122. The breeding actuarial model is used to estimate the number of offspring of breeding livestock at any time in the future. By estimating the number of offspring of breeding livestock, it is possible to grasp the current productivity of breeding livestock one by one and manage the breeding toward the target productivity.

In addition, the basic financial information storage unit 123 that stores basic financial information related to farm management, the basic financial information, the information on livestock for breeding included in the basic information on livestock for breeding, and the production estimated by the number of offspring estimation unit 111. Estimated number of offspring by further providing a management accounting department 116, which uses the number of offspring to calculate management accounting information, including information on the benefits arising from the breeding of the breeding livestock associated with the breeding livestock. Since it is possible to recognize information on the benefits generated from breeding livestock based on the above and grasp the livestock management status, it is possible for the farmer who is the user to take measures such as taking breeding measures to improve those conditions. become able to.

In addition, using basic livestock information for breeding, management accounting information including information on benefits generated from breeding livestock calculated by the management accounting department 116, and the number of offspring estimated by the number of offspring estimation unit 111. , Optimal birth history showing a combination of birth history consisting of multiple breeding livestock to optimize at least either the total number of offspring on the farm or the benefits resulting from breeding livestock at a given time. In order to improve the total number of offspring on the farm or the benefits generated from breeding livestock by further providing the composition optimization unit 117, which generates composition information, the birth history composition of the mother pigs held on the farm. Can be improved.

In addition, the composition optimization unit 117 obtains at least one of the total number of offspring of the farm or the benefit generated from the breeding of the breeding livestock at a predetermined timing according to the basic information on the breeding livestock and the optimum birth history composition information. Proposing abandoned candidate livestock candidates for optimization By generating abandoned candidate livestock information, it becomes easier for the farmer who is the user to understand what kind of sow should be abolished, and it is owned by the farm. It will be easier to improve the composition of the sow.

In addition, the waste livestock selection information selected based on the waste candidate livestock information and the candidate livestock number information stored in the livestock information storage unit 121 should be newly prepared as candidate livestock for breeding livestock. By further providing the candidate livestock number calculation unit 118 for calculating the number of candidate livestock for preparation, the farmer who is the user can prepare the candidate pigs that will be needed in the future in advance.

Further, the mathematical model storage unit 122 stores at least a breeding mathematical model according to the breed information of the livestock, and the offspring number estimation unit 111 stores a breeding mathematical model according to the breed included in the breeding basic information of the livestock. By selecting from the mathematical model storage unit 122 and estimating the number of offspring of breeding livestock using the selected breeding mathematical model, the accuracy of estimating the number of offspring is improved.

In addition, the breeding mathematical model is generated by adding at least the weight information included in the breeding environment information and / or the information on the physical condition to the breeding history information as learning data, so that the state of the mother pig can be obtained. Information such as the weight and physical condition to be expressed can be incorporated into the mathematical model as the internal reproductive variables of the mother pig derived from the individual mother pig, and the accuracy of estimating the number of offspring is improved.

In addition to the information on the temperature, the breeding mathematical model is generated by adding information including at least one of the humidity, air volume, light volume, sound, and odor of the farm acquired from the measuring device as training data. By doing so, it becomes possible to incorporate the factors of the environment in which the mother pig is raised into the mathematical model as environmental variables, and the accuracy of estimating the number of offspring is improved.

In addition, the breeding mathematical model is generated by adding at least one of information on the source of the breeding livestock and information on the age of the breeding livestock as learning data as basic information on the breeding livestock. As a result, it becomes possible to incorporate it into a mathematical model as an internal reproductive variable of the mother pig derived from the individual mother pig such as the origin and aging status of the mother pig, and the accuracy of estimation of the number of offspring is improved.

In addition, the breeding actuarial model is used as breeding history information, further information on the feeding type and / or feeding amount for the breeding livestock, information on the water supply amount for the breeding livestock, and medication history for the breeding livestock (used for promoting reproduction). ), At least one of the information is added as training data and generated, so that it is incorporated into the mathematical model as a breeding variable that indicates whether the energy and nutrients necessary for giving birth to a piglet are sufficiently ingested. Is possible, and the accuracy of estimating the number of offspring is improved.

In addition, the breeding actuarial model can be used as breeding history information, such as disease morbidity history of breeding livestock, vaccine administration history of breeding livestock, medication history of breeding livestock (used for treatment or prevention of diseases), and breeding livestock. At least one of the normal or abnormal history of childbirth is added and generated as learning data, so that measures are taken for the treatment or prevention of the disease, or the childbirth has been performed normally so far. Since it is possible to incorporate it into a mathematical model as a soma pig health condition variable, such as whether or not it is present, the accuracy of estimating the number of offspring is improved.

In addition, the breeding mathematical model is generated by adding at least one of the breeding history information, the seeding date for the breeding livestock, the semen that has been seeded, and the semen property information as learning data. As a result, it becomes possible to incorporate the reproductive performance including the fertility of the sow into the mathematical model as an external reproductive variable of the mother pig, and thus the accuracy of estimating the number of offspring is improved.

In addition, the breeding number estimation unit 111 estimates each breeding number of the plurality of breeding livestock for the plurality of breeding livestock managed on the farm by a breeding actuarial model applied to each breeding livestock and corrected. By doing so, by calculating the total number of offspring of the farm, it is possible to use the breeding actuarial model applied by each breeding mother pig, and the accuracy of estimating the number of offspring is improved.

In addition, for the breeding actuarial model generated by using the historical information of the number of offspring of a plurality of breeding livestock and the historical information of the breeding environment information of a plurality of breeding livestock as learning data, the breeding livestock Mathematical model management unit 115 for updating the breeding actuarial model using the breeding environment information, the breeding history information of the breeding livestock, and the actual value of the number of offspring of the breeding livestock. The model can remain responsive to the latest data state.

Further, an image capturing unit 311 for acquiring an image of the breeding livestock is further provided, and using the image of the breeding livestock acquired by the imaging unit 311, the weight information of the breeding livestock and / or the information on the physical condition, By further providing the livestock state estimation unit 112 for estimating the above, it is possible to reduce the burden on the farmer who is the user to actually measure the weight.

Further, an alert generation unit 113 for generating alert information for the user is further provided, and the alert generation unit 113 generates alert information when the weight information of the breeding livestock and / or the information on the physical condition satisfy a predetermined condition. By doing so, the farmer who is a user who recognizes that the condition of the sow is unfavorable can take measures such as taking breeding measures to solve these problems.

Further, an alert generation unit 113 for generating alert information for the user is further provided, and the alert generation unit 113 generates alert information when the number of offspring estimated by the number of offspring estimation unit 111 satisfies a predetermined condition. Therefore, if the estimated number of offspring to be produced in the future is predicted to be less than the standard number of offspring such as the average number of offspring, the farmer who is the user will take measures to solve these problems. You will be able to take measures such as hitting.

(program)
FIG. 15 is a schematic block diagram showing the configuration of the computer 801. The computer 801 includes a CPU 802, a main storage device 803, an auxiliary storage device 804, and an interface 805.

Here, the details of the program for realizing each function constituting the livestock information management server 101 according to each embodiment will be described.

The livestock information management server 101 is mounted on the computer 801. Then, the operation of each component of the livestock information management server 101 is stored in the auxiliary storage device 804 in the form of a program. The CPU 802 reads the program from the auxiliary storage device 804, expands it to the main storage device 803, and executes the above processing according to the program. Further, the CPU 802 secures a storage area corresponding to the above-mentioned storage unit in the main storage device 803 according to the program.

Specifically, the program is a livestock information management program that manages livestock information on the computer 801. The breeding livestock basic information linked to the stored breeding livestock and the acquired breeding livestock. Estimating the number of offspring of breeding livestock at a predetermined timing by the breeding performance actuarial model stored using the breeding environment information regarding the breeding environment of the breeding environment and the breeding history information regarding the breeding history of the stored breeding livestock. It is a program to be realized.

The auxiliary storage device 804 is an example of a tangible medium that is not temporary. Other examples of non-temporary tangible media include magnetic disks, magneto-optical disks, CD-ROMs, DVD-ROMs, semiconductor memories and the like connected via the interface 805. When this program is distributed to the computer 801 via the network NW, the distributed computer 801 may expand the program to the main storage device 803 and execute the above processing.

Further, the program may be for realizing a part of the above-mentioned functions. Further, the program realizes the above-mentioned function in combination with another program already stored in the auxiliary storage device 804, that is, a so-called difference file (difference program).
May be.

Although the embodiments of the present invention have been described above, these embodiments can be implemented in various other embodiments, and various omissions, replacements, and changes may be made without departing from the gist of the invention. Can be done. These embodiments and variations thereof shall be included in the scope of the invention described in the claims and the equivalent scope thereof, as well as in the scope and gist of the invention. Further, in the description of the embodiment, pigs are taken as an example as livestock, but the livestock information management system may be applied to other livestock such as cows and birds and poultry.

1 Livestock information management system 101 Livestock information management server 111 Livestock number estimation department 112 Livestock condition estimation department 113 Alert generation department 114 Mathematical model construction department 115 Mathematical model management department 116 Management accounting department 117 Configuration optimization department 118 Candidate livestock number calculation department 119 Information acquisition unit 120 Information storage unit 121 Livestock information storage unit 122 Mathematical model storage unit 123 Financial basic information storage unit 124 Alert condition storage unit 201 Terminal device 211 Input unit 212 Display unit 301 Livestock barn 310 Measuring device 311 Imaging unit 320 Feeding device 801 computer 802 CPU
803 Main storage 804 Auxiliary storage 805 Interface

Claims (19)

It is a livestock information management system that manages breeding livestock, which is a mother pig.
A measuring device for acquiring breeding environment information regarding the breeding environment of the breeding livestock installed on the farm where the breeding livestock is raised, and
Basic information on breeding livestock, including at least breed information of the breeding livestock, associated with the breeding livestock stored in the livestock information storage unit.
The breeding environment information including at least information on the temperature of the farm acquired by the measuring device, and
The breeding history information regarding the breeding history of the breeding livestock including at least the history information of the number of offspring of the breeding livestock stored in the livestock information storage unit, and
At least, the information on the temperature of the farm included in the breeding environment information, the breed information included in the breeding livestock basic information, and the history information of the number of offspring of the breeding livestock included in the breeding history information are learned. A mathematical model construction unit that generates a breeding mathematical model for estimating the number of offspring by birth history of the breeding livestock by learning from a mathematical model as data.
A mathematical model storage unit that stores the breeding mathematical model generated by the mathematical model construction unit, and a mathematical model storage unit.
Using the breeding mathematical model stored in the mathematical model storage unit, a breeding number estimation unit for estimating the number of offspring of the breeding livestock at an arbitrary timing in the future is provided.
The livestock number estimation unit is a livestock information management system that estimates the number of offspring of the breeding livestock using the breeding mathematical model. The Treasury Basic Information Storage Department, which stores basic financial information related to farm management,
The above-mentioned basic financial information, the information on the breeding livestock included in the basic information on breeding livestock, and the number of offspring estimated by the livestock number estimation unit are used and linked to the breeding livestock. The livestock information management system according to claim 1, further comprising a management accounting department that calculates management accounting information including information on benefits arising from breeding livestock. Using the management accounting information including the basic information on breeding livestock, the information on the profit generated from the breeding of the breeding livestock calculated by the management accounting department, and the number of offspring estimated by the birth number estimation department. hand,
Optimal production shown for a combination of birth history consisting of a plurality of the breeding livestock to optimize at least the total number of offspring on the farm or the benefits resulting from the breeding of the breeding livestock at a given timing. The livestock information management system according to claim 2, further comprising a configuration optimization unit for generating history configuration information. The composition optimization unit is either at least the total number of offspring of the farm or the benefit generated from the breeding of the breeding livestock at a predetermined timing according to the breeding livestock basic information and the optimum birth history composition information. The livestock information management system according to claim 3, which generates information on livestock candidates for disuse, which proposes livestock candidates to be discontinued in order to optimize the above. Preparations to be newly prepared as candidate livestock for breeding using the abandoned livestock selection information selected based on the abandoned candidate livestock information and the candidate livestock number information stored in the livestock information storage unit. The livestock information management system according to claim 4, further comprising a candidate livestock number calculation unit for calculating the number of candidate livestock. The mathematical model storage unit stores at least a breeding mathematical model according to the breed information of livestock, and stores the breeding mathematical model.
The breeding number estimation unit selects a breeding mathematical model according to the breed included in the breeding livestock basic information of the livestock from the mathematical model storage unit, and uses the selected breeding mathematical model to produce the breeding livestock. The livestock information management system according to any one of claims 1 to 5, which estimates the number of offspring. The actuarial model according to claim 1 to 5, wherein the breeding actuarial model is generated by adding at least the weight information included in the breeding environment information and / or the information on the physical condition included in the breeding history information as learning data. The livestock information management system according to any one of the items. In the breeding mathematical model, in addition to the information on the air temperature, information including at least one of the humidity, air volume, light amount, sound, and odor of the farm acquired from the measuring device is added as training data. The livestock information management system according to any one of claims 1 to 5, which is generated in the above-mentioned. The breeding actuarial model is generated by adding at least one of information on the source of the breeding livestock and information on the age of the breeding livestock as learning data as the breeding livestock basic information. The livestock information management system according to any one of claims 1 to 5. In the breeding actuarial model, as the breeding history information, information on the feeding type and / or feeding amount for the breeding livestock, information on the water supply amount for the breeding livestock, and medication history for the breeding livestock (used for promoting reproduction). ), The livestock information management system according to any one of claims 1 to 5, which is generated by adding at least one piece of information as learning data. In the breeding actuarial model, as the breeding history information, the disease morbidity history of the breeding livestock, the vaccine administration history to the breeding livestock, the medication history to the breeding livestock (used for the treatment or prevention of the disease), and the breeding livestock The livestock information management system according to any one of claims 1 to 5, wherein at least one of the normal or abnormal history of childbirth is added as learning data and generated. The breeding actuarial model is generated by adding at least one of the breeding history information, the seeding date for the breeding livestock, the semen that has been seeded, and the semen property information as learning data. The livestock information management system according to any one of items 1 to 5. The breeding number estimation unit estimates the number of offspring of each of the plurality of breeding livestock for a plurality of breeding livestock managed on the farm by a breeding actuarial model applied to each breeding livestock and corrected. The livestock information management system according to any one of claims 1 to 5, wherein the total number of offspring of the farm is calculated. The breeding livestock is used with respect to the breeding actuarial model generated by using the historical information of the number of offspring of the plurality of breeding livestock and the historical information of the breeding environment information of the plurality of breeding livestock as learning data. A request for further providing a mathematical model management unit for updating the breeding actuarial model using the breeding environment information, the breeding history information of the breeding livestock, and the actual value of the number of offspring of the breeding livestock. The livestock information management system according to any one of Items 1 to 5. Further equipped with an imaging unit for acquiring an image of the breeding livestock,
Claim 7 further includes a livestock state estimation unit for estimating weight information and / or information on physical condition of the breeding livestock using the image of the breeding livestock acquired by the imaging unit. The described livestock information management system. It also has an alert generator for generating alert information for the user.
The livestock information management system according to claim 15, wherein the alert generation unit generates alert information when the weight information of the breeding livestock and / or the information regarding the physical condition satisfy a predetermined condition. It also has an alert generator for generating alert information for the user.
The livestock information management system according to claim 1, wherein the alert generation unit generates alert information when the number of offspring estimated by the number of offspring estimation unit satisfies a predetermined condition. It is a livestock information management method that manages breeding livestock that are mother pigs.
A measuring device for acquiring breeding environment information regarding the breeding environment of the breeding livestock installed on the farm where the breeding livestock is raised, and
Basic information on breeding livestock, including at least breed information of the breeding livestock, associated with the breeding livestock stored in the livestock information storage unit.
Breeding environment information including at least the temperature information of the farm acquired by the measuring device, and
Using at least the breeding history information regarding the breeding history of the breeding livestock including the history information of the number of offspring of the breeding livestock stored in the livestock information storage unit, the mathematical model construction unit uses at least the breeding environment information. The temperature information of the farm included in the above, the breed information included in the basic information on breeding livestock, and the history information on the number of offspring of the breeding livestock included in the breeding history information are used as training data for learning in a mathematical model. By doing so, a step of constructing a mathematical model to generate a breeding actuarial model for estimating the number of offspring by birth history of the breeding livestock, and
The livestock production unit comprises a breeding actuarial model generated by the mathematical model construction unit to estimate the number of offspring of the breeding livestock at an arbitrary timing in the future. Information management method. It is a livestock information management program that manages breeding livestock that are mother pigs.
A measuring device for acquiring breeding environment information regarding the breeding environment of the breeding livestock installed on the farm where the breeding livestock is raised, and
Basic information on breeding livestock, including at least breed information of the breeding livestock, associated with the breeding livestock stored in the livestock information storage unit.
Breeding environment information including at least the temperature information of the farm acquired by the measuring device, and
Using at least the breeding history information regarding the breeding history of the breeding livestock including the history information of the number of offspring of the breeding livestock stored in the livestock information storage unit,
The mathematical model construction unit has at least the temperature information of the farm included in the breeding environment information, the breed information included in the breeding livestock basic information, and the history of the number of offspring of the breeding livestock included in the breeding history information. A mathematical model construction step to generate a breeding mathematical model for estimating the number of offspring by breeding history of the breeding livestock by training the information into a mathematical model as training data.
The offspring number estimation unit executes on a computer a step of estimating the number of offspring to estimate the number of offspring of the breeding livestock at an arbitrary timing in the future by using the breeding mathematical model generated by the mathematical model construction unit. Livestock information management program to let you.

Images