CN111027822A

CN111027822A - Method and device for determining feed type

Info

Publication number: CN111027822A
Application number: CN201911164681.4A
Authority: CN
Inventors: 陶兴源; 陆杰; 吴明辉
Original assignee: Miaozhen Information Technology Co Ltd
Current assignee: Miaozhen Information Technology Co Ltd
Priority date: 2019-11-25
Filing date: 2019-11-25
Publication date: 2020-04-17
Anticipated expiration: 2039-11-25
Also published as: CN111027822B

Abstract

The application provides a method and a device for determining a feed type, wherein the method comprises the following steps: before feeding livestock, collecting the current weight value and the current feeding time of the livestock; determining the type of feed to be fed to the livestock according to the current precursor weight value and the current feeding time; and the current feeding time length is the difference value between the current time and the initial feeding time. The technical scheme can determine the feed type of a single pig, thereby realizing individual feeding.

Description

Method and device for determining feed type

Technical Field

The invention relates to the field of computers, in particular to a method and a device for determining feed types.

Background

In the traditional breeding industry, the characteristics of pigs are researched by experts, and a long-term tracking test is carried out to obtain a time node for replacing the feed of the pigs. The replacement of the feed is to save cost and ensure that the pigs only grow up quickly. In the traditional method, experts are required to randomly select a plurality of groups of test pigs to carry out a large number of comparison tests, and the time node conclusion of feed replacement is obtained by carrying out the tests under various fattening conditions.

In the prior art, the feed used by a certain type of pigs is mainly judged through expert tests. From suckling pig to weaning pig, from weaning pig to growing pig and from growing pig to fattening pig, all the stages of feeding which kind of feed are operated according to the conclusion obtained by expert test.

The following problems exist in the manner of expert guidance: 1) the experience data is obtained through experiments by experts with abundant experience, the period is long, and the ordinary pig farm hardly requires the experts to stay for a long time for guidance. 2) In the traditional method, multiple pigs serve as a test group, a plurality of test groups are subjected to comparative tests, and the conclusion obtained by the tests is usually specific to a certain type of pigs, so that the individual conclusion of a single pig cannot be obtained. 3) For a pig farm without expert guidance, the pigs in the farm may be different from the initial condition or growth condition of the pigs tested by the expert, and if the expert's experience value is applied, there is generally an error of about 10 days. 4) Experts have limited accuracy in their test data.

Disclosure of Invention

The technology to be solved by the application is to provide a method and a device for determining the feed type, which can determine the feed type of a single pig, thereby realizing individual feeding.

In order to solve the above technical problem, the present application provides a method for determining a feed type, the method comprising:

before feeding livestock, collecting the current weight value and the current feeding time of the livestock;

determining the type of feed to be fed to the livestock according to the current precursor weight value and the current feeding time;

and the current feeding time length is the difference value between the current time and the initial feeding time.

Optionally, the determining the type of feed corresponding to the livestock according to the current body weight value and the current feeding time period comprises:

and inputting the current precursor weight value and the current feeding time into a trained feed prediction model to obtain the type of feed to be fed by the livestock.

Optionally, the feed prediction model is a model trained according to a farm to which the livestock belong.

Optionally, the feed prediction model is trained by:

for a plurality of test livestock selected in the feedlot, collecting weight value data and feed type information of feeding of each test livestock from initial feeding to qualified slaughtering;

training to obtain a feed prediction model corresponding to the feeding farm according to the set reward corresponding to each experimental livestock for each feeding;

wherein, the reward of each feeding is the sum of the following three items: the weight value coefficient, the reciprocal of the feed price coefficient and the reciprocal of the natural logarithm of the feeding time;

the weight value coefficient is the absolute value of the difference value of the weight values acquired by the livestock twice;

the feed price coefficient is obtained by normalizing the unit price of fed feed to the interval of [0,1 ].

Optionally, the method further comprises:

storing the collected identification information, weight value and feeding duration of the livestock into a model training data pool;

and when the data in the model training data pool meets the set conditions, updating the feed prediction model according to the data in the model training data pool.

The present application also provides an apparatus for determining a type of feed, the apparatus comprising: a memory and a processor;

the memory is used for storing a program for determining the type of the feed;

the processor is used for reading and executing the program for determining the feed type and executing the following operations:

Optionally, the feed prediction model is trained by:

Optionally, the processor is configured to read and execute the program for determining the type of the feed, and further perform the following operations:

The application includes: before feeding livestock, collecting the current weight value and the current feeding time of the livestock; determining the type of feed to be fed to the livestock according to the current precursor weight value and the current feeding time; and the current feeding time length is the difference value between the current time and the initial feeding time. The technical scheme can determine the feed type of a single pig, thereby realizing individual feeding.

Drawings

The accompanying drawings are included to provide an understanding of the present disclosure and are incorporated in and constitute a part of this specification, illustrate embodiments of the disclosure and together with the examples serve to explain the principles of the disclosure and not to limit the disclosure.

FIG. 1 is a flow chart of a method for determining feed type according to a first embodiment of the present invention;

FIG. 2 is a schematic structural diagram of a device for determining the type of feed according to a first embodiment of the present invention;

FIG. 3 is another flow chart of a method for determining feed type according to a first embodiment of the present invention;

fig. 4 is a schematic training diagram of a feed prediction model according to a first embodiment of the present invention.

Detailed Description

The present application describes embodiments, but the description is illustrative rather than limiting and it will be apparent to those of ordinary skill in the art that many more embodiments and implementations are possible within the scope of the embodiments described herein. Although many possible combinations of features are shown in the drawings and discussed in the detailed description, many other combinations of the disclosed features are possible. Any feature or element of any embodiment may be used in combination with or instead of any other feature or element in any other embodiment, unless expressly limited otherwise.

The present application includes and contemplates combinations of features and elements known to those of ordinary skill in the art. The embodiments, features and elements disclosed in this application may also be combined with any conventional features or elements to form a unique inventive concept as defined by the claims. Any feature or element of any embodiment may also be combined with features or elements from other inventive aspects to form yet another unique inventive aspect, as defined by the claims. Thus, it should be understood that any of the features shown and/or discussed in this application may be implemented alone or in any suitable combination. Accordingly, the embodiments are not limited except as by the appended claims and their equivalents. Furthermore, various modifications and changes may be made within the scope of the appended claims.

Further, in describing representative embodiments, the specification may have presented the method and/or process as a particular sequence of steps. However, to the extent that the method or process does not rely on the particular order of steps set forth herein, the method or process should not be limited to the particular sequence of steps described. Other orders of steps are possible as will be understood by those of ordinary skill in the art. Therefore, the particular order of the steps set forth in the specification should not be construed as limitations on the claims. Further, the claims directed to the method and/or process should not be limited to the performance of their steps in the order written, and one skilled in the art can readily appreciate that the sequences may be varied and still remain within the spirit and scope of the embodiments of the present application.

Example one

As shown in fig. 1, the present embodiment provides a method of determining a feed type, the method comprising:

step S101, collecting the current weight value and the current feeding time of livestock before feeding the livestock;

step S102, determining the type of feed to be fed to the livestock according to the current weight value and the current feeding time;

Optionally, the determining the type of the feed corresponding to the livestock according to the current weight value and the current feeding time period may include:

Alternatively, the feed prediction model may be a model trained from a farm to which the livestock belong.

Alternatively, the feed prediction model may be trained by:

Optionally, the method may further include:

According to the technical scheme, the feed type of a single pig can be determined without expert intervention, so that individual feeding is realized.

As shown in fig. 2, the present embodiment also provides an apparatus for determining a type of feed, the apparatus comprising: a memory 10 and a processor 11;

the memory 10 is used for storing a program for determining the type of the feed;

the processor 11 is configured to read and execute the program for determining the type of the feed, and perform the following operations:

Alternatively, the feed prediction model may be trained by:

Optionally, the processor is configured to read and execute the program for determining the type of the feed, and may further perform the following operations:

The method of determining feed type of the present application is further illustrated below by taking a pig farm as an example, as shown in fig. 3.

S301, training a corresponding feed prediction model of a pig farm;

as shown in fig. 4, which is a schematic diagram of the feed prediction model training, when the model is trained, a certain number of test pigs can be selected from the pig farm, and data information in the complete process from the beginning of feeding to the qualified slaughter of each test pig is collected through a sensor, where the data information includes weight information of each test pig and the type of feed fed each time.

The reward per feeding can be set as the sum of the following three items: the weight value coefficient, the reciprocal of the feed price coefficient and the reciprocal of the natural logarithm of the feeding time; wherein the weight value coefficient is the absolute value of the difference value of the weight values acquired by the livestock twice; the feed price coefficient is obtained by normalizing the unit price of fed feed to the interval of [0,1 ]. By the set reward, when the trained feed prediction model aims at achieving the slaughtered weight, the larger the weight increase, the better the feed cost and the shorter the feeding date. The feed prediction model had the input of weight data and feeding duration for each test pig and the output was the feed type.

For each step in each training round, the state value network predicts the reward expected to be obtained by feeding each feed type, selects the feed type with the maximum predicted reward, and stores the current state, the selected feed type, the obtained reward and the next state in a playback memory unit. And randomly reading part of records from the playback memory unit, and carrying out gradient back propagation training model by using the residual error of the target value network and the state value network. And assigning the parameters of the state value network to the target value network for updating at regular intervals of training steps. And when the weight value of the test pig reaches the slaughtering standard weight, the feed prediction model finishes one round of training and returns the successful reward value of the feed prediction model.

S302, loading a feed prediction model corresponding to a feeding farm to which the livestock belongs;

the trained feed prediction model of the feeding farm can be loaded into local pork pig fattening feed changing prediction online service, so that the type of the fed feed of each pig is predicted through the feed prediction model.

Step S303, collecting the current weight value and the current feeding time of the pig before feeding the pig;

s304, determining the type of the feed to be fed by the pig by using a feed prediction model;

in this example, the current weight and the current feeding time of the pig can be collected by the sensor. Then, the current weight and the current feeding time of the pig are input into a pork pig fattening feed changing prediction online service (the pork pig fattening feed changing prediction online service is loaded with a feed prediction model), the pork pig fattening feed changing prediction online service can predict the optimal feed type for the current pig individual, and then the determined feed type is fed to the current pig individual.

And S305, feeding each pig with the determined corresponding feed type.

The technical scheme has the following beneficial technical effects:

1) in the aspect of collecting data information of a test pig, the difference between the example and the traditional method is that the data information is collected for a single pig, and the application uses a big data method and takes each pig as a test unit, so that test data can be used more efficiently. Meanwhile, the abnormal value detection method of big data is utilized, so that the problem of abnormal test data of individual pigs can be effectively avoided. Through the technical scheme, the utilization rate of data can be improved, and the time period required by the test is further saved.

2) Each pig farm can train a model suitable for the pig farm, and the problem of expert knowledge deviation caused by different environments (temperature, humidity, longitude and latitude, feed and the like) is avoided, so that the optimal model can be trained for each pig farm.

3) In the example, a reinforcement learning model is trained according to test data of each type of pig, and the trained model can be remotely loaded into a local system of a pig farm in a network deployment mode. Thereby solving the problem that a plurality of pig farms are difficult to invite experts to stay on site to guide the replacement of pig feed.

4) The acquisition of the test data of the present example requires no expert intervention. If the farmer wants to customize the model for his own pig farm, but no expert provides prior knowledge to select the initial value of the test parameter (for example, the time range of the test for refueling can be preliminarily selected by the expert as the initial value), the system can give a suggested initial value based on the statistical result accumulated by the big data. Therefore, the example can help a pig farm without expert stationing, and a corresponding reinforcement learning model is trained according to the type of the pig farm.

5) When judging which kind of feed type the pig should only feed, the reinforcement learning model judges for each individual pig, and by utilizing the characteristics of the collected individual pig, the reinforcement learning model transmits the judged feed type to the feed feeding system so as to feed the pig according to the determined feed type. Thereby realizing accurate feeding.

It should be noted that the above technical solution is also applicable to other breeding plants, such as cattle farms, sheep farms, etc.

It will be understood by those of ordinary skill in the art that all or some of the steps of the methods, systems, functional modules/units in the devices disclosed above may be implemented as software, firmware, hardware, and suitable combinations thereof. In a hardware implementation, the division between functional modules/units mentioned in the above description does not necessarily correspond to the division of physical components; for example, one physical component may have multiple functions, or one function or step may be performed by several physical components in cooperation. Some or all of the components may be implemented as software executed by a processor, such as a digital signal processor or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). The term computer storage media includes volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, as is well known to those of ordinary skill in the art. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, Digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can accessed by a computer. In addition, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media as known to those skilled in the art.

Claims

1. A method of determining a feed type, the method comprising:

2. The method of claim 1, wherein determining the type of feed corresponding to the livestock based on the current weight and current feed length comprises:

3. The method of claim 2, wherein:

the feed prediction model is a model obtained by training according to a feedlot to which the livestock belongs.

4. The method of claim 3, wherein:

the feed prediction model is obtained by training in the following way:

5. The method of any of claims 2 to 4, further comprising:

6. An apparatus for determining a type of feed, the apparatus comprising: a memory and a processor; the method is characterized in that:

the memory is used for storing a program for determining the type of the feed;

7. The apparatus as claimed in claim 6, wherein the determining the type of feed corresponding to the livestock according to the current weight and the current feeding time period comprises:

8. The apparatus of claim 7, wherein:

9. The apparatus of claim 8,

the feed prediction model is obtained by training in the following way:

10. The apparatus of any one of claims 7 to 9, wherein the processor, configured to read and execute the program for determining the type of feed, further performs the following operations: