CN114674782A - Method and device for predicting the content of feces in dairy cows - Google Patents

Abstract

The invention provides a method and a device for predicting the content of fecal components of dairy cows, wherein the method comprises the following steps: acquiring spectral data of a sample to be detected; inputting the spectral data of the sample to be detected into a milk cow manure component content prediction model to obtain the component content of the sample to be detected output by the milk cow manure component content prediction model; the dairy cow fecal component content prediction model is obtained through regression analysis according to a spectrum data set and a component content label set, the spectrum data set comprises a plurality of subsets, and the subsets correspond to labels of the component content label set in a one-to-one mode. The method can quickly predict the content of the components in the excrement of the dairy cow, can calculate the digestibility of the dairy cow more quickly and conveniently according to the content of each component in the excrement, judges the physical condition of the dairy cow, and further modifies the feed formula of each dairy cow in real time, reduces the feed waste and improves the production benefit.

Description

Method and device for predicting the content of feces in dairy cows

technical field

The invention relates to the technical field of component analysis, in particular to a method and a device for predicting the component content of dairy cow feces.

Background technique

Cow feces and feed residues and wastes excreted by dairy cows that cannot be digested and absorbed, including water, feed residues, gastrointestinal microbes, gastrointestinal mucosal cells and their own secretions, etc. Therefore, cow feces carry feed intake A large amount of information such as the situation, feed digestibility and own health status provides rich and reliable data for efficient feeding in production.

At present, routine fecal nutrition evaluation needs to be carried out by laboratory wet chemical analysis methods. The wet chemical analysis methods for various nutrients are different. Usually, various chemical reagents are added during the operation, accompanied by dangerous operations such as burning and heating. The operation process is complex, difficult, and the workload is large. higher quality requirements. In addition, a large number of chemical reagents need to be used in the analysis process, which not only causes huge waste and increases the analysis cost, but also pollutes the environment. Some toxic reagents can even affect the health of operators. It is difficult to promote the use in production.

Near-infrared spectroscopy is a technique for rapidly predicting the content of organic matter by means of near-infrared spectroscopy. It only takes a few minutes to scan a sample. The traditional wet chemical analysis method in the laboratory is time-consuming and labor-intensive. Usually, it takes dozens of hours of measurement time and hundreds of yuan to determine the content of all components. This technology has been widely used in agricultural products such as feed raw materials, but it is still rarely used in livestock and poultry manure.

SUMMARY OF THE INVENTION

The invention provides a method and a device for predicting the content of feces of dairy cows, which are used to solve the defect of complex and difficult to predict the composition of feces of dairy cows under the condition of house feeding in the prior art, and realize accurate measurement.

The present invention provides a method for predicting the content of dairy cow feces, comprising:

Obtain the spectral data of the sample to be tested;

Inputting the spectral data of the sample to be tested into a cow feces component content prediction model to obtain the component content of the tested sample output by the cow feces component content prediction model;

Wherein, the cow feces component content prediction model is obtained by regression analysis based on a spectral data set and a component content label set, the spectral data set includes a plurality of subsets, and the subset and the label of the component content label set are one by one. correspond.

According to a method for predicting component content of dairy cow feces provided by the present invention, the regression analysis is performed according to a spectral data set and a component content label set, including:

According to the subsets included in the spectral data set and the component content labels corresponding to the subsets, regression analysis is performed on the cow feces component content prediction model in turn, and the regression method adopts the partial least squares method.

According to the method for predicting the content of feces of dairy cows provided by the present invention, the subset is obtained by performing spectral preprocessing according to the original spectral data.

According to a method for predicting the content of dairy cow feces components provided by the present invention, the spectral preprocessing includes derivative processing and at least one of the following:

Detrending, standard normal correction, and standard multivariate scatter correction.

According to a method for predicting the content of feces of dairy cows provided by the present invention, the derivative processing method is first-order derivative processing, the spectral interval is 4 nm or 8 nm, the primary smoothing interval is 4 or 8, and the secondary smoothing interval is 1 .

According to a method for predicting the composition content of dairy cow feces provided by the present invention, the composition content label includes: dry matter, crude protein, crude fat, neutral detergent fiber, acid detergent fiber and starch.

The present invention also provides a device for predicting the content of dairy cow feces, comprising:

The acquisition module is used to obtain the spectral data of the sample to be tested;

a prediction module, configured to input the spectral data of the sample to be tested into a cow feces component content prediction model to obtain the component content of the tested sample output by the cow feces component content prediction model;

Wherein, the cow feces component content prediction model is obtained by regression analysis based on a spectral data set and a component content label set, the spectral data set includes a plurality of subsets, and the subset and the label of the component content label set are one by one. correspond.

The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, when the processor executes the program, the dairy cow feces composition described above can be realized by the processor. Content prediction method.

The present invention also provides a non-transitory computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, implements any one of the above-mentioned methods for predicting the content of dairy cow feces.

The present invention also provides a computer program product, including a computer program, which, when executed by a processor, implements any one of the above-mentioned methods for predicting the content of dairy cow feces.

The method and device for predicting the content of cow feces provided by the present invention can rapidly predict the content of cow feces through near-infrared technology. , and then make real-time modifications to the feed formula of each cow to reduce feed waste and improve production efficiency. Precise feeding for each individual, instead of feeding cows in groups of different stages through the traditional grouping method, allows dairy farming production to gradually change from the previous group feeding to individual feeding, from static nutrition to dynamic nutrition .

Description of drawings

In order to explain the present invention or the technical solutions in the prior art more clearly, the following will briefly introduce the accompanying drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are the For some embodiments of the invention, for those of ordinary skill in the art, other drawings can also be obtained according to these drawings without any creative effort.

Fig. 1 is the schematic flow sheet of cow feces component content prediction provided by the present invention;

Fig. 2 is the near-infrared scanning original spectral data diagram of cow feces sample provided by the present invention;

3 is a schematic structural diagram of a device for predicting the content of dairy cow feces components provided by the present invention;

FIG. 4 is a schematic structural diagram of an electronic device provided by the present invention.

Detailed ways

In order to make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention. , not all examples. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

Below in conjunction with Fig. 1-describe a kind of dairy cow manure component content prediction method of the present invention, including:

Step 101, obtaining spectral data of the sample to be tested;

Step 102: Input the spectral data of the sample to be tested into a cow feces component content prediction model to obtain the component content of the tested sample output by the cow feces component content prediction model;

Wherein, the cow feces component content prediction model is obtained by regression analysis based on a spectral data set and a component content label set, the spectral data set includes a plurality of subsets, and the subset and the label of the component content label set are one by one. correspond.

The method for predicting the component content of cow feces in the embodiment of the present invention can predict the content of each component of the sample to be tested by identifying the near-infrared spectral data of the sample to be tested. In this way, the digestibility of dairy cows can be quickly predicted, the condition of dairy cows can be judged, and the feed ratio can be adjusted in time. Compared with the method of calculating the digestibility of dairy cows after the traditional wet chemical method is used to determine the content of fecal components, this method can save a lot of time, realize real-time regulation and precise feeding of nutrition of dairy cows, achieve "one cow for one", and increase the number of cows. Feed utilization efficiency and improve production efficiency.

It should be noted that due to the large differences in the feed used by different dairy farms and dairy cows at different stages, the feces of different dairy cows vary greatly. Therefore, in order to establish a set of prediction models suitable for adult cows at various stages of most cattle farms, a large number of cow dung samples at different stages of different cattle farms are needed as support. Therefore, the spectral data set in the embodiment of the present invention is selected from different cattle farms and mature cows at different stages in various regions of the country, which ensures the representativeness and breadth of the data.

In at least one embodiment of the present invention, the regression analysis is performed according to the spectral data set and the component content label set, including:

According to the subsets included in the spectral data set and the component content labels corresponding to the subsets, regression analysis is sequentially performed on the prediction model of the component content of the cow feces, and a calibration model is established. Regression analysis was performed using the partial least squares method.

It should be noted that the preprocessing of spectral data includes: selecting samples with GH>10.0 and NH<2.5 in the sample set as abnormal samples and eliminating them, and other samples after the elimination form a new sample set.

In at least one embodiment of the present invention, the subset is obtained after spectral preprocessing according to the original spectral data.

In at least one embodiment of the present invention, the spectral preprocessing includes derivative processing and at least one of the following:

Detrending, standard normal correction, standard multivariate scatter correction, or scatter-free processing.

In at least one embodiment of the present invention, the derivative processing method is first-order derivative processing, the spectral interval is 4 nm or 8 nm, the primary smoothing interval is 4 or 8, and the secondary smoothing interval is 1.

The spectral preprocessing method of the dry matter calibration model adopts the detrend processing, the derivative processing adopts the 1st derivative, the spectral interval point is 8 nm, the smoothing interval point value is 8 points, and the second smoothing interval point value is 1; The spectral preprocessing method of the protein calibration model adopts standard normal correction (SNV) processing, the derivative processing adopts the first derivative, the spectral interval point is 4 nm, the smoothing interval point value is 4 points smooth, and the second smoothing interval point value is 1; The spectral preprocessing method of the crude fat calibration model adopts the standard multivariate scatter correction (MSC) processing, the derivative processing adopts the 1st derivative, the spectral interval point is 4nm, the smoothing interval point value is 4 points, and the second smoothing interval point value is 1 The spectral preprocessing method of the neutral detergent fiber calibration model adopts standard normal correction (SNV) processing, the derivative processing adopts the first derivative, the spectral interval point is 8 nm, and the smoothing interval point value is 8 points for smoothing, and the second smoothing interval point The value is 1; the spectral preprocessing method of the acid washing fiber calibration model adopts standard normal correction (SNV) and detrend (Detrend) processing, the derivative processing adopts the first derivative, the spectral interval is 4 nm, and the smoothing interval is 4. Point smoothing, the value of the second smoothing interval point is 1; the starch calibration model spectral preprocessing method adopts the detrend (Detrend) processing, the derivative processing adopts the 1st derivative, the spectral interval point is 4 nm, and the smoothing processing interval point value is 4 points smoothing , the quadratic smoothing interval point value is 1. The spectral data and wet chemical components were regressed by partial least squares method, and a near-infrared prediction model for the content of cow feces was established.

After the calibration is completed, the calibration results can be verified through calibration standard analysis error (SEC), calibration determination coefficient (RSQ), cross-validation standard error (SECV), and cross-validation correlation coefficient (1-VR). The accuracy and stability of the model were evaluated. The smaller the SEC and SECV, the closer the RSQ and 1-VR are to 1, the better the model regression is.

In at least one embodiment of the present invention, the ingredient content label includes: dry matter (DM), crude protein (CP), crude fat (EE), neutral detergent fiber (NDF), acid detergent fiber (ADF), and starch.

In at least one embodiment of the present invention, the whole process of establishing a prediction model for the content of feces of dairy cows by near-infrared technology under house feeding conditions is given:

Step 201, collecting cow feed samples;

Collect different total mixed rations (TMR) for dairy cows, and store the samples at -20 °C for storage. After the test, they are placed in a 65 °C oven for 48 hours. After drying, they are placed at room temperature for 24 hours, and then pulverized by a micro pulverizer. Air-dried samples were made and stored for testing.

Step 202, the collection of cow feces samples

The feces were collected by the rectal method, 250 g each time, 12 times in total, and the samples were stored at -20°C for cryopreservation. After sampling, thaw the fecal samples, mix the fecal samples of the same cow evenly, take 400 g, add 100 g of 10% tartaric acid, and then place the fecal samples in a 65°C oven for 48 hours, and then place them at room temperature to regain moisture for 24 hours. The air-dried samples were made into air-dried samples after being pulverized by a micro-pulverizer and stored for testing.

Step 203, measuring the conventional nutrient content of the cow feed and feces samples;

Determination of Dry Matter (DM), Crude Protein (CP), Crude Fat (EE), Neutral Detergent Fiber (NDF), Acid Detergent Fiber (ADF), Starch (Starch) in Cow Feed and Fecal Samples Using Laboratory Wet Chemistry content. See Table 1 for details.

Table 1 Laboratory wet chemical determination of nutrients

Step 204, collecting the spectral data of the cow feces sample;

FOSS NIRS DS 2500F was used to scan the fecal samples by near-infrared spectroscopy. The pretreated dairy cow feces samples were evenly placed into the FOSS special sample cup, and the cup was filled with samples that accounted for about 1/2 of the cup volume, and the scanning was started. The spectral scanning band range is 850-2500nm, and the wavelength accuracy is 2nm. The original near-infrared scanning spectrum of the cow feces sample is shown in Figure 1.

As shown in Figure 2, in the near-infrared scanning raw data spectrum of the cow feces sample, each sample represents a curve. It can be seen that the absorption spectrum trends of each sample are the same, and the composition of absorption peaks is also relatively close; the absorption spectrum curve fluctuates up and down. Therefore, the composition of fecal samples is similar, but the content of the components varies greatly.

Step 205, establishing a prediction model for the content of dairy cow feces;

Using WinISIⅢ analysis software, the spectral data set was compared and regression analysis with the data determined by laboratory wet chemical method, and the prediction model of dairy cow dung composition content was established. The model can be used directly in a near-infrared spectroscopy analyzer to rapidly predict dry matter, crude protein, crude fat, neutral detergent fiber, acid detergent fiber, and starch in cow manure.

Table 2 The optimal scaling model for predicting the content of feces in dairy cows

It can be seen from Table 2 that the results of the calibration and verification of the prediction model for the content of dairy cow feces are as follows:

SEC of dry matter was 0.3364, SECV was 0.3939, RSQ was 0.9809, 1-VR was 0.9737; SEC of crude protein was 0.6937, SECV was 0.7761, RSQ was 0.9255, 1-VR was 0.9063; SEC of neutral washed fiber was 2.4412 , SECV is 2.6024, RSQ is 0.8148, 1-VR is 0.7897; SEC of acid detergent fiber is 0.8874, SECV is 1.2103, RSQ is 0.7416, 1-VR is 0.5565; SEC of crude fat is 0.2622, SECV is 0.2909, RSQ is 0.9140, 1-VR was 0.8936; SEC of starch was 0.1012, SECV was 0.1219, RSQ was 0.7924, and 1-VR was 0.6978. The calibration model of dry matter, crude protein and crude fat has good effect, and the calibration effect of neutral detergent fiber, acid detergent fiber and starch is better.

The following describes the device for predicting the content of feces of dairy cows provided by the present invention, and the device for predicting the content of feces of cows described below and the method for predicting the content of feces of cows described above can be referred to each other correspondingly.

As shown in FIG. 3 , an embodiment of the present invention discloses a device for predicting the content of dairy cow feces, including:

The acquisition module 301 is used for acquiring spectral data of the sample to be tested;

A prediction module 302, configured to input the spectral data of the sample to be tested into a cow feces component content prediction model to obtain the component content of the tested sample output by the cow feces component content prediction model;

Wherein, the cow feces component content prediction model is obtained by regression analysis based on a spectral data set and a component content label set, the spectral data set includes a plurality of subsets, and the subset and the label of the component content label set are one by one. correspond.

The device for predicting the component content of cow feces in the embodiment of the present invention can identify the components of the sample to be tested and the content of each component by identifying the near-infrared spectral data of the sample to be tested. In this way, the digestibility can be predicted, the condition of dairy cows can be judged, and the feed ratio can be adjusted in time to improve production efficiency.

In at least one embodiment of the present invention, the regression analysis is performed according to the spectral data set and the component content label set, including:

According to the subsets included in the spectral data set and the component content labels corresponding to the subsets, regression analysis is performed on the dairy cow feces component content prediction model in sequence.

In at least one embodiment of the present invention, the subset is obtained after spectral preprocessing according to the original spectral data.

In at least one embodiment of the present invention, the spectral preprocessing includes partial least squares and at least one of the following:

Detrending, standard normal correction, and standard multivariate scatter correction.

In at least one embodiment of the present invention, the ingredient content label includes: dry matter, crude protein, crude fat, neutral detergent fiber, acid detergent fiber, and starch.

FIG. 4 illustrates a schematic diagram of the physical structure of an electronic device. As shown in FIG. 4 , the electronic device may include: a processor (processor) 410, a communication interface (Communications Interface) 420, a memory (memory) 430, and a communication bus 440, The processor 410 , the communication interface 420 , and the memory 430 communicate with each other through the communication bus 440 . The processor 410 can invoke logic instructions in the memory 430 to execute a method for predicting the content of feces in dairy cows, the method comprising:

Obtain the spectral data of the sample to be tested;

Inputting the spectral data of the sample to be tested into a cow feces component content prediction model to obtain the component content of the tested sample output by the cow feces component content prediction model;

Wherein, the cow feces component content prediction model is obtained by regression analysis based on a spectral data set and a component content label set, the spectral data set includes a plurality of subsets, and the subset and the label of the component content label set are one by one. correspond.

In addition, the above-mentioned logic instructions in the memory 430 can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention can be embodied in the form of a software product in essence, or the part that contributes to the prior art or the part of the technical solution. The computer software product is stored in a storage medium, including Several instructions are used to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, Read-Only Memory (ROM, Read-Only Memory), Random Access Memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program codes .

In another aspect, the present invention also provides a computer program product, the computer program product includes a computer program, the computer program can be stored on a non-transitory computer-readable storage medium, and when the computer program is executed by a processor, the computer can Execute the method for predicting the content of dairy cow feces provided by the above methods, and the method includes:

Obtain the spectral data of the sample to be tested;

Inputting the spectral data of the sample to be tested into a cow feces component content prediction model to obtain the component content of the tested sample output by the cow feces component content prediction model;

Wherein, the cow feces component content prediction model is obtained by regression analysis based on a spectral data set and a component content label set, the spectral data set includes a plurality of subsets, and the subset and the label of the component content label set are one by one. correspond.

In yet another aspect, the present invention also provides a non-transitory computer-readable storage medium on which a computer program is stored, and the computer program is implemented when executed by a processor to execute the methods for predicting the content of feces of dairy cows provided by the above methods. Methods include:

Obtain the spectral data of the sample to be tested;

Inputting the spectral data of the sample to be tested into a cow feces component content prediction model to obtain the component content of the tested sample output by the cow feces component content prediction model;

Wherein, the cow feces component content prediction model is obtained by regression analysis based on a spectral data set and a component content label set, the spectral data set includes a plurality of subsets, and the subset and the label of the component content label set are one by one. correspond.

The device embodiments described above are only illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in One place, or it can be distributed over multiple network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution in this embodiment. Those of ordinary skill in the art can understand and implement it without creative effort.

From the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and certainly can also be implemented by hardware. Based on this understanding, the above-mentioned technical solutions can be embodied in the form of software products in essence or the parts that make contributions to the prior art, and the computer software products can be stored in computer-readable storage media, such as ROM/RAM, magnetic A disc, an optical disc, etc., includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform the methods described in various embodiments or some parts of the embodiments.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, but not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that it can still be The technical solutions described in the foregoing embodiments are modified, or some technical features thereof are equivalently replaced; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims (10)

1. A method for predicting the content of fecal components of dairy cows is characterized by comprising the following steps:

acquiring spectral data of a sample to be detected;

inputting the spectral data of the sample to be detected into a milk cow manure component content prediction model to obtain the component content of the sample to be detected output by the milk cow manure component content prediction model;

the dairy cow fecal component content prediction model is obtained through regression analysis according to a spectrum data set and a component content label set, the spectrum data set comprises a plurality of subsets, and the subsets correspond to labels of the component content label set in a one-to-one mode.

2. The method for predicting the fecal component content of dairy cows according to claim 1, wherein the performing regression analysis based on the spectral data set and the component content tag set comprises:

and performing regression analysis on the cow excrement component content prediction model in sequence according to the subsets contained in the spectral data set and the component content labels corresponding to the subsets, wherein the regression method adopts a partial least square method.

3. The method for predicting the fecal component content of dairy cows according to claim 1 or 2, wherein the subset is obtained by performing a spectrum preprocessing on the raw spectrum data.

4. The method of predicting the fecal component content of dairy cows of claim 3, wherein the spectral pre-processing comprises derivative processing and at least one of:

detrending, standard normal correction, and standard multivariate scatter correction.

5. The method for predicting the fecal component content of dairy cows according to claim 4, wherein the derivative processing method is first derivative processing, the spectral interval point is 4nm or 8nm, the first smoothing interval value is 4 or 8, and the second smoothing interval point value is 1.

6. The method for predicting the fecal component content of dairy cows according to claim 1 or 2, wherein the component content label comprises: dry matter, crude protein, crude fat, neutral detergent fiber, acid detergent fiber, and starch.

7. A milk cow excrement composition content prediction device which characterized in that includes:

the acquisition module is used for acquiring spectral data of a sample to be detected;

the prediction module is used for inputting the spectral data of the sample to be detected into a milk cow excrement component content prediction model to obtain the component content of the sample to be detected output by the milk cow excrement component content prediction model;

the dairy cow fecal component content prediction model is obtained through regression analysis according to a spectrum data set and a component content label set, the spectrum data set comprises a plurality of subsets, and the subsets correspond to labels of the component content label set in a one-to-one mode.

8. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method for predicting the content of cow's dung components according to any one of claims 1 to 6 when executing the program.

9. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the dairy cow fecal component content prediction method according to any one of claims 1 to 6.

10. A computer program product comprising a computer program, wherein the computer program, when executed by a processor, implements the dairy cow fecal component content prediction method according to any one of claims 1 to 6.

Images