Poultry health state intelligent diagnosis method based on dynamic step counting

Abstract

The invention discloses a poultry health state intelligent diagnosis method based on dynamic step counting, which comprises the following steps: installing a step counting foot ring on a leg of poultry to be monitored to obtain the step number of the poultry at regular time, and obtaining the body temperature of the poultry through a wing label; constructing a poultry database; generating a breeding environment big data thermodynamic diagram based on a preset grid size on the basis of the individual positions, the exercise amount, the feed consumption, the environmental parameters and the poultry health state of the poultry, constructing a poultry group density distribution prediction model by using a hierarchical learning network, and predicting poultry group density distribution; and in combination with the birth days of the poultry and the growth and development rules of the poultry, constructing an intelligent diagnosis model of the exercise amount, the feed dosage, the environmental parameters and the abnormal state by using a graphic space-time network, and predicting the health state of the poultry. The poultry health detection precision and accuracy can be effectively improved, and the poultry health condition can be visually reflected.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and particularly to an intelligent diagnosis method for the health status of poultry based on dynamic step counting. Background Art

[0002] With the gradual improvement of people's living standards, the demand for poultry is increasing. However, in the process of breeding, many poultry die because their diseases are not detected in time, and even if they are rescued in time, they do not have sufficient nutritional value. Therefore, there is an urgent need for a method that can immediately diagnose the health of poultry and display the diagnosis information in the form of a score to improve the survival rate of free-range poultry. Summary of the Invention

[0003] The purpose of the present invention is to provide an intelligent diagnosis method for the health status of poultry based on dynamic step counting, which can effectively improve the accuracy and precision of poultry health detection, and at the same time can intuitively reflect the health status of poultry.

[0004] To achieve the above technical purpose, the technical solution adopted by the present invention is as follows:

[0005] The present invention discloses an intelligent diagnosis method for the health status of poultry based on dynamic step counting, and the method includes the following steps:

[0006] S1. Install a step-counting ankle bracelet on the leg of the poultry to be monitored to obtain the number of steps of the poultry at regular intervals, and obtain the body temperature of the poultry through a wing tag; preprocess the collected data to obtain the individual position, exercise amount, and body temperature change degree of the poultry;

[0007] S2. Construct a poultry database, which stores the breed, birth days, environmental parameters, feed consumption, individual position, exercise amount, body temperature change degree, and health status of the poultry;

[0008] S3. Based on the individual position, exercise amount, feed consumption, environmental parameters, and health status of the poultry, generate a big data heat map of the breeding environment based on a preset grid size, and use a hierarchical learning network to construct a prediction model for the density distribution of the poultry population to predict the density distribution of the poultry population;

[0009] S4. Combine the birth days of the poultry and the growth and development laws of the poultry, and use a graphical spatio-temporal network to construct an intelligent diagnosis model for the exercise amount, feed consumption, environmental parameters, and abnormal status to predict the health status of the poultry.

[0010] Further, the environmental parameters include forest and grass parameters, temperature, and humidity.

[0011] Further, the health status of the poultry includes growth conditions and disease infection conditions.

[0012] Further, the poultry health status is comprehensively calculated from the amount of exercise, feed consumption, body temperature change, and the density distribution of the poultry population.

[0013] Further, the grid size is 1 meter * 1 meter.

[0014] Further, the poultry population density distribution prediction model includes an input module, a first convolutional module, a second convolutional module, a third convolutional module, a fourth convolutional module, and a BLSTM module connected in sequence;

[0015] The input module constructs a big data heat map of the breeding environment with a preset grid size based on the input grid density, individual position, amount of exercise, diet, environmental parameters, and poultry health status data; the first convolutional module, the second convolutional module, and the third convolutional module, and the fourth convolutional module sequentially perform feature extraction on the input big data heat map of the breeding environment, and input the finally extracted feature map into the BLSTM module to predict the density distribution of the poultry population for a period of time in the future.

[0016] Further, the method further includes:

[0017] Manually calibrating the intelligent diagnosis model by means of manual labeling.

[0018] Further, the manual labeling method includes labeling abnormal data, obvious cases, and misjudged data

[0019] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0020] The intelligent diagnosis method for poultry health status based on dynamic step counting of the present invention takes poultry as the research object, collects the amount of exercise of poultry as estimation data through the acceleration sensor configured on the ankle ring, preprocesses the collected acceleration sensor data to obtain high-precision poultry exercise amount data; combines the body temperature and air temperature of the same individual at different time periods for statistics and data distribution, conducts big data modeling, obtains a poultry health relationship step model through deep learning and other methods, and improves the accuracy of the big data model and ensures the reliability of the model through manual labeling processing, sets a scoring system, scores poultry with different health levels, and makes the health status of poultry more intuitive. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 is a flowchart of the intelligent diagnosis method for poultry health status based on dynamic step counting of the present invention;

[0022] Figure 2 is a schematic structural diagram of the poultry population density distribution prediction model of the present invention;

[0023] Figure 3Schematic diagram of the intelligent diagnosis model of the present invention. Detailed implementation manners

[0024] The embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0025] The present invention discloses an intelligent diagnosis method for the health status of poultry based on dynamic step counting. The method includes the following steps:

[0026] S1, install a step-counting anklet on the leg of the poultry to be monitored to obtain the number of steps of the poultry at regular intervals, and obtain the body temperature of the poultry through a wing tag; preprocess the collected data to obtain the individual position, exercise amount, and body temperature change degree of the poultry;

[0027] S2, construct a poultry database, which stores the breed, birth days, environmental parameters, feed consumption, individual position, exercise amount, body temperature change degree, and health status of the poultry;

[0028] S3, based on the individual position, exercise amount, feed consumption, environmental parameters, and health status of the poultry, generate a big data heat map of the breeding environment based on a preset grid size, and use a hierarchical learning network to construct a prediction model for the density distribution of the poultry population to predict the density distribution of the poultry population;

[0029] S4, combine the birth days of the poultry and the growth and development laws of the poultry, and use a graphical spatio-temporal network to construct an intelligent diagnosis model for the exercise amount, feed consumption, environmental parameters, and abnormal status to predict the health status of the poultry.

[0030] The present invention performs big data modeling on the data and establishes a big data model belonging to poultry through deep learning and neural networks. Preferably, the present invention can also manually calibrate the big data model by means of artificial labels to improve the accuracy of the big data model.

[0031] The big data model modeling includes using deep learning and neural network modeling methods. The data processing performed by the big data model includes step stability, body temperature change degree, high temperature weather value, statistical analysis by growth stage, and feed consumption, etc. See Figure 2 , to scientifically identify the spatio-temporal distribution of the poultry population in the leisure ranch and support refined breeding, based on GIS grids and technologies, using data such as the individual position, exercise amount, diet amount, environmental parameters, forest and grass system, and growth status of the poultry, generate a big data heat map of the breeding environment based on a grid (1*1 meter), and use the AI algorithm of a hierarchical learning network (CBLSTM) to construct a prediction model for the density distribution of the poultry population. See Figure 3 , combine the animal growth and development laws, and use the AI algorithm of a graphical spatio-temporal network (GSTN) to construct an intelligent diagnosis model for the exercise amount, diet amount, environmental parameters, and abnormal status to support the grid-based and digital management of the leisure ranch.

[0032] The present invention takes poultry as the research object, collects the exercise amount of poultry as estimation data through the acceleration sensors configured on the leg rings, preprocesses the collected acceleration sensor data to obtain high-precision poultry exercise amount data; combines the body temperature and air temperature of the same individual at different time periods for statistics and data distribution, conducts big data modeling, obtains a poultry health relationship step model through methods such as deep learning, and improves the accuracy of the big data model and ensures the reliability of the model after artificial labeling processing. A scoring system is set up to score poultry with different health levels, making the health status of poultry more intuitive.

[0033] Although the preferred embodiments of the present application have been described, those skilled in the art can make additional changes and modifications to these embodiments once they learn the basic creative concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications falling within the scope of the present application.

[0034] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application is also intended to include these modifications and variations.

Claims

1. An intelligent diagnosis method for poultry health status based on dynamic step counting, characterized in that: The method comprises the following steps: S1, install a pedometer foot ring on the leg of the poultry to be monitored to regularly obtain the number of steps of the poultry, and obtain the body temperature of the poultry through the wing mark; pre-process the collected data to obtain the individual position, movement amount and body temperature change of the poultry; S2, constructing a poultry database, wherein the poultry database stores the breed, birth days, environmental parameters, feed usage, individual location, exercise volume, body temperature variation and health status of the poultry; S3, based on the individual location, exercise volume, feed dosage, environmental parameters and health status of poultry, generates a big data heat map of the breeding environment based on a preset grid size, and uses a hierarchical learning network to build a poultry population density distribution prediction model to predict the poultry population density distribution; S4, combining the days of birth of poultry and the laws of poultry growth and development, uses a graph spatiotemporal network to build an intelligent diagnostic model of exercise volume, feed usage, environmental parameters and abnormal conditions to predict the health status of poultry.

2. The intelligent diagnosis method for poultry health status based on dynamic pedometer according to claim 1, characterized in that: The environmental parameters include forest and grass parameters, temperature and humidity.

3. The intelligent diagnosis method for poultry health status based on dynamic pedometer according to claim 1, characterized in that: The poultry health status includes growth status and disease infection status.

4. The intelligent diagnosis method for poultry health status based on dynamic pedometer according to claim 1, characterized in that: The poultry health status is obtained by comprehensive calculation of the amount of exercise, feed usage, body temperature variation and poultry population density distribution.

5. The intelligent diagnosis method for poultry health status based on dynamic pedometer according to claim 1, characterized in that: The grid size is 1 meter*1 meter.

6. The intelligent diagnosis method for poultry health status based on dynamic pedometer according to claim 1, characterized in that: The poultry population density distribution prediction model comprises an input module, a first convolution module, a second convolution module, a third convolution module, a fourth convolution module and a BLSTM module which are connected in sequence; The input module constructs a breeding environment big data heat map of a preset grid size according to the input grid density, individual position, exercise volume, diet volume, environmental parameters and poultry health status data; the first convolution module, the second convolution module, the third convolution module, and the fourth convolution module extract features from the input breeding environment big data heat map in turn, and input the finally extracted feature map into the BLSTM module to predict the poultry population density distribution in the future.

7. The intelligent diagnosis method for poultry health status based on dynamic pedometer according to claim 1, characterized in that: The method further comprises: The intelligent diagnosis model is manually calibrated through manual labeling.

8. The intelligent diagnosis method for poultry health status based on dynamic pedometer according to claim 7, characterized in that: Manual labeling methods include marking abnormal data, obvious cases, and misjudged data.

Images