CN-122004168-A - Unmanned accurate feeding method, unmanned accurate feeding device, unmanned accurate feeding system, unmanned accurate feeding electronic equipment and unmanned accurate feeding storage medium

CN122004168ACN 122004168 ACN122004168 ACN 122004168ACN-122004168-A

Abstract

The invention provides an unmanned accurate feeding method, device, system, electronic equipment and storage medium for a farm, which relate to the technical field of livestock feeding, the method combines the historical feeding information to feed, can reduce the deviation from the current feed formula to the actual intake of livestock, improve the feed utilization rate, reduce the carbon emission and reduce the feeding cost. The method also considers the current feeding requirements of livestock, can realize accurate and personalized feeding management, and can further ensure the high consistency of the latest feeding formula and the actual intake of the livestock by continuously optimizing the current feed formula for feeding, thereby improving the production performance of the livestock, avoiding the waste of the feed, ensuring the maximization of the cultivation income, and simultaneously ensuring the health and the production efficiency of the livestock. Moreover, the method can realize automatic feeding by using a feeding operation system, greatly reduces manual intervention, improves the efficiency and accuracy of feeding management, and provides powerful support for large-scale livestock cultivation.

Inventors

LI BIN
YUE JIANMIN
WANG HAIFENG
ZHU JUN
ZHAO YULIANG
WANG ZHIBIN
ZHAO WENWEN
JIA NAN
YU CHENGLONG

Assignees

北京市农林科学院智能装备技术研究中心

Dates

Publication Date: 20260512
Application Date: 20251226

Claims (10)

1. An unmanned precise feeding method for a farm is characterized by comprising the following steps: acquiring a current farm state, wherein the current farm state comprises a current feed formula of livestock, a current grouping index and individual carbon emission data; mixing materials to produce current feed based on the current feed formula, the current grouping index and the individual carbon emission data, and feeding the current feed to a feeding area where the livestock is located by using a feeding operation system based on historical feeding information acquired in the previous feeding process of the livestock to obtain current feeding information in the current feeding process; Based on the current feeding information and the current feeding requirement of the livestock, adopting a reinforcement learning algorithm, and updating the current feed formula with the aim of minimizing a carbon emission state value function in the current farm state to generate the latest feed formula of the livestock; And updating the current farm state based on the latest feed formula.
2. The unmanned precise feeding method for farms according to claim 1, wherein the feeding operation system comprises a spreader robot; Based on the historical feeding information collected in the previous feeding process of the livestock, the feeding operation system is applied to throw the current feed into a feeding area where the livestock is located for feeding, and the method comprises the following steps: Acquiring colony house information of the livestock; And controlling the material spreading robot to spread the current feed onto the feeding area based on the colony house information and the historical feeding information.
3. The unmanned precise feeding method for farms according to claim 2, wherein the feeding operation system further comprises a pushing robot, the pushing robot is loaded with a visual sensing unit, the controlling the feeding robot to spread the current feed to the feeding area based on the housing information and the historical feeding information further comprises: controlling the pushing robot to push materials at regular time so as to concentrate the feed on the feeding area; The obtaining the current feeding information in the current feeding process comprises the following steps: In the pushing process of the pushing robot, controlling the visual sensing unit to collect images of a feeding area; And receiving the feeding area image, and identifying the feeding area image to obtain the current feeding information.
4. The unmanned precise feeding method for farms according to claim 3, wherein the feeding operation system further comprises a feeding robot, and the current feeding information comprises feeding positions of the livestock and the feed remaining amount of the feeding area; the control of the pushing robot pushes the feed at regular time so as to concentrate the feed on the feeding area, and then the control further comprises: and controlling the feeding robot to feed based on the feeding position and the residual feed amount.
5. The unmanned precise feeding method for farms according to claim 2, wherein the feeding operation system further comprises a receiving robot; The control of the spreader robot to spread the current feed to the feeding area based on the housing information and the historical feeding information, further comprises: After a preset time period, controlling the material collecting robot to recover the residual feed on the feeding area, and collecting residual feed information; and receiving the residual feed information, and determining the current feeding information and the next feeding amount based on the residual feed information.
6. The unmanned precise feeding method for farms according to any one of claims 1 to 5, wherein the mixing production of the current feed based on the current feed formulation, the current grouping index and the individual carbon emission data comprises: determining the required amount and the water required amount of each feed component based on the current feed formula, the current grouping index and the individual carbon emission data; controlling each bin in a mixing operation system to provide corresponding weight of each feed component for mixing equipment based on the required quantity, and controlling a water valve to supply water for the mixing equipment based on the required quantity of water; based on detection equipment installed on the mixing equipment, carrying out online real-time detection and evaluation decision on uniformity, granularity and freshness of the mixture in the mixing equipment; and if the uniformity, granularity and freshness all reach preset values, controlling the mixing equipment to stop to obtain the current feed.
7. The unmanned precise feeding method for farms according to claim 6, further comprising: receiving first operation information of the mixing operation system and second operation information fed back by the feeding operation system; and optimizing the mixing operation of the mixing operation system in the next mixing process based on the first operation information, and optimizing the feeding operation of the feeding operation system in the next feeding process based on the second operation information.
8. Unmanned accurate feeding device in plant, its characterized in that includes: The system comprises a state acquisition module, a state analysis module and a control module, wherein the state acquisition module is used for acquiring the current state of a farm, and the current state of the farm comprises the current feed formula of livestock, the current grouping index and individual carbon emission data; the feeding operation module is used for mixing materials to produce current feed based on the current feed formula, the current grouping index and the individual carbon emission data, and feeding the current feed to a feeding area where the livestock is positioned by applying a feeding operation system based on historical feeding information acquired in the previous feeding process of the livestock to acquire current feeding information in the current feeding process; The formula updating module is used for updating the current feed formula by adopting a reinforcement learning algorithm based on the current feeding information and the current feeding requirement of the livestock and aiming at minimizing a carbon emission state value function in the current farm state, and adjusting and generating the latest feed formula of the livestock; and the state updating module is used for updating the current farm state based on the latest feed formula.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor, when executing the computer program, implements the unmanned precise feeding method of a farm as claimed in any of claims 1 to 7.
10. A computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the unmanned precision feeding method of a farm as claimed in any of claims 1-7.

Description

Unmanned accurate feeding method, unmanned accurate feeding device, unmanned accurate feeding system, unmanned accurate feeding electronic equipment and unmanned accurate feeding storage medium Technical Field The invention relates to the technical field of livestock feeding, in particular to an unmanned accurate feeding method, device and system for a farm, electronic equipment and a storage medium. Background Traditional livestock breeding relies on artificial experience for feeding management, and has the problems of feed waste, unbalanced nutrition, high carbon emission (such as methane emission) and the like. For example, when a theoretical feeding formula designed by a nutritionist is processed and stirred by a total mixed ration (Total Mixed Ration, TMR) device, the actual daily ration is possibly deviated from the theoretical formula due to the problems of feeding precision, mixing uniformity and the like, and finally, the actual ingested nutrient components are deviated from the fed daily ration again due to the picky and eating behaviors of animal individuals, ingestion competition and the like, and the deviation further aggravates unbalanced nutrition and resource waste. Extensive feeding results in low resource utilization, overfeeding exacerbates environmental stress (such as nitrogen and phosphorus pollution), and insufficient feeding affects animal health and productivity. . Disclosure of Invention The invention provides an unmanned accurate feeding method, device and system for a farm, electronic equipment and a storage medium, which are used for solving the defects in the related technology. The invention provides an unmanned precise feeding method for a farm, which comprises the following steps: acquiring a current farm state, wherein the current farm state comprises a current feed formula of livestock, a current grouping index and individual carbon emission data; mixing materials to produce current feed based on the current feed formula, the current grouping index and the individual carbon emission data, and feeding the current feed to a feeding area where the livestock is located by using a feeding operation system based on historical feeding information acquired in the previous feeding process of the livestock to obtain current feeding information in the current feeding process; Based on the current feeding information and the current feeding requirement of the livestock, adopting a reinforcement learning algorithm, and updating the current feed formula with the aim of minimizing a carbon emission state value function in the current farm state to generate the latest feed formula of the livestock; And updating the current farm state based on the latest feed formula. According to the unmanned accurate feeding method for the farm, the feeding operation system comprises a spreader robot; Based on the historical feeding information collected in the previous feeding process of the livestock, the feeding operation system is applied to throw the current feed into a feeding area where the livestock is located for feeding, and the method comprises the following steps: Acquiring colony house information of the livestock; And controlling the material spreading robot to spread the current feed onto the feeding area based on the colony house information and the historical feeding information. According to the unmanned precise feeding method for the farm, the feeding operation system further comprises a pushing robot, and the pushing robot is loaded with a visual sensing unit; The control of the spreader robot to spread the current feed to the feeding area based on the housing information and the historical feeding information, further comprises: controlling the pushing robot to push materials at regular time so as to concentrate the feed on the feeding area; The obtaining the current feeding information in the current feeding process comprises the following steps: In the pushing process of the pushing robot, controlling the visual sensing unit to collect images of a feeding area; And receiving the feeding area image, and identifying the feeding area image to obtain the current feeding information. According to the unmanned accurate feeding method for the farm, the feeding operation system further comprises a feeding robot, and the current feeding information comprises the feeding position of the livestock and the feed remaining amount of the feeding area; the control of the pushing robot pushes the feed at regular time so as to concentrate the feed on the feeding area, and then the control further comprises: and controlling the feeding robot to feed based on the feeding position and the residual feed amount. According to the unmanned accurate feeding method for the farm, the feeding operation system further comprises a material collecting robot; The control of the spreader robot to spread the current feed to the feeding area based on the housing information and the historical feeding information, fur