Pig house environment control method and device based on digital twinning and reinforcement learning and medium

By constructing a digital twin model of a pigsty and using reinforcement learning methods, the problems of real-time and accuracy in pigsty environmental control were solved, achieving efficient and dynamic regulation of the pigsty environment. This avoided the economic losses caused by training in a real pigsty and improved environmental suitability and energy conservation and emission reduction effects.

CN122195178APending Publication Date: 2026-06-12ZHEJIANG UNIV
0 Cites 0 Cited by

Patent Information

Application Number
CN202610338938.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-03-19
Publication Date
2026-06-12

AI Technical Summary

Technical Problem

Existing methods for controlling the environment in pigsties are insufficient for achieving real-time, dynamic, and precise adjustments, resulting in deficiencies in environmental suitability and energy conservation and emission reduction.

Method used

A digital twin model of the pigsty environment is constructed. An improved LSTM model is used to train the sample pigsty state sequences. A reinforcement learning model is combined to generate and update virtual environment control commands. Through training the digital twin model and the experience playback pool, the collection of real pigsty actions is reduced, and the accuracy and real-time performance of environmental control are improved.

Benefits of technology

It enables real-time, dynamic, and efficient control of the pig house environment, avoiding the economic losses caused by training in real pig houses, improving the accuracy and real-time nature of environmental control, and enhancing the ability to coordinate control of multiple environmental factors.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses a pigsty environment control method and device based on digital twinning and reinforcement learning and a medium, and relates to the technical field of large-scale breeding, and the method comprises the steps: extracting a plurality of data subsets from a data set; for each data subset, inputting a sample hog house state at the last moment in the data subset into a network of a reinforcement learning model to obtain a virtual environment control instruction and update an equipment state, and inputting the updated data subset into a digital twinborn model to obtain an updated hog house internal environment state; updating an experience playback pool based on the updated data subsets and the updated environment state in the hog house, training the network until all the data subsets are traversed, and determining the trained network as an environment control model; and inputting the pig house state at the current moment into the environment control model to obtain an environment control instruction at the current moment so as to control the pig house environment. According to the invention, the accuracy and real-time performance of pig house environment regulation and control can be improved.
Need to check novelty before this filing date? Find Prior Art