Check patentability & draft patents in minutes with Patsnap Eureka AI!

Large-time-lag coupling system dynamic modeling method based on deep reinforcement learning

A technology of reinforcement learning and coupled systems, which is applied in the field of dynamic modeling of large time-delay coupled systems based on deep reinforcement learning, can solve problems such as difficult to establish accurate mathematical models, reduce intelligent decision-making ability, shorten parameter search time, Accurately built effects

Inactive Publication Date: 2021-04-06
CHINA SOUTH IND GRP AUTOMATION RES INST +1
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0002] The large time-delay coupled system has the characteristics of multi-physics field parameter coupling and obvious time-delay, so it is difficult to establish an accurate mathematical model

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Large-time-lag coupling system dynamic modeling method based on deep reinforcement learning
  • Large-time-lag coupling system dynamic modeling method based on deep reinforcement learning
  • Large-time-lag coupling system dynamic modeling method based on deep reinforcement learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0025] Such as figure 1 As shown, the feature extraction of the data is performed using a recurrent-convolutional neural network, the RNN is used to extract the temporal features of the data, and the CNN is used to extract the spatial local features of the data. refer to figure 1 , where P={p 1, p 2, ...,p i ,…,p n} represents the recorded data of the monitored n state parameters p at k consecutive time points, which are used as input data and input into the network from the left and right channels respectively, where the left channel is a CNN network, which is used to extract the spatial correlation of different data Features, the last layer of the CNN network is activated by n neurons to obtain a 1×n-dimensional feature vector in, is the i-th feature extracted from the input data by the CNN network, that is, the output of the i-th neuron in the last layer. The right channel is a RNN network, and the input data is the transposition P of the input data of the left cha...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a large-time-lag coupling system dynamic modeling method based on deep reinforcement learning, and the method comprises the steps: firstly carrying out the dynamic coarse positioning of a model parameter based on the fusion of historical information and current information, so as to reduce the time-lag impact and shorten the parameter searching time; secondly, adopting a method for dynamically and accurately adjusting model parameters based on a deep reinforcement learning network to meet the requirement of improving the model precision. According to the method, the recurrent-convolutional neural network and deep reinforcement learning are combined, decoupling, time-lag influence reduction and intelligent decision-making capabilities are achieved, the model can be constructed more quickly and accurately, the problem of model construction of a large time-lag coupling system can be well solved, and the method can be widely applied to the fields of molding control and industrial production and manufacturing.

Description

technical field [0001] The invention belongs to the technical field of dynamic modeling of large time-delay coupled systems, and more specifically relates to a dynamic modeling method of large time-delay coupled systems based on deep reinforcement learning. Background technique [0002] The large time-delay coupling system has the characteristics of multi-physics field parameter coupling and obvious time-delay, so it is difficult to establish an accurate mathematical model. However, most systems in industrial production are coupled systems with large time-delays, so it is extremely important to study the dynamic modeling methods of coupled systems with large time-delays. [0003] As a widely used technology in the field of artificial intelligence today, deep learning combines low-level features to form more abstract high-level representation attribute categories or features to discover the distributed feature representation of data, which has great advantages in data process...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06F30/27G06N3/04G06N3/08
CPCG06F30/27G06N3/08G06N3/045
Inventor 高丰黄求安陈勇史慧芳伍凌川潘尧杰刘越智
Owner CHINA SOUTH IND GRP AUTOMATION RES INST
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com