Distributed training method of linear dynamic system model

A technology of dynamic system model and training method, applied in character and pattern recognition, instrument, calculation, etc., can solve problems such as time series data link breakage, and achieve the effect of improving computational efficiency, accurate learning results, and high prediction accuracy

Pending Publication Date: 2022-08-05
CHINA UNIV OF PETROLEUM (EAST CHINA)
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Aiming at the deficiencies of the prior art, the present invention provides a distributed training method of a linear dynamic system model, which establishes the dynamic relationship between auxiliary variables and key variables in the form of a linear dynamic system model, and effectively solves the dynamic and large-scale data problems caused by industry. The problem of broken training and time series data chains, and the information contained in labeled samples and unlabeled samples are simultaneously mined through semi-supervised learning, making model training more reliable

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
  • Distributed training method of linear dynamic system model
  • Distributed training method of linear dynamic system model
  • Distributed training method of linear dynamic system model

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0091] The distributed training method for a linear dynamic system model of the present invention will be further elaborated below with reference to specific embodiments. It should be pointed out that the described embodiments are only intended to enhance the understanding of the present invention, and do not limit the present invention in any way.

[0092] The distributed training method of the linear dynamic system model of the present invention, such as figure 1 shown, including the following steps:

[0093] (1) Select the auxiliary variable x∈R associated with the key variable y v , where v represents the number of auxiliary variables;

[0094] This embodiment is based on the low temperature conversion unit (such as figure 2 The 7 variables that have the greatest impact on CO concentration are selected as auxiliary variables, which are the flow rate of the inlet gas (x 1 ), intake air temperature (x 2 ), the top temperature of the reactor (x 3 ), the temperature of ...

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 distributed training method for a linear dynamic system model. The distributed training method comprises the following steps: firstly, dividing large-scale time series data into a plurality of continuous-time data blocks; and then, solving model parameters through an expectation maximization algorithm in a parameter server mode to realize efficient data mining. By applying the method, the training duration of the linear dynamic system model can be obviously shortened, and discontinuous time sequence data can be fully mined, so that the method has remarkable advantages in the aspects of calculation efficiency, prediction precision and model interpretability compared with an existing centralized training method; technical support and guarantee are provided for improving the product quality control quality, reducing the cost, monitoring the process and making decisions.

Description

technical field [0001] The invention belongs to the field of process system soft measurement modeling and application, in particular to a distributed training method for a linear dynamic system model. Background technique [0002] Soft sensing techniques have been widely used to measure difficult-to-measure quality-related variables in process systems, such as various compositional variables, polypropylene melt indices, diesel and gasoline quality indices, and more. Soft sensing technology is essentially a predictive mathematical model that takes easily measurable auxiliary variables such as temperature, flow, pressure and level as input and provides predictions of key quality-related variables. Therefore, online estimation of key variables by soft sensing can be as fast as the sampling rate of auxiliary variables, with essentially no measurement delay compared to offline laboratory analysis. In addition, soft sensing offers huge economic advantages over field measurement a...

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): G06K9/62
CPCG06F18/29G06F18/214
Inventor 邵伟明赵东亚李友高
Owner CHINA UNIV OF PETROLEUM (EAST CHINA)
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap