Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Gas compressor characteristic prediction method based on Kriging model optimization and neural network

A kriging model and neural network technology, applied in the field of compressor characteristic prediction, to achieve high calculation accuracy

Inactive Publication Date: 2019-08-09
HARBIN ENG UNIV
View PDF6 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Due to various reasons, manufacturers can only provide some component characteristics in the high-speed region, which has become a major obstacle to building a component-based gas turbine 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
  • Gas compressor characteristic prediction method based on Kriging model optimization and neural network
  • Gas compressor characteristic prediction method based on Kriging model optimization and neural network
  • Gas compressor characteristic prediction method based on Kriging model optimization and neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0048] The present invention is described in more detail below in conjunction with accompanying drawing example:

[0049] combine figure 1 , choose two kriging methods, namely the univariate ordinary kriging method and the univariate universal kriging method. According to the drift of the data source, different kriging methods need to be selected. For each Kriging method, 3 semivariograms are used to solve the Kriging equations, 6 sets of weighting coefficients are obtained, and 6 compressor characteristic prediction surfaces are obtained by fitting.

[0050] Using the improved leave-one-out cross-validation method, the original sample data is divided into a training sample set and a test sample set, the model is first trained on the training sample set, and then the prediction error is obtained on the test sample set. The improved leave-one-out cross-validation method means that in order to ensure that each prediction is an interpolation prediction, the data on the speed lin...

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 aims to provide a gas compressor characteristic prediction method based on Kriging model optimization and a neural network. The gas compressor characteristic prediction method mainly comprises the steps that two Kriging methods are selected, three semi-variation functions are selected for each Kriging method, a Kriging equation set is solved to obtain six sets of weighting coefficients, and six prediction curved surfaces are obtained by fitting the weighting coefficients; a prediction error average value of each Kriging method is obtained by adopting an improved reserved cross validation method, and an optimal Kriging method is selected; and the characteristics of the gas compressor are identified by using an optimal Kriging method, and finally he characteristics of the gas compressor under all working conditions are predicted by using a neural network method. The method can predict, encrypt and extrapolate the characteristics of the gas compressor under the condition that the characteristic lines of the gas compressor are incomplete or sparse, and has the advantages of high precision and high calculation speed. The method is universal, and has reference significancefor characteristic prediction of impeller machinery such as an axial-flow turbine and the like.

Description

technical field [0001] The invention relates to a method for predicting characteristics of a compressor. Background technique [0002] Gas turbines often operate under non-design conditions, such as start-up, acceleration and deceleration, shutdown, and changes in environmental conditions. The overall performance of a gas turbine mainly depends on the characteristics of the compressor, combustion chamber, turbine and other components, and the compressor has the most critical influence on the performance of the gas turbine. Having complete characteristic curves of core components is the basis for performance analysis of gas turbines under variable operating conditions. Due to various reasons, manufacturers can only provide some component characteristics in the high-speed region, which has become a major obstacle to building a component-based gas turbine model. Therefore, how to use a small number of characteristic curves to obtain component characteristic curves in other wo...

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
IPC IPC(8): G06F17/50G06N3/08
CPCG06N3/084G06F30/17G06F30/20
Inventor 李淑英季念坤王志涛刘瑞戚万领张君鑫李铁磊于海超刘硕硕高楚铭张靖凯
Owner HARBIN ENG UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com