Data-driven soft sensor modeling method based on feature selection

DONG Ze; JIA Hao; JIANG Wei

doi:10.19666/j.rlfd.201901004

English Version Reader LoginJournal Login

您当前的位置：

首页 >

文章列表页 >

Data-driven soft sensor modeling method based on feature selection

更新时间：2025-09-28

- Data-driven soft sensor modeling method based on feature selection
- Thermal Power Generation Vol. 48, Issue 9, Pages: 83-89(2019)
- 作者机构：
  
  1. 华北电力大学河北省发电过程仿真与优化控制技术创新中心
  2. 华北电力大学控制与计算机工程学院
- 作者简介：
- 基金信息：
- DOI：10.19666/j.rlfd.201901004
  CLC： TP274.2
- Published Online：05 September 2019，
  
  Published：2019
- 稿件说明：
移动端阅览
DONG Ze, JIA Hao, JIANG Wei. Data-driven soft sensor modeling method based on feature selection[J]. Thermal power generation, 2019, 48(9): 83-89.
DOI：

DONG Ze, JIA Hao, JIANG Wei. Data-driven soft sensor modeling method based on feature selection[J]. Thermal power generation, 2019, 48(9): 83-89. DOI： 10.19666/j.rlfd.201901004.

摘要

由于测量环境恶劣

分析仪器成本昂贵

测量延迟大

发电机组一些关键参数难以实现在线测量。本文提出一种基于皮尔森相关系数和最小角回归算法相结合的特征选择方法

并以此为基础建立电厂烟气含氧量软测量模型。首先

根据机理分析确定烟气含氧量模型的初始输入变量

使用皮尔森相关系数对多测点数据进行数据融合;然后

使用最小角回归算法对输入变量进行特征选择

引入高斯过程回归(GPR)模型建立了烟气含氧量的软测量模型;最后将该模型用于某超超临界1 000 MW机组实际稳定工况数据进行仿真。结果表明:该模型预测结果与实测值误差较小

模型精度较高;使用该特征选择方法可以有效减少模型二次变量数量

降低模型输入复杂度

提高计算效率

同时可以改善模型的泛化能力

提高软测量的静态准确度。

Abstract

Due to poor measurement environment

expensive analytical instruments and large measurement delay

some key variables of power generating units are difficult to realize on-line measurement. To solve this problem

a feature selection method based on the Pearson correlation coefficient and the least angle regression algorithm(LARS) is presented. On this basis

a soft sensing model of oxygen content in flue gas of power plants is established. Firstly

the initial variables of oxygen content in flue gas are determined according to the mechanism study

and the Pearson correlation coefficient is used to fuse the data of multiple measurement points. Then

the least angle regression algorithm is used to select the input variables

and the soft sensor model of oxygen content in flue gas is established by introducing the Gauss process regression(GPR) model. Finally

this model is applied to simulate the actual steady condition data of an ultra supercritical 1 000 MW unit. The results show that

the predicted results have little error with the actually measured values

and the precision of the model is higher. Applying this feature selection method can effectively reduce the number of the secondary variables used in modeling

decrease the input complexity of the model and improve the calculation efficiency. Moreover

it can enhance the static accuracy and generalization ability of the flue gas oxygen content model.

关键词

Keywords

references

Views

303

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

CALCULATION OF PULVERIZED COAL CONCENTRATION BASED ON DATA FUSION METHOD OF STATE,SPACE AND DISTANCE

ANALYSIS AND APPLICATION OF UNCERTAINTY IN ANTI-BALANCING THERMAL EFFICIENCY OF BOILERS

GA-SVM-based soft-sensor model for oxygen content in flue gas of utility boilers

Data feature selection for steam turbine based on Copula function and LU decomposition method

Flame stability detection method for co-firing of biomass fuels based on digital image processing

Related Author

PAN Jing-bin2(1.North China Electric Power University

XIE Hui2

TIAN Liang1

CHANG Tai-hua1

LIU Huan-zhang1

LIU Jingyan2

ZHUO Xunjia1

WANG Jiangwei1

Related Institution

Taicang Power Plant

North China Electric Power University

Ningxia Electric Power Company Electric Power Science Research Institute

School of Energy and Power Engineering, Xi'an Jiaotong University

State Grid Beijing Electric Power Company

AI问答

⁰