Combustion optimization for CFB boiler based on least square support vector machine and modified fruit fly optimization algorithm

ZHANG Wenguang; ZHANG Yue; SUN Yazhou; GAO Mingming; LI Baogui

English

您当前的位置：

首页 >

文章列表页 >

Combustion optimization for CFB boiler based on least square support vector machine and modified fruit fly optimization algorithm

Thermal Power Generation Vol. 45, Issue 7, Pages: 44-49(2016)

更新时间：2025-09-28

- Combustion optimization for CFB boiler based on least square support vector machine and modified fruit fly optimization algorithm
- 作者机构：
  
  1. 华北电力大学新能源电力系统国家重点实验室
  2. 中国人民解放军93469部队
- 作者简介：
- 基金信息：
- DOI：
  CLC： TM621.2
- Published：2016
- 稿件说明：
移动端阅览
ZHANG Wenguang, ZHANG Yue, SUN Yazhou, et al. Combustion optimization for CFB boiler based on least square support vector machine and modified fruit fly optimization algorithm[J]. Thermal Power Generation, 2016, 45(7): 44-49.
DOI：

ZHANG Wenguang, ZHANG Yue, SUN Yazhou, et al. Combustion optimization for CFB boiler based on least square support vector machine and modified fruit fly optimization algorithm[J]. Thermal Power Generation, 2016, 45(7): 44-49. DOI：

摘要

针对电厂循环流化床(CFB)锅炉降低污染物排放和提高锅炉燃烧效率的问题

本文首先应用最小二乘支持向量机(LS-SVM)建立了锅炉效率、NO_x和SO_2排放特性的软测量模型

并对比了LS-SVM和BP神经网络模型的性能;然后基于LS-SVM建立的模型

提出了3种优化策略

采用改进果蝇优化算法(MFOA)在一定范围内对CFB锅炉运行工况的可调参数进行优化。结果表明:LS-SVM模型与BP神经网络模型相比

训练时间较短

预测精度较高

泛化能力较强;CFB锅炉效率最多提高了0.61%

NO_x和SO_2排放质量浓度最多降低了7.88%和18.13%。

Abstract

To reduce the pollutant emissions and enhance combustion efficiency of circulating fluidized bed(CFB)boiler

three soft measurement models

the boiler efficiency

the NO_x emission and SO_2 emission

were established based on the least square support vector machine(LSSVM).The performance of the LSSVM was compared with that of the conventional BP neural network.Then

on the basis of the above LSSVM models

three optimization strategies were proposed to optimize the adjustable parameters using the modified fruit fly optimization algorithm(MFOA)within a certain range.The simulation results show that

compared with the conventional BP neural network

the LSSVM model needs less training time and has higher prediction accuracy with better generalization ability.Moreover

the result also proves the effectiveness of the combustion optimization strategies for CFB boilers

the boiler efficiency was increased by0.61% and the NO_xand SO_2 emission were respectively reduced by 7.88% and 18.13% at most.

关键词

Keywords

references

Views

261

下载量

CSCD

Alert me when the article has been cited

提交

Publicity Resources

APPLICATION OF FLUE GAS AUTOMATIC CONTINUOUS EMISSION MONITORING SYSTEM

TEST STUDY ON EMISSION OF NO_x AND SO₂ FROM GASIFIED INCINERATION OF ORGANIC SOLID WASTES

PREDICTION MODEL FOR POLLUTANTS SO₂ AND NO_x FROM LARGE-SCALE COAL-FIRED POWER PLANT BASED ON NEURAL NETWORK

Combustion Optimization for Utility Boilers Based on WLS-SVM-SFS Model

Application of ultra-low emission technology in a 600 MW coal-fired unit

Related Author

DING Hui

KONG Fan-xin

FU Cong-wei

JIA Mei-qin

MA Yan

KONG Guo-qiangHuaju Energy Source Jidong Branch Company

GAO Ning-bo1

LI Ai-min1

Related Institution

Huaju Energy Jidong Branch

Dalian University of Technology

Shenyang Institute of Aeronautics and Astronautics

Zhejiang University Energy Engineering Design and Research Institute

Guodian Dongsheng Thermal Power Plant

AI问答

⁰