Prediction of key indicators of utility boiler based on multi-task uncertainty loss

WANG Yufei; LI Nan; XIE Gang; ZHANG Xiaohong; NIE Xiaoyin; ZHOU Yong

doi:10.19666/j.rlfd.202408222

English Version Reader LoginJournal Login

您当前的位置：

首页 >

文章列表页 >

Prediction of key indicators of utility boiler based on multi-task uncertainty loss

更新时间：2025-09-26

- Prediction of key indicators of utility boiler based on multi-task uncertainty loss
- Thermal Power Generation Vol. 54, Issue 5, Pages: 132-139(2025)
- 作者机构：
  
  1. 太原科技大学先进控制与工业智能山西省重点实验室
  2. 山西华光发电有限责任公司
- 作者简介：
- 基金信息：
- DOI：10.19666/j.rlfd.202408222
  CLC：
- Published：2025
- 稿件说明：
移动端阅览
WANG Yufei, LI Nan, XIE Gang, et al. Prediction of key indicators of utility boiler based on multi-task uncertainty loss[J]. Thermal power generation, 2025, 54(5): 132-139.
DOI：

WANG Yufei, LI Nan, XIE Gang, et al. Prediction of key indicators of utility boiler based on multi-task uncertainty loss[J]. Thermal power generation, 2025, 54(5): 132-139. DOI： 10.19666/j.rlfd.202408222.

摘要

随着电站锅炉灵活运行需求的增加，频繁的变负荷运行导致污染物含量和烟气参数大范围波动，对单一污染物或烟气参数等关键指标进行建模已无法满足电厂实际需求，因此需要考虑多种关键指标的耦合性进行协同预测建模。基于燃煤电厂的历史运行数据，通过核函数映射进行特征提取，构建硬参数共享结构的长短时记忆神经网络进行多任务预测建模，利用不确定性损失的方法优化预测模型。实验结果表明，所提出的预测模型在变负荷工况下表现出较高的预测精度，对于所涉及的关键指标空气预热器出口烟气含氧量、烟气温度、炉膛出口NO_x质量浓度的预测，均方根误差分别降低了25.5%、41.8%和4.7%。所提方法能够在变负荷工况下对电站锅炉多个关键指标进行预测，能够辅助电厂实现污染控制和燃烧效率优化，可为电厂智能化运行提供技术支持。

Abstract

With the increasing demand for flexible operation of power plant boilers

frequent variable-load operation leads to a wide range of fluctuations in pollutant concentrations and flue gas parameters. Modeling of key indicators such as single pollutant or flue gas parameter can no longer meet the actual demand

so it is necessary to consider the coupling of multiple key indicators for synergistic predictive modeling. Based on the historical operation data of coal-fired power plants

feature extraction is performed through kernel function mapping

and a long short-term memory neural network with a hard parameter sharing structure is constructed for multi task prediction modeling. The prediction model is optimized using uncertainty loss methods. The experimental results show that

the proposed prediction model exhibits high prediction accuracy under variable load conditions

and the prediction errors for the key metrics involved in this study are reduced by 25.5%

41.8%and 4.7%

respectively. The proposed method is capable of predicting several key indicators of utility boilers under variable load conditions

which can assist power plants to achieve pollution control and optimize the thermal efficiency of combustion

and provide technical support for intelligent operation of power plants.

关键词

Keywords

references

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Multi-parameter collaborative prediction model of boilers based on long-short-term memory neural network

NO_x emission prediction for coal-fired boilers based on RBF Neural network and BP Neural network

Prediction model for NO_x emissions from thermal power plants based on long-short-term memory neural network

Prediction model of main steam temperature of coal-fired boiler considering time delay characteristic optimization

Advancements in the application of machine learning in circulating fluidized bed boiler technology

Related Author

JIN Zhiyuan

LI Shengnan

TAN Peng

YAN Xingchu

LIU Yunong

ZHANG Cheng

CHEN Gang

YU Tingfang

Related Institution

State Key Laboratory of Coal Combustion

Huazhong University of Science and Technology

School of Mechanical & Electrical Engineering

School of Control and Computer Engineering

State Grid Hebei Energy Technology Service Co.Ltd.

AI问答

⁰