中国电机工程学报2025，Vol.45Issue(7)：2710-2718,中插22,10.DOI:10.13334/j.0258-8013.pcsee.232281
基于树状结构Parzen估计器优化长短期记忆神经网络的燃煤机组NOx生成浓度预测
Prediction of NOx Generation in Coal-fired Units Based on Tree-structure Parzen Estimator Optimized Long Short-term Memory Neural Network
摘要
Abstract
Developing an efficient prediction model for NOx generation is of great significance for reducing NOx emissions and denitrification costs in coal-fired units.The NOx model is built based on related variables and depends on the design of the model structure,with the parameters of the model structure referred to as hyperparameters.Setting these hyperparameters appropriately can substantially enhance the accuracy and generalization capabilities of NOx prediction models.This paper presents a NOx generation prediction model based on tree-structure parzen estimator optimized long short-term memory neural network(TPE-LSTM).Using historical operational data from a 330 MW coal-fired unit,model structural parameters are combined with time series data window length and the number of principal components of NOx generation related variables,thus creating a new type of hyperparameter.The improved hyperparameters are then optimized to construct NOx generation prediction model based on long short-term memory(LSTM)neural networks.Comparing the proposed hyperparameter optimized NOx prediction model with the unoptimized LSTM model and the typical optimization algorithm particle swarm optimization optimized LSTM(PSO-LSTM)model,the prediction results reveal that the TPE-LSTM prediction model demonstrates superior accuracy and generalization abilities.关键词
燃煤锅炉/NOx生成浓度预测/树状结构Parzen估计器/超参数优化/长短期记忆神经网络Key words
coal-fired boiler/NOx generation/tree-structure parzen estimator/hyperparameters optimized/long short-term memory(LSTM)neural network分类
能源科技引用本文复制引用
陈东升,梁中荣,郑国,何荣强,屈可扬,甘云华..基于树状结构Parzen估计器优化长短期记忆神经网络的燃煤机组NOx生成浓度预测[J].中国电机工程学报,2025,45(7):2710-2718,中插22,10.基金项目
国家自然科学基金项目(52376108) （52376108）
广东省省级科技计划项目(2022A0505050004).Project Supported by National Natural Science Foundation of China(52376108) （2022A0505050004）
Guangdong Science and Technology Plan(2022A0505050004). （2022A0505050004）
