基于LSTM-SAFCN模型的生物质锅炉NOx排放浓度预测
LSTM-SAFCN model based NOx emission prediction for biomass boilers
何德峰 1刘明裕 1孙芷菲 1王秀丽 1李廉明2
作者信息
- 1. 浙江工业大学信息工程学院 杭州 310023
- 2. 嘉兴新嘉爱斯热电有限公司 嘉兴 314016
- 折叠
摘要
针对生物质锅炉燃烧过程的动态特性,提出一种改进的长短期记忆-自注意力机制全卷积神经网络(LSTM-SAFCN)模型用于预测NOx排放浓度.首先利用完全自适应噪声集合经验模态分解法(CEEMDAN)对数据进行预处理,消除数据噪声对NOx排放浓度预测的影响;其次融合自注意力机制与长短时记忆-全卷积神经网络(LSTM-FCN)进行特征提取与预测建模,该拓展方法能够同时兼顾时间序列数据的局部细节与长期趋势特征;最后,利用生物质热电联产系统的实际运行数据验证了所提算法的有效性.
Abstract
In view of the dynamic characteristics of the biomass boiler combustion process,this paper proposes a long short-term memory-self attention fully convolutional network(LSTM-SAFCN)to predict NOx emission.Firstly,a complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN)is applied to preprocess the noise existing in input data.Secondly,the long short-term memory fully convolutional network(LSTM-FCN)is combined with self-attention method for feature extraction and prediction modeling,which takes both the local de-tails of series data and long-term prediction tendency into account.Finally,the effectiveness of the proposed algo-rithm is verified on a biomass cogeneration system.
关键词
生物质锅炉/NOx排放浓度预测/经验模态分解/长短时记忆-全卷积神经网络(LSTM-FCN)/自注意力机制Key words
biomass boiler/NOx emission prediction/empirical mode decomposition/long short-term memo-ry fully convolutional network(LSTM-FCN)/self-attention mechanism引用本文复制引用
基金项目
浙江省重点研发计划(2021C03164)
出版年
2024
国家科技期刊平台
NSTL
万方数据