摘要
本文利用某燃煤电厂湿法烟气脱硫系统的数据对比随机森林、极致梯度提升、支持向量回归、深度神经网络、长短时记忆神经网络五种模型的预测效果,结果显示长短时记忆神经网络的效果优于其它四种模型。基于该模型结合多目标粒子群优化算法,进一步对脱硫系统相关参数开展多目标优化,优化结果验证了本文提出的数据驱动的脱硫系统多目标优化模型的有效性。
In this paper,data from the wet flue gas desulfurization system in a coal-fired power plant is used to compare the results show that this prediction model is more precise than static modeling and models that do not use first order differential prediction method.The predictive performances of five models,including random forest,extreme gradient boosting,support vector regression,deep neural network,and long short-term memory neural network.The results show that the long short-term memory neural network performs better than the other four models.Based on this model and multi-objective particle swarm optimization algorithm,multi-objective optimization of desulfurization system parameters is carried out,and the optimization results verify the effectiveness of the proposed data-driven desulfurization system multi-objective optimization model.
作者
马永涛
邢江宽
罗坤
樊建人
MA Yongtao;XING Jiangkuan;LUO Kun;FAN Jianren(State Key Laboratory of Clean Energy Utilization,Zhejiang University,Hangzhou 310058,China)
出处
《能源工程》
2024年第1期1-12,共12页
Energy Engineering
基金
国家杰出青年科学基金资助项目(51925603)。
关键词
数据驱动
机器学习
脱硫系统
多目标粒子群优化
data-driven
machine learning
desulfurization system
multi-objective particle swarm optimization