基于大数据平台的SO_(2)排放GWO-N-BEATS预测算法  

SO_(2) Emission Prediction by GWO-N-BEATS Algorithm based on Big Data Platform

作　　者：曾庆华 冉鹏[2] 董坤 刘旭 ZENG Qing-hua;RAN Peng;DONG Kun;LIU Xu(Hunan Province Key Laboratory of Efficient&Clean Thermal Power Generation Technologies,State Grid Hunan Electric Power Corporation Limited Research Institute,Changsha,China,410007;Hebei Key Laboratory of Low Carbon and High Efficiency Power Generation Technology,North China Electric Power University,Baoding,China,071003;School of Energy Power and Mechanical Engineering,North China Electric Power University,Baoding,China,071003)

机构地区：[1]国网湖南省电力有限公司电力科学研究院,高效清洁发电技术湖南省重点实验室,湖南长沙410007 [2]华北电力大学河北省低碳高效发电技术重点实验室,河北保定071003 [3]华北电力大学能源动力与机械工程学院,河北保定071003

出　　处：《热能动力工程》2024年第3期125-131,共7页Journal of Engineering for Thermal Energy and Power

基　　金：国家自然科学基金资助项目(51506052)。

摘　　要：为了更精确地预测SO_(2)排放质量浓度,解决非线性随机预测问题,提出了一种基于随机森林特征选择的GWO-N-BEATS算法。通过随机森林算法筛选输入参数的特征,使用灰狼优化算法对N-BEATS算法的超参数进行优化;与长短期记忆网络(Long Short-Term Memory, LSTM)、门控循环神经网络(Gated Recurrent Unit, GRU)以及N-BEATS算法对比分析,验证了GWO-N-BEATS算法的有效性。将本算法应用于某大型电网公司大数据平台,探索了复杂智能算法在大数据平台上开展污染物排放预测的可行性。研究结果表明,相较于长短期记忆网络、门控循环神经网络和N-BEATS方法,GWO-N-BEATS算法预测误差更小,其中平均绝对百分比误差MAPE为1.50%,相对均方误差RMSE为0.42,平均绝对误差MAE为0.33,决定系数R^(2)为0.97。To predict SO_(2)emission mass concentration more accurately and to solve the nonlinear stochastic prediction problem,a novel grey wolf optimization(GWO)deep learning architecture N-BEATS algorithm based on random forest feature selection was proposed.The features of the input parameters were screened by the random forest algorithm,and the hyperparameters of the N-BEATS model were optimized using the GWO algorithm;the effectiveness of the proposed algorithm was verified by comparing it with long short-term memory network(LSTM),gated recurrent unit(GRU)and N-BEATS.The algorithm was applied to a large power grid company's big data platform to explore the feasibility of complex intelligent algorithms to carry out pollutant emission prediction on a big data plaform.The results show that the GWO-N-BEATS algorithm has less error compared to LSTM,GRU and N-BEATS methods,where MAPE is 1.50%,RMSE is 0.42,MAE is 0.33,and R^(2)is 0.97.

关 键 词：随机森林 特征选择 灰狼优化算法 大数据平台 N-BEATS SO2预测 

分 类 号：TK284[动力工程及工程热物理—动力机械及工程]