基于多尺度特征Informer模型的受热面积灰预测研究  

Research on Prediction of Heating Surface Ash Accumulation Using Informer Model with Multi-scale Features

作　　者：王鲁君 孙永华 刘洪涛 于秋红 郝浚杰 WANG Lujun;SUN Yonghua;LIU Hongtao;YU Qiuhong;HAO Junjie(Guoneng Zhejiang Ninghai Power Generation Co.,Ltd.,Ningbo 315612,China;Shandong Luruan Digital Technology Co.,Ltd.,Jinan 250001,China)

机构地区：[1]国能浙江宁海发电有限公司,浙江宁波315612 [2]山东鲁软数字科技有限公司,山东济南250001

出　　处：《山东电力高等专科学校学报》2024年第3期35-40,共6页Journal of Shandong Electric Power College

摘　　要：针对电厂变负荷工况频发,污染因子波动较大的问题,提出一种融合多尺度特征的Informer预测模型,首先通过小波变换对传感器数据进行去噪预处理,然后对污染因子、机组负荷以及其他相关参数进行建模,预测锅炉的积灰状态。利用某电厂2022年1月至9月间的锅炉相关数据对模型进行训练和验证,结果表明融合多尺度特征的Informer模型的预测误差显著降低,验证了模型的有效性。In allusion to the frequent load variations and severe fluctuations in pollution factors in power plants,a predictive model utilizing the Informer model with multi-scale features is proposed.Base on sensor data denoised by wavelet transformation,pollution factors,unit loads,and other relevant parameters are modelled to forecast boiler fouling status.The model is trained and validated using boiler data of a power plant from January to September in 2022.The results indicate the prediction errors of the Informer model with multi-scale features reduce significantly,validating its effectiveness.

关 键 词：污染因子 Informer预测模型 多尺度特征融合 

分 类 号：TK22[动力工程及工程热物理—动力机械及工程]