考虑属性重要度的空冷系统背压特性模型.pdfVIP

第38 卷 第5 期 中 国 电 机 工 程 学 报 Vol.38 No.5 Mar. 5, 2018 1518 2018 年3 月5 日 Proceedings of the CSEE ©2018 Chin.Soc.for Elec.Eng. DOI ：10.13334/j.0258-8013.pcsee.170408 文章编号：0258-8013 (2018) 05-1518-07 中图分类号：TK 124 考虑属性重要度的空冷系统背压特性模型 李晓恩，王宁玲，张雨檬，杨志平，杨勇平 (华北电力大学国家火力发电工程技术研究中心，北京市 昌平区 102206) Back Pressure Model of the Direct Air-cooling System Considering Attribute Significance LI Xiaoen, WANG Ningling, ZHANG Yumeng, YANG Zhiping, YANG Yongping (National Thermal Power Engineering Technology Research Center, North China Electric Power University, Changping District, Beijing 102206, China) ABSTRACT: The direct air-cooling condenser is more 精度均较高，均方根误差分别为 0.70、0.76kPa 。同时，结 sensitive to the changing of various boundary conditions than 果表明，加权过程可在不影响模型精度的情况下，有效降低 the traditional water-cooling system. By considering the 模型计算量，提高模型泛化能力。 significance of each attributes, a back pressure model was built 关键词：直接空冷系统；背压；属性重要度；加权支持向量 with the practical operating data. The historical data in 回归；模糊粗糙集 one-month operation was used to calculate the weight of each 0 引言 condition attributes based on fuzzy rough sets (FRS). The weighted support vector regression (SVR) method was 燃煤电站在我国能源结构中占据主导地位。传 introduced to build the regression model. The result shows that 统燃煤电厂对水资源需求较高，富煤少水的西北地 the condition attributes, consisting of mass flowrate and 区则广泛采用空气作为排汽冷却介质。空气冷却系 pressure of main steam, ambient temperature, frequency of air fan, wind speed and atmospheric pressure, well reflect the back 统的背压更易受边界条件影响，包括可控的设备运 pressure with a RMSE of 0.70kP