华电技术 ›› 2015, Vol. 37 ›› Issue (3): 10-13.

• 研究与开发 • 上一篇    下一篇

基于熵聚类的改进算法在电厂脱硫系统中的应用

  

  1. (1.国电南瑞科技股份有限公司,南京〓210000; 2.江苏智方建设工程有限公司,南京〓210000)
  • 出版日期:2015-03-25 发布日期:2015-05-14

Application of improved entropy fuzzy clustering in power plant desulfurization system

  1. (1.NARI Technology Corporation Limited, Nanjing 210000,China; 2.Jiangsu Zhifang Construction Engineering Company Limited, Nanjing 210000,China)
  • Online:2015-03-25 Published:2015-05-14

摘要: 摘要:石灰石-石膏湿法脱硫技术采用石灰或石灰石作为脱硫剂,影响其脱硫性能的主要因素有石灰石原料、机组运行参数和可调参数等方面。在参数约减后,将浆液密度、浆液的pH值、钙硫比(Ca/S)可调参数控制在其最优值范围内,得到影响脱硫效率的因素为机组负荷、入口SO2质量浓度和浆液泵运行台数。对基本的熵模糊聚类(EFC)进行相关改进,提出阀值β变尺度EFC算法,并验证了算法的准确性。利用改进算法在各典型负荷邻域区间内对实际运行数据分别进行多参量同步聚类,挖掘出不同负荷所对应的入口SO2质量浓度、浆液泵开启台数与脱硫效率之间的关系,并提出浆液循环泵优化调度方案。

关键词: 关键词:湿法脱硫, 脱硫效率, 阀值β变尺度EFC聚类, 浆液循环泵

Abstract: Abstract:Limestonegypsum wet FGD technology, lime or limestone as the desulfurizer, is mainly affected by raw limestone, unit’s running parameters and adjustable parameters. After reduction of the parameters, constrain adjustable parameters as slurry density, the pH value of the slurry and the calcium to sulfur ratio (Ca/S) within the optimal range of values, and the factors affecting the desulfurization efficiency are concluded as unit load, inlet SO2 concentration and slurry pump operating units. Threshold variable metric entropy fuzzy clustering (EFC) algorithm is proposed on base of the basic EFC, and its reliability is verified. The proposed algorithm is used for multiparameters’clustering on actual data within several typical load neighborhoods, to dig out the relationship between the inlet SO2 concentration, slurry pumps and the desulfurization efficiency according to different loads, furthermore, propose slurry circulating pump optimization scheduling scheme.

Key words: Keywords:wet FGD technology;desulfurization efficiency;threshold variable metric EFC, slurry circulating pump