基于声学测温与最小二乘支持向量机的锅炉炉膛灰污监测方法  

Method of Furnace Fouling Monitoring Based on Acoustic Pyrometry and Least Squares Support Vector Machine

作　　者：马美倩[1] 吕伟为[1] 沈国清[1] 安连锁[1] 张世平[1] 

机构地区：[1]华北电力大学电站设备状态监测与控制教育部重点实验室,北京102206

出　　处：《电力科学与工程》2012年第7期46-50,70,共6页Electric Power Science and Engineering

基　　金：国家自然科学基金项目(50976034);中央高校基本科研业务费专项资金资助(09QG43)

摘　　要：研究电站锅炉炉膛灰污监测问题,提出了基于声学测温和最小二乘支持向量机的电站锅炉炉膛灰污监测方法。该方法采用声学测温装置获得实际运行状态下锅炉炉膛出口烟温,用最小二乘支持向量机获得实际运行状态下锅炉炉膛清洁时的潜在炉膛出口烟温,运用上述两参数定义灰污特征参数来表征锅炉炉膛整体灰污状况。建立了监测模型,从电厂采集数据对模型进行了训练和验证,并对获得的灰污特征参数进行了分析,结果表明:基于声学测温和最小二乘支持向量机的锅炉炉膛灰污监测方法可以较准确地实现电站锅炉炉膛的灰污监测,为炉膛的吹灰优化打下了良好的基础。This paper studies power station boiler furnace pollution monitoring,puts forward the method of boiler hearth fouling monitoring based on the acoustic pyrometry and least squares support vector machine,in this method,we get the actual boiler furnace outlet gas temperature with acoustic pyrometry device and gain the potential cleaning outlet gas temperature under the practical operating condition with the least squares support vector machine,then using the two parameters define fouling characteristic parameter to show the conditions of boiler furnace fouling.Collect data from the power plant to train and certification the model,analyze the fouling characteristic parameter,results show that method of boiler hearth fouling monitoring based on the acoustic pyrometry and least squares support vector machine can more accurately realize boiler furnace fouling monitoring and it is a good foundation for furnace blowing optimization.

关 键 词：炉膛出口烟温 声学测温 最小二乘支持向量机 灰污特征参数 灰污监测 

分 类 号：TM621.2[电气工程—电力系统及自动化]