基于改进支持向量机的风机性能在线监测平台研究

摘要

Monitoring the fan performance and running state in real time is very important for the safety and economy of power plant.Besides, an accurate measurement of fan flow is the key of real time monitoring .Considering the current situation of the lack of accurate methods to measure the flow of large pipe , the mapping principle of steady operating parameters of fan based on dimensionless performance maps is discussed and the flow measurement models of fan based on different modeling intelligence al -gorithms are derived, we find that the accuracy of the modified support vector machine (SVM) model is higher.Finally, the on-line monitoring platform of fan is established by using Visual Basic language mixed with Matlab programming on the basis of im -proved SVM model , which can monitor the flow and performance of fan in different conditions on-line.%风机性能实时监测对电厂安全、经济运行具有重要意义，而对风机流量的准确实时监测是风机性能在线监测的关键。针对现场缺乏大管径流量准确测量手段的现状，本文采用风机无因次性能曲线中性能参数之间稳定的映射规律，并通过不同智能算法建立了风机全程流量测量模型，结果发现改进的支持向量机模型拟合精度更高，效果更好。最后，在改进支持向量机模型基础上，采用Visual Basic语言与Matlab混合编程方法，建立了风机性能在线监测平台，实现了对风机实际工况下的流量、性能的实时监测。