摘要
实现磨煤机预知维修对电厂经济、安全运行具有重要意义,而磨辊磨损程度的判断是重要一环,但其难以直接测量是当前最大难题。电厂海量历史数据中蕴含了极有价值的信息,在机理分析和数据分析的基础上,利用小波变换对磨煤机单耗进行多尺度分析,从中提取出反映磨辊磨损程度的趋势分量,实验验证了该方法的有效性。
Achieving of predictive maintenance of the mill is very useful for the power plants＇ economic and safe operation.The judgment of wear characteristics is an important part and great challenge,which is difficult to measure directly.The mass historical data contain lots of valuable information.Based on the mechanism analysis and data analysis,trend component of wear characteristics was extracted from specific energy using wavelet transform and multi-scale analysis.Effectiveness was proved by the experiment.
出处
《华北电力大学学报（自然科学版）》
CAS
北大核心
2011年第2期37-42,共6页
Journal of North China Electric Power University：Natural Science Edition
基金
国家自然科学基金重点资助项目(51036002)
关键词
预知维修
磨辊磨损
多尺度分析
小波变换
趋势分量
predictive maintenance
wear characteristics
multi-scale analysis
wavelet transform
trend component
