基于ARMA模型和趋势分析的铁谱定量参数分析与预测.pdfVIP

学兔兔 化 工 进 展 2014年第33卷增刊1 CHEMICAL INDUSTRY AND ENGINEERING PROGRESS # 。 p ； {研究开发2 — 1 — 1卢 基于ARMA模型和趋势分析的铁谱定量参数分析与预测 曹海峰 ，韩念琛 ，宣征南 ( 太原理工大学化学化工学院，山西 太原 030024； 广东石油化工学院教务处，广东茂名 525000) 摘 要：在自行搭建的齿轮箱磨损实验平台的基础上，获取了38组代表齿轮磨损程度的铁谱直读读数，并以此为 基础介绍了处理该数据的ARMA模型理论及趋势分析方法。通过对二者原理及优缺点的分析提出了将 ARMA模 型理论和趋势分析方法组合的数据处理方法。利用该方法对获取的数据进行分析：对前35组数据进行时序建模 并获取后3组数据的预测值，并利用趋势分析法将获取的监测值与预测值进行对比比较。结果表明：该组合分析 方法在对设备磨损值的预测及磨损状态的判定和预测方面具有较好的精度，可作为一种新的数据处理方式运用到 实际的设备故障监测工作中。 关键词：趋势分析；ARMA模型；铁谱定量分析；预测；参数估值 Ferrographic quantitative parameter analysis and prediction based on ARMA model and trend analysis CAO Haifeng ，HAN Nianshen ，XUAN Zhengnan ( School of Chemistry and Chemical Engineering，Taiyuan University of Technology，Taiyuan 030024，Shanxi，China； 。Office of Teaching Affairs，Guangdong University of Petrochemical Technology，Maoming 525000，Guangdong，China) Abstract：On the wear experiment platform for gear box，38 groups of ferrographic quantitative readings which could characterize the wear degree of the gear were obtained，and the ARMA model theory and the trend analysis method often used to process the ferrographic quantitative readings were introduced． Through the analysis of the principles，advantages and disadvantages of the ARMA model theory and the trend analysis method，a combination of the two methods was proposed for data processing．Time—series modeling was made with the first 35 groups of data and the last 3 groups were predicted by using the proposed method．Trend analysis was made to compare the monitored values with the predicted values．The proposed combination for failure prediction of equipment had good precision