基于不等距灰预测模型的变频调速设备故障预测  

Fault Prediction of Variable Frequency Speed Regulation Based on Non-isometric Grey Forecast Model

作　　者：王裕清[1] 李婧[2] 王新[2] 

机构地区：[1]河南理工大学机械与动力工程学院,河南焦作454001 [2]河南理工大学电气工程学院,河南焦作454001

出　　处：《河南科技大学学报（自然科学版）》2009年第3期43-46,共4页Journal of Henan University of Science And Technology:Natural Science

基　　金：河南省重点科技攻关基金项目(072102240006);河南理工大学博士基金项目(B050801);河南省高校科技创新人才支持计划项目(2008HASTIT022)

摘　　要：灰色预测模型具有要求样本数据少、计算方便、预测精度较高并且可对结果进行校验等优点,因此被广泛应用于设备的故障诊断中。本文针对在数据采集过程中出现的不等距数据无法用一般等距模型来预测的情况,建立了一般累加生成差分灰预测模型UGM(1,1)的改进模型,将其用于变频调速设备的状态监测实践中。结果表明该预测模型能使预测精度提高了10%以上,同时计算量并未明显增加。该方法还可以广泛应用于工业企业、医疗卫生等行业各种设备的故障诊断。Grey forecasting models have some advantages,such as modeling with less data,easy operation,high precision and being able to test results,so it’s widely used in fault diagnosis of equipments.In this paper,an improved un-equal interval grey model of UGM(1,1) is introduced,and used on the condition that monitoring of variable frequency speed regulation equipment to solve the problem of non-equidistant data from the data acquisition process that can not be processed by isometric model to predict the general situation.The results show that the forecast model is simple and reliable.The method can also be widely used in industrial enterprises,health care and other industries of various equipment fault diagnosis.

关 键 词：故障诊断 灰色理论 不等距灰然模型 变频调速 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置] TM401.1[自动化与计算机技术—控制科学与工程]