一类基于FARIMA过程的电梯导轨振动模型  

A kind of vibration model of elevator's guide based on FARIMA process

作　　者：安德洪[1] 丁春蕾[2] 刘嘉焜[2] 许树荆 

机构地区：[1]天津大学管理学院,天津300072 [2]天津大学理学院,天津300072 [3]天津市华苑电力有限公司,天津300384

出　　处：《机械设计》2004年第4期31-33,41,共4页Journal of Machine Design

基　　金：南开大学天津大学刘徽应用数学中心资助项目 (T0 8) ;天津市华苑电力有限公司资助项目

摘　　要：研究电梯运行中导轨的随机振动时 ,将电梯运行中测得的导轨间距离DGB看作一时间序列 ,发现具有长相关性。常用的整数自回归模型、分数噪声模型都只能片面地描述该类数据的短相关性或长相关性。给出了利用FARIMA(自回归分数整合滑动平均模型 )拟合DGB的方法 ,该模型可同时刻画实测数据DGB的长相关和短相关特性 ,并通过对实测数据的实验 ,证明了模型的优效性。While the stochastic vibration of guide during the movement of elevator is being studied, the data of distance between guides (DGB) measured when the elevator is in motion is looked upon as a time series, and it has been found that the data is possessing a long dependence. The commonly used integer autoregressive model and fractional noise model could only describe one-sidedly the short dependence or long dependence of this kind of data. A method utilizing the fractional autoregressive integration moving (sliding) average (FARIMA) model to match the DGB was presented. This model could depict simultaneously both the long and short dependences of the actually measured data (DGB), and the property of optimization and effectiveness of this model was verified by means of an experiment on the measured data.

关 键 词：电梯 导轨 随机振动 长相关性 整数自回归模型 自回归分数整合滑动平均模型 FARIMA 

分 类 号：TU857[建筑科学] TH113.1[机械工程—机械设计及理论]