表面粗糙度在线识别的理论与实践  

Theory and practice of on-line identification of roughness of a work-piece

作　　者：王爱囡[1] 

机构地区：[1]合肥工业大学理学院,安徽合肥230009

出　　处：《合肥工业大学学报（自然科学版）》2004年第4期425-429,共5页Journal of Hefei University of Technology：Natural Science

摘　　要：工件表面粗糙度,反映了加工设备的动态特性。通过观察加工中工件表面粗糙度的变化,可以用作在线识别设备工况的一种手段。根据时间序列分析理论所建立的系统的等价模型,是建立在输出等价原则上的,可以不管输入和输出的因果关系。所建模型能随时间的推移具有某些统计特性,能用于在线识别。对于自回归AR(n)时序模型的建模过程、模型阶次的确定、评定所建模型的适用性及被识别系统工况的特征参数均进行了讨论,试验结果证明,所建模型能快速、准确判别设备工况变化,达到了在线识别的目的。The dynamic characteristics of a machining device is reflected by the surface roughness of a work-piece,so observing the surface roughness's changing of a processing work-piece is a way of on-line identifying the operating conditions of the device. Based on the theory of time-series analysis, the equivalent model of the identification system is built on the principle of output being equivalent without considering the causality between input and output. The built model shows certain statistical regularity as time goes on, hence it can be used for on-line identification. Also discussed are the modeling process of autoregressive AR(n) time-series model,determining of the model order, evaluating of the applicability of the built model, and the characteristic parameters of the identified system's operating conditions. Experimental research demonstrates that the built model is quick and accurate in ascertaining the changing of the device's operating conditions,and the on-line identification object is attained.

关 键 词：在线识别 自回归模型 时间序列 建模 

分 类 号：O433.54[机械工程—光学工程]