摘要
随着系统设备和功能的日益复杂化,复杂机械系统和过程的监测诊断问题是一个广泛的研究课题,该文针对复杂机械系统故障现象的特点,分析了自回归(AR)模型分析技术在复杂机械系统监测诊断中的应用。研究了AR模型的建模过程及其参数估计,讨论了典型模型的定阶准则,建立了模型识别信息距离判别函数,并就非平稳时间序列异常值分析、非平稳残差修正方法等问题进行了探讨。
With increasing complexity of equipment and its function, monitoring and diagnosis of complex mechanical system and process is a expensive subject. According to the characteristics of complex systems, this paper analyzes the application of auto-regression model in monitoring and diagnosis of mechanical system. AR model is established and the estimation of AR model parameter is studied. Phase criterion of model is discussed. The information distance discriminating function for pattern recognition is established. Furthermore a non-stationary time series model abnormal value and non-stationary residual revise method are stated.
出处
《组合机床与自动化加工技术》
2005年第10期45-47,共3页
Modular Machine Tool & Automatic Manufacturing Technique
关键词
监测诊断
模式识别
时间序列
自回归模型
monitoring and diagnosis
pattem recognition
time series analysis
AR model
