小型无人直升机辨识建模试验信号的预处理  

The Preprocessing for Robot Helicopter's Model Identification Signals

作　　者：赵志刚[1] 吕恬生 

机构地区：[1]兰州交通大学机电工程学院,甘肃兰州730070 [2]上海交通大学机械与动力工程学院,上海200030

出　　处：《兰州交通大学学报》2010年第4期40-43,48,共5页Journal of Lanzhou Jiaotong University

基　　金：国家自然科学基金(60475039)

摘　　要：分析了小型直升机的模型结构并讨论了对其进行辨识建模的必要性,设计了小型直升机的地面飞行实验台架用于辨识信号的采集,用7点二阶前推插值算法对野值进行识别和剔除,用拉格朗日插值算法对剔除后的野值进行补正,用一个低通滤波器对试验数据作低通滤波,滤除高频噪声信号,随后祛除掉趋势项和中位值.预处理后所得数据可用来辨识小型直升机系统的模型.The high complexity and instability of the small-scale helicopter＇s model are analyzed,so the system identification for the model helicopter is necessary.The ground stand is designed and made for collecting the data and validating control methods.The 7 points second-order derivation forward insert data algorithm is used to recognize and eliminate the outliers,and the Lagrange insert data algorithm is used to remedy the data.A low-pass filter is used to trail off the high-band noise.Subsequently,the mean and trend are peeled off.The resulting data is used to identify the model of small-scale helicopter.

关 键 词：小型直升机 系统辨识 预处理 滤波 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TP15[自动化与计算机技术—控制科学与工程]