小型直升机航向模型的增广递推最小二乘辨识  

Model Identification of Robot Helicopter’s Yaw Model Using Recursive Extended Least Squares Method

作　　者：赵志刚[1] 吕恬生 田耕[2] 

机构地区：[1]上海交通大学机械工程学院,上海200023 [2]金川集团公司镍钴研究设计院,甘肃金昌737100

出　　处：《系统仿真学报》2007年第19期4386-4388,4406,共4页Journal of System Simulation

基　　金：国家自然科学基金资助项目(60475039)

摘　　要：分析确定了模型直升机航向通道结构,考虑了噪声模型的存在,采用增广最小二乘法进行了参数辨识,从而避免了一般最小二乘法把误差模型隐含在系统模型中所造成的辨识精度不高的缺点。结果表明,虽然采用增广最小二乘法要考虑噪声模型,但是它具有速度快、辨识结果准确的特点,所得模型直升机的航向模型很好的表示了其系统特性,可以用来进行控制系统的设计和直升机的分析理解。The structure of the small-scale helicopter yaw model was analyzed and established. The subsisting of error model was considered so that recursive extended least squares method （RELS）, which avoided to the error of identification precision from the least squires method, was used. The results indicate that though the noise model is considered in recursive extended least squares method, the identification process is rapidly and the resulting model is accurately enough. The resulted model expresses the characteristics of the yaw and is tised to design the control system and analyze the helicopter.

关 键 词：模型直升机 增广递推最小二乘法 系统辨识 噪声模型 

分 类 号：TP271.64[自动化与计算机技术—检测技术与自动化装置] TP18[自动化与计算机技术—控制科学与工程]