【作者】 柳明； 刘雨； 苏宝库； 蒋效雄；

【Author】 Liu Ming Liu Yu Su Baoku Jiang Xiaoxiong (Space Control and Inertial Technology Research Center,Harbin Institute of Technology,Harbin 150001,China)

【机构】 哈尔滨工业大学业空间控制与惯性技术研究中心；

【摘要】 在进行惯导平台误差模型辨识的过程中,真实的输入加速度难以获得,因此不得不使用测量值来代替,这使得输入输出观测数据中均含有测量噪声。应用传统的最小二乘方法进行辨识将会得到有偏的结果。总体最小二乘法是解决输入、输出观测数据中均含有噪音的信号处理问题的一种有效方法。因此提出应用总体最小二乘方法进行惯导平台模型辨识工作。首先给出了惯导平台的误差模型并推导了视加速度误差方程,讨论了总体最小二乘方法,并同时应用最小二乘法和总体最小二乘法进行仿真对比研究。仿真结果表明,在输入输出观测数据均含有噪声的情况下,总体最小二乘法对惯导平台误差模型辨识精度较最小二乘法要高,可消除输入数据中噪声的影响。更多还原

【Abstract】 Because the real input acceleration can not be obtained during the error model identification of the inertial navigation platform,so the measurement value is used to substitute the input data.Then both of the input and output data contain noises.If the traditional least squares method is used to do the work of error model identification of inertial navigation platform,the result will be a biased result.The total least squares method is one of the effective methods to deal with the problem that both of the input and output data con- tain noises.So the total least squares method is proposed to identify the error model of inertial navigation platform.Firstly,models of inertial navigation platform error and observation error are proposed,and the total least squares method is introduced.Both the least squares method and the total least squares method are used to do the simulation research work.The simulation results show that when both of the input and output data contain noises,the efficiency of the total least squares method is better than the least squares method.更多还原