一种基于光纤捷联惯性系统大样本统计的参数有效期评估方法

标定参数有效期是武器系统在使用过程中的一项重要指标，传统方法使用单套惯组的多次测试数据来评估标定参数的有效期。但这种抽测方法存在样本较少、耗时长的缺点。为克服以上问题，提出了一种利用正态分布和卡方检验分析多套惯组的单次测试数据来评估标定参数有效期的方法，样本量大、置信度高。用此评估方法对75套光纤惯组单次的测试数据进行统计分析，结果表明标定参数满足有效期要求，证明了本评估技术的有效性、可行性。同时，还分析了部分标定参数误差的漂移机理，为今后设计出长期稳定性更高的光纤惯组提供了理论支撑。

A Evaluation Method of Validity Period of Calibration Parameters Based on Large Sample Fiber Optic IMU

The validity period of calibration parameters is an important index in the use of weapon systems. The multiple test data of single IMU is used to evaluate validity period of calibration parameters by the traditional method, but this method has the shortcomings of fewer samples and long time-consuming. In order to overcome the above problems, a method which analyzes multiple sets of IMU''s single test data with normal distribution and Chi-square test to evaluate validity period of calibration parameters is proposed, and this method has larger samples and higher confidence. Using this method, 75 sets of IMU''s single test data are analyzed. The results show that calibration parameters meet the requirements of the validity period, which proves the validity and feasibility of this evaluation. The drift mechanism of some calibration parameter errors is analyzed, which provides theoretical support for the future design of fiber optical IMU with higher long-term stability.