将非平稳时间序列的状态空间建模方法用于陀螺过渡过程的分析.基于平滑先验约束的概念，使用卡尔曼滤波和赤池的AIC方法拟合全局模型，得到陀螺漂移模型的若干数值结果并用于陀螺系统分析.由于观测序列的趋势项、不规则分量可同时建模，因此比分别建模在统计上更加准确有效.-Will be non-stationary time series state space modeling method for the analysis of the transition process gyro. Priori smoothness constraint based on the concept, the use of Kalman filtering and Akaike s AIC method of fitting the overall model, the gyro drift model and a number of numerical results in the gyro system analysis. Due to the trend observed sequence, the irregular component may be at the same time modeling, modeling than were statistically more accurate and effective.