In order to reduce bias drift in startup phase of fiber-optic gyroscope（ FOG） and improve measurement precision

a compensation method for FOG startup drift based on statistical product and service solutions（ SPSS） is proposed. The SPSS statistical analysis software is applied to modeling of FOG startup drift at room temperature. It is founded that the main factors affecting FOG start-up drift are temperature change rate

temperature and temperature square. The temperature change rate is selected as the independent variable after considering many factors. A linear compensation model for startup drift is established. By the test

the compensation model is effective and can improve the FOG measurement accuracy. The alignment test results show that the accuracy of course angle is improved by about 70 % on average after compensation.