通过对干涉型光纤陀螺误差来源及其影响因素的分析,建立了与温度变化量及温度变化速率相关的角速率误差概率密度函数,并提出一种针对慢时变与弱非线性随机过程的分段等概率误差补偿方法.该方法采用非线性手段,对一系列具有统计平稳性的随机过程进行参数估计与预测,能够实现模型动态跟踪,近似程度较高,且充分发挥了并行计算的优势,可避免由补偿过程所造成的输出时间滞后.几种典型工作状况下的实际补偿结果表明其能有效地减小角速率及航向角误差.
To set a probability density function for drift caused by variation of environmental temperature, the error model of interferometric fiber optic gyro were analyzed. Based on this work, a segmented error compensation strategy was presented, which uses non-linear method to estimate and predicate parameters for non-stationary stochastic process, and then does error compensation according to the probability density function. It has a good approaching capability to dynamic model, and also supports parallel computing, that can avoid time delay brought by compensation process. Results under several typical work conditions showed good effect on error compensation for both angle rate and heading angle calculating.