捷联姿态算法非互易误差的分析与修正

在捷联惯性导性导航系统中，陀螺仪的输出是迭代时间内载体的角增量的量化值；由其解算出的载体姿态存不可交换性误差，该误差的大小与捷联姿态算法密切相关，研究的目的在于选择适当的捷联算汉，减小姿态算法中的不可交换误差，提高捷联算法的精度，以鱼雷捷联惯性姿态基准为例，对四元数法的捷联姿态算法误差和基于等效旋转矢量的优化的三子样捷联姿态算法误差进行了分析比较，数字仿真结果表明，三子样法比四元数法的捷联姿态算精度高7至8个数量级，而前者的采样周期却低一个数量级，显著地提高了捷联姿态算法的制度。

Analysis and correcting of irreplaceable errors of strapdown attitude algorithms

In the strapdown inertial system, the Gyro outputs are quantized value of the vehicle angle increment during the iteration interval and there are irreplaceable errors which are related closely to the strapdown algorithms in the vehicle attitude resolution obtained from the Gyro outputs. This paper aims at selecting the right algorithm to reduce the irreplaceable errors in attitude. With the torpedo SINS as an example, two kinds of strapsown attitude errors,coming from respectively the quaternion algorithms and the optimizing three samples based on the equivalent rotation vector,are analyzed and compared with each other. The results of digital simulation indicate that the strapdown attitude precision of the three samples is 7 to 8 grades higher than that of the quaternion algorithms,and the sample period of the former is 1 grade smaller than that of later,and the strapdown algorithm has obviously higher accuracy.