论文部分内容阅读
陀螺仪是惯性导航系统的重要组成部分,其精度依赖于惯性导航系统的精度.为了提高陀螺仪的精度,针对陀螺随机漂移非线性、弱平稳性引起的随机误差,以激光陀螺仪随机漂移时间序列数据为研究对象,首先通过对陀螺仪建模的分析和对激光陀螺仪实时数据的分析和预处理,得到了陀螺漂移误差的离散时间序列;然后对其基于遗传规划(GP)建模,得出了当前时刻陀螺漂移数据和前几时刻的陀螺漂移数据之间的非线性数学模型;最后利用遗传算法(GA)对该模型有数学关系的参数进行优化,得到更高精度的模型.仿真结果表明:与经典自回归(AR)建模优化方法相比,GP+GA建模能够更加有效地反映陀螺仪的随机漂移特性,陀螺仪的方差降低了73.72%,与经典自回归(AR)建模方法相比效果提高了4.72%.该建模方法有效补偿了陀螺仪的随机漂移误差,提高了系统的稳定性.
Gyroscope is an important part of inertial navigation system, and its accuracy depends on the accuracy of inertial navigation system.In order to improve the accuracy of gyroscope, aiming at the random error caused by random drift of gyroscope and weak stationaryness, gyroscope’s random drift time Firstly, the discrete time series of gyroscope drift error is obtained through the analysis of the gyroscope modeling and the analysis and preprocessing of the laser gyroscope real-time data. Then, based on the genetic programming (GP) modeling, The nonlinear mathematical model between the gyro drifting data and the gyro drifting data at the previous moment is obtained. Finally, the parameters which have the mathematical relation with the genetic algorithm (GA) are optimized to get the model with higher accuracy. The results show that the GP + GA modeling can reflect the random drift of the gyroscope more effectively than the classical autoregression (AR) modeling and optimization method. The variance of the gyroscope is reduced by 73.72% Compared with the simulation results, the modeling method improved by 4.72%. This modeling method effectively compensates for the random drift error of the gyroscope and improves the system stability.