基于自适应神经网络的激光陀螺温度补偿研究

摘要

针对捷联惯性导航系统中激光陀螺的输出信号随温度漂移的问题,研究了激光陀螺的零偏与温度的关系,建立了一种新的考虑温度变化率的零偏温度补偿模型.在分析BP神经网络的基础上,提出了一种基于线性再励的自适应变步长神经网络算法进行激光陀螺的零偏温度模型系数的辨识.仿真结果表明,该方法能够有效地进行温漂补偿,从而提高惯组系统的导航精度.%In view of the problem of bias error of output signal caused by temperature drift of laser gyro in SINS, the relationship of them is studied. The new temperature compensating model which takes the tempera-ture changing rate into account is established. On the base of analyzing the BP neural network, a new neural network algorithm based on adaptive variable stepsize via linear reinforcement is proposed to identify the model coefficient. Simulation results show that this method can improve the accuracy of inertial navigation system by the means of compensating the bias error effectively.