论文部分内容阅读
温度漂移是存在于光纤陀螺系统中使得输出信号产生较大偏置误差的一种不可忽略因素,如何准确地辨识漂移并有效地对其进行补偿直接关系到陀螺的测量精度.文中比较了前馈网络中的BP网络和径向基函数(RBF)网络,采用RBF网络进行温漂辨识.温漂辨识可以通过离线事先学习,因而在多种学习方法中选择了简单易行、精度高且运算速度快的正交最小二乘(OLS)法.通过仿真验证,采用RBF网络及其OLS学习算法可以快速、有效、高精度地辨识并补偿温漂
Temperature drift is a non-negligible factor in the fiber optic gyroscope system, which causes large offset error of the output signal.However, how to accurately recognize the drift and compensate it effectively is directly related to the measurement accuracy of the gyroscope.This paper compares the performance of the feedforward BP neural network and Radial Basis Function (RBF) network in the network, using RBF network for temperature drift identification. Temperature drift identification can be learned offline beforehand, so a variety of learning methods to choose a simple, high precision and speed of operation Fast Orthogonal Least Squares (OLS) method.By simulation, RBF network and its OLS learning algorithm can be used to quickly and effectively identify and compensate the temperature drift