摘要
为了消除光纤陀螺的温度效应并提高陀螺的精度,将神经网络应用于光纤陀螺温度漂移模型的辨识和温度补偿中.采用神经网络算法中的误差反向传播算法,在采集的光纤陀螺漂移数据样本的基础上对其进行了非线性辨识,通过对所采集数据的学习训练出了有效的温度补偿网络,并用实验测试数据验证了该方法的有效性.数据分析表明,神经网络温度补偿方法可以明显改善陀螺的零偏稳定性能.
Artificial Neutral Network was applied into temperature drift model identification and temperature compensation to eliminate the temperature effect and improve its precision. Adopting the arithmetic of error Back Propagation method to identify the FOG drift nonlinear model on the base of the bias data, and trained an effect network which can realize the temperature compensation of FOG by studying the bias data, then validated it with the test bias data. The analysis suggested that the ANN compensation method can improve FOG performance.
出处
《陕西科技大学学报（自然科学版）》
2008年第5期95-98,共4页
Journal of Shaanxi University of Science & Technology
关键词
光纤陀螺
温度补偿
神经网络
fiber optic gyro
temperature compensation
Artificial Neutral Network
