采用BP神经网络的四频激光陀螺零偏的光强补偿方法

• 作 者: 赵洪常 肖光宗 汪之国 张斌
• 机 构: 国防科学技术大学光电科学与工程学院光电工程系,
• 出 处: 红外与激光工程
• 发表日期: 2013年02期
• 关键词: 四频激光陀螺 零偏 BP神经网络 光强 four-frequency laser gyro bias BP neural network light inten

总结了环境因素对四频激光陀螺光强和零偏影响的3种常见方式,指出光强变化和零偏变化都是陀螺经环境因素作用后的外在表现,光强更能全面细致地反映环境因素对陀螺零偏的影响,提出了以光强为基准采用BP神经网络对零偏进行补偿的方法。建立了陀螺零偏的BP神经网络,采用该神经网络针对开机零漂、跳模和高低温冲击等3种不同的零漂产生机理分别进行了补偿实验,实验表明:该方法能够有效辨识产生零漂的不同机理并进行补偿,对提高复杂环境下陀螺的零漂性能有一定的参考意义。

Summarizes the environmental factors on the quadrupole laser gyroscope light intensity and zero bias effects of the three common ways that the light intensity changes and zero bias are the gyroscopic environmental factors after the external manifestations of light more comprehensive and meticulous reflect the environment Factor on gyro zero bias, a method of compensating zero offset using BP neural network based on light intensity is proposed. The gyro zero-bias BP neural network is established. The compensation experiment is carried out using the neural network for three different zero-drift mechanisms, such as zero drift, mode hopping and high and low temperature shocks. The experimental results show that this method can effectively identify zero- Drift of different mechanisms and compensation, to improve the zero drift performance of gyroscope in complex environment has some reference value.