基于神经网络的压电陀螺的零位漂移补偿  

Zero point drifts compensation for piezoelectric gyro based on BP neural network

作　　者：邱建勋[1] 潘英俊[1] 刘宇[1] 黎蕾蕾[1] 朱 

机构地区：[1]重庆大学光电工程学院,重庆400044

出　　处：《仪器仪表学报》2006年第z1期842-844,共3页Chinese Journal of Scientific Instrument

摘　　要：针对低精度的压电陀螺静态零位输出随温度漂移严重的问题,提出了用BP神经网络进行补偿的方法,该方法具有很强的非线性映射能力,能够有效减少压电陀螺的零位输出误差。实验结果与传统拟合方法比较证明,在不改变陀螺结构的情况下,该方法能有效地将陀螺的静态零位漂移误差减小一个数量级以上,从而使该类型陀螺能够应用于中、高精度惯性测量领域中。In the view of serious temperature excursion of piezoelectric gyro,we propose the way of BP neural network to compensate the temperature excursion.This method is very functional in non-linear mapping and it could reduce the output error of piezoelectric gyro.We compare the result with that of traditional fitting method,it is proved that this method could reduce the gyro's static bias temperature output error at least 1 order even we do not change the gyro's structure,so we could apply this type of gyro to medium high accuracy system.

关 键 词：压电陀螺 静态零位漂移 BP神经网络 建模 补偿 

分 类 号：TH7-55[机械工程—仪器科学与技术]