基于数据融合的MEMS陀螺信号处理研究  

Research of the MEMS Gyro Signal Processing Based on Multi-sensor Data Fusion

作　　者：刘宇[1] 张小林[1] 段耀宇[1] 贺涛[1] 胡少伟[1] 曾燎燎[1] 

机构地区：[1]重庆邮电大学光纤通信重点实验室,重庆400065

出　　处：《压电与声光》2011年第4期554-556,565,共4页Piezoelectrics & Acoustooptics

基　　金：国家自然科学基金资助项目(10926072);重庆市科委自然科学基金资助项目(CSTG2007BB2448);重庆市教委科研基金资助项目(HJ080518;KJ110507)

摘　　要：针对当前MEMS陀螺仪信号为非平稳、非线性、不同步和信号量大的特点,提出了一种基于数字信号处理器(DSP)信号高速实时处理平台的、并结合多传感器数据融合技术的MEMS陀螺信号处理方案。该方案设计了基于SCC1300MEMS陀螺和DSP的信号采集处理平台,并运用正交基神经网络数据融合算法对采集的陀螺数据进行融合处理。实验结果表明,该方案在保证对陀螺信号进行高速实时采集的基础上,能提高信号处理精度达37.45%。Since the signals of the current MEMS gyroscope have the features of nonstationarity, nonlinearity, nonsynchronization and larger semaphore,a MEMS gyro signal processing scheme based on the high-speed and realtime signal processing plalform with the digital signal processor（DSP） and combined with the multi-sensor data fusion technology has been proposed in this paper. A signal acquisition processing platform based on SCC 1300 MEMS gyroscope and I）SP has been designed. The acquired MEMS gyroscope data were processed by using the neural net- work fusion algorithm with orthogonal basis function. The experimental results indicated that this program could ensure that the MEMS gyro signal was collected in high-speed and the signal processing accuracy has been increased by 37.45G.

关 键 词：MEMS陀螺仪 数字信号处理器 多传感器数据融合 神经网络 

分 类 号：V241.5[航空宇航科学与技术—飞行器设计]