基于MEMS惯性器件的微型船用组合测姿系统设计  

Design of Minisize Integrated Attitude Measurement System Based on MEMS Inertial Sensors

作　　者：赵琳[1] 刘付强[2] 王文晶[1] 徐博[1] 

机构地区：[1]哈尔滨工程大学自动化学院,哈尔滨150001 [2]中国空间技术研究院总体部,北京100094

出　　处：《系统仿真学报》2009年第14期4433-4436,共4页Journal of System Simulation

基　　金：国家自然科学基金(60474046)

摘　　要：首先构造了基于Kalman滤波的MEMS陀螺仪、加速度计和磁强计的姿态测量系统,在证明船舶机动运动产生的干扰加速度会引起Kalman滤波发散的基础上,构造了径向基函数神经网络姿态系统以克服干扰加速度的这种不利影响,并提出了基于神经网络与kalman滤波的组合姿态测量系统的设计方案,对设计的组合姿态测量系统进行了仿真试验。试验结果表明,组合姿态测量系统即使在干扰加速度存在的情况下也可以完成船舶姿态的实时准确估计,可满足船用设备中低精度姿态测量系统使用要求。An attitude measurement system of MEMS gyros, accelerometers and magnetometer was constructed based on Kalman filter firstly. After demonstrating that the disturbance acceleration induced by ship maneuvers would cause the Kalman filter to diverge, an attitude measurement system based on radial basis function neural network was constructed to overcome such disadvantages, an integrated attitude measurement system project based on the neural network and Kalman filter was designed, and simulations were used to test the designed integrated system. The simulation results show that even when the disturbance acceleration exists, the designed integrated attitude measurement system can also achieve real-time and precise estimates of ship attitude, and meet the needs of the attitude measurement system of middling or low precision.

关 键 词：MEMS惯性器件 KALMAN滤波 姿态测量系统 神经网络 

分 类 号：U666.12[交通运输工程—船舶及航道工程]