单兵导航多传感器信息融合自适应滤波算法

基于足绑式INS的单兵导航系统,通过将惯性导航系统、人体运动学约束、磁强计等多传感器信息进行融合得到准确的单兵导航信息。对于匀速步行时的单兵导航,可采用普通Kalman滤波算法进行多传感器信息融合,但不能满足跑步等激烈运动模式下单兵导航需求。提出衰减记忆新息自适应Kalman滤波算法,可满足多步态模式下的单兵导航多源信息融合的需求,实验结果验证了算法的有效性。

Adaptive Filtering Algorithm of Multi-Sensor Information Fusion for Individual Navigation

Individual Navigation based on foot-mounted inertial navigation system(INS) can get accurate individual navigation information through the INS,human kinematics constraints,magnetometer and other multi-sensor information. For the walking at a constant speed in individual navigation,Kalman filter algorithm can be used for multi-sensor information fusion,however,the requirement can not be met in running or other intense movement modes. In this paper,an adaptive Kalman filter algorithm is proposed,which can meet the requirement of multi-source information fusion in multi-step mode. The experimental results verify the effectiveness of the algorithm.