一种优化的零速检测行人导航算法

A Novel Pedestrian Navigation Algorithm Based on ZVU

针对军事上对单兵定位和民用领域为行人提供位置服务的需求,提出了足部固定MIMU的行人导航算法。利用人体行走过程中脚部与地面相接触的静止时间段,采用零速更新(Zero Velocity Update,ZVU)对INS继续对准或标定,而零速检测又是实现零速修正的基础。研究对比了在足部固定MIMU进行零速检测的常用方法,针对常用算法阈值自适应性差,无法检测步速变化转弯等一些异常情况对行人零速检测产生误判等问题,提出优化的零速判断算法,使得加速度计信号和陀螺仪信号综合应用到零速检测。最后,利用改进算法设计行人室内平地正常行走(大约5km/h)包含步速变化和转弯过程的实验。实验结果表明,优化的零速检测算法相比之前的常用算法,能更准确地检测零速并进行修正,对导航轨迹和导航误差有更好的修正效果。

In order to meet the needs of the location of a solider in military fields and provide location service for pedestrian in civilian fields.We propose foot pedestrian navigation algorithm based on MIMU.We use zero-speed updates to the alignment of the INS or calibration at the stance period when the foot landing on the ground.The zero speed detection is the basis of zero velocity update.This paper compares the traditional approach in the foot fixed MIMU zero speed detection.As for threshold of conventional algorithms is not adaptive,and can not detect the impact of the problems caused by the change of velocity.We suggest an modified algorithm for zero speed judgment,making the accelerometer signal and the gyroscope signal in maximize application to zero speed detection.Finally,we carry out the experiment of indoor pedestrian walking in normal speed(about 5km/h).The experimental results show that the new zero speed detection algorithm can detect the zero speed more accurately than the traditional algorithm and correct the navigation error and the navigation error is better than the traditional algorithm.