K均值聚类自适应零速检测人员定位技术

零速区间检测准确度直接制约着零速修正算法对于改善人员定位系统提供位置精度的能力.针对现有零速检测方案存在阈值设定以及训练模型准确性等问题,提出一种基于加速度幅值滑动方差开展行走步态零速区间检测方法,采用K均值聚类方法自适应纠正初始检测结果中的误检状态,构造Kalman滤波器并在零速区间以惯性系统解算的速度信息为观测量进行量测更新来限制导航误差积累.开展人员多种运动状态下的行走测试,实验结果表明K均值聚类自适应算法能对零速区间进行有效地检测,获取的位置解算误差小于2%.

Pedestrian Positioning Technology with K-means Clustering Adaptive algorithmPedestrian Positioning Technology with K-means Clustering Adaptive Zero Velocity Detection

The detection accuracy of Zero velocity interval is the premise to implement Kalman filtering to clear velocity error and correct position error in time.For the existing zero velocity detection scheme,there are some problems such as threshold setting and training model accuracy.This paper proposes a zero velocity interval detection method in walking gaits based on moving variance of acceleration amplitude,an adaptive K-means clustering method is used to correct the false identified states in initial detection results.A Kalman filter is constructed in zero velocity interval,which adopt the velocity information calculated by inertial system as measurements and carry out measure-ment update to restrict navigation error accumulation.Series experiments are carried out,experimental results show that K-means adaptive algorithm can effectively detect zero velocity interval,and the error of the obtained position calculation is less than 2%.