一种基于微机械电子系统惯性传感器室内行人航位推算算法

摘要

The pedestrian dead reckoning (PDR) system is widely applied in indoor localization for its no requirement of external auxiliary information.Aiming at the problem of step length limitation and other issues in traditional PDR,a pedestrian adaptive algorithm based on micro electro mechanical system (MEMS) sensors is proposed.The six-position testing method and the Kalman filter (KF) are selected as the scheme to deal with the errors of the original data of accelerometers and gyroscopes.The motion state of feet is calculated by zero-crossing detection and gait short-time invariance.The travel distance is fitted adaptively in the presence of the motion state conjunct accelaration.The trajectory of the pedestrian is calculated through the position estimation.Simulation results show that the pedestrian adaptive algorithm is suitable for practical indoor pedestrian positioning with good accuracy and flexibility which error less than 1.5 m in 95 m motion distance.%在室内行人定位中,行人航位推算(pedestrian dead reckoning,PDR)由于不需要外部辅助信息,而被广泛应用.针对传统室内PDR存在步长局限性等问题,提出了一种基于微机械电子系统(micro electro mechanical system,MEMS)传感器的行人航位自适应拟合推算算法.该算法选用六位置法和卡尔曼滤波器(Kalman filter,KF)作为对加速度计和陀螺仪原始数据误差处理方案.通过过零检测和步态短时不变性计算脚的运动状态;并结合加速度自适应拟合行进距离,最后利用位置推算解算行人的运动轨迹.仿真结果表明,该算法在95 m运动距离内,最大误差不超1.Sm,具有良好精确性和灵活性,适用于实际的室内行人定位.