室内自校准导航定位方法

Indoor Self-Calibration Navigation and Positioning Methods

提出一种室内信号强度路径损耗自校准模型,与传统的因墙壁等障碍物引起的信号强度路径损耗模型相比,该自校准模型既不需要大量试验,而且精度还更高。在此基础上,推导出距离量测方程的协方差矩阵,建立一种基于RSSI的室内高精度定位方法。同时,提出一种新的步长估计模型和行人航位自校准推算方法,该模型通过综合利用X轴与Y轴加速度极差的平方和进行步长估计,与通过Z轴加速度极差估计步长的Weinberg模型相比,它不但克服了因地面不平(如上下坡或上下楼等)导致的步长误差,而且在平地上也具有更高的精度。并建立一种基于RSSI的室内自校准导航方法,能够通过自校准滤波对步长参数K进行实时估计,解决了传统方法需通过大量训练来确定参数K的问题,显著提高导航精度。

An indoor signal strength path loss self—calibration model is proposed.Compared with the traditional signal strength path loss model caused by obstacles such as walls,this self—calibration model does not require a lot of tests and has higher accuracy.On this basis,the covariance matrix of the distance measurement equations is derived,and an RSSI—based indoor high—accuracy positioning method is established.Meanwhile,a new step length estimation model and a pedestrian self—calibration dead—reckoning method are proposed.This model estimates the step length by comprehensively utilizing the sum of squares of acceleration ranges on X—axis and Y—axis.Compared with the Weinberg model which estimates the step length by using the acceleration range on Z—axis,the presented model not only overcomes the step length errors caused by ground slope(such as going uphill,downhill,upstairs or downstairs),but also has higher accuracy on flat ground.Moreover,an RSSI—based indoor self—calibration navigation method is established.In this method,the step length parameter K can be estimated in real time by self—calibration filters,which solves the problem that traditional approaches need plenty of training to determine the parameter K,and improves navigation accuracy significantly.