自适应权重更新的两步定位算法

Two-Stage Locating Algorithm with Adaptively Updated Weight

在室内环境中,影响定位精度的测量误差包括接收设备自身引起的误差以及信号非视距传播和多径效应所引起的测量正偏差。针对室内环境中测量数据包含测量误差服从正均值高斯分布的特性,提出了一种自适应权重更新的两步定位算法。该算法使用卡尔曼滤波和自适应权重更新的加权最小二乘算法进行两步定位,通过对每个测量距离分配不同的权重,克服了固定权重分配需在特定环境下方能获得良好定位精度的缺点。仿真结果表明,该算法定位精度优于两步定位算法和EKF算法,且对环境适应性更强。

The main impact on position accuracy of wireless indoor positioning systems is measure- ment error. The measurement error consists of an error caused by the receiver and a bias component in rich multipath indoor radio propagation environment. This paper proposes a two-stage locating method which adaptively updates the weighting matrix to deal with the issue of measurement error subject to the positive mean of the Gaussian distribution characteristics. The two-stage locating algo- rithm is composed by Kalman filtering and weighted least square algorithm with adaptively updated weight. By assigning different weight to each distance measurement, the proposed algorithm overcomes the shortage that satisfying location performance can only be achieved in certain environment by two-stage locating algorithms with fixed weighting matrix. Simulations indicate that the proposed method obtains higher accuracy compared with two-stage locating algorithms of fixed weighting matrix and Extended Kalman filter algorithms,and has better applicability to various environments.