基于MDS-MAP和非线性滤波的传感器网络定位算法研究木.pdfVIP

基于MDS-MAP和非线性滤波的传感器网络定位算法研究 1,2 1 2 陈岁生 ，卢建刚，楼晓春 1. 江大学控制科学与工程学系，工业控制技术国家重点实验室，杭州 310027 E-mail: css@hzvtc.edu.cn 2. 杭州职业技术学院友嘉机电学院， 杭州 310018 摘 要：本文在MDS-MAP定位算法的基础上，根据测量距离与传感器节点之间存在的非线性关系，引入扩展 卡尔曼滤波(EKF)求精算法和不敏卡尔曼滤波(UKF)求精算法对MDS-MAP 求得的节点坐标进行求精。对 MDS-MAP定位算法、MDS-MAP和EKF相结合的定位算法（MDS-EKF ）、MDS-MAP和UKF相结合的定位方 法(MDS-UKF)就定位精度进行了比较，实验结果表明，EKF和UKF等非线性滤波方法能提高定位精度，并且 在相同条件下，MDS-UKF定位算法的定位精度更高，其生成的网络拓扑图最接近于实际网络拓扑图。 关键词：节点自定位，MDS-MAP ，EKF ，UKF 中图分类号：TP393 文献标识码：A Research on the Sensor Networks Localization Algorithm Based on MDS-MAP and Nonlinear Filtering 1,2 1 2 Chen Sui-sheng Lu Jian-gang Lou Xiao-chun 1. Department of Control Science and Engineering, State Key Laboratory of Industrial Control Technology, Zhej iang University, Hangzhou, 310027; 2. Fair Friend School of Electromechanics, Hangzhou Vocational Technical College, Hangzhou, 310018 Abstract Based on the MDS-MAP localization algorithm, according to the nonlinear relationship between the measurement of distance and sensor node, we apply the Extended Kalman Filter(EKF) and the Unscented Kalman Filter(UKF) to refine the locaized coordinates which obtained by MDS-MAP. The location accuracy were compaird between the MDS-MAP localization algorithm, the algorithm which combines MDS-MAP and EKF(MDS-EKF), and the algorithm which combines MDS-MAP and UKF(MDS-UKF). Experiments results show that the EKF and UKF and other nonlinear filtering algorithms can improve the location accu