基于Newton插值与混合灰狼优化SVR的RFID定位算法

Indoor Positioning Algorithm for RFID Based on Newton Interpolation and Support Vector Regression Optimized by Hybridizing Grey Wolf Optimization

针对RFID定位系统中,传统的LANDMARC定位算法定位精度不高且需要布置大量参考标签的问题,提出了一种基于牛顿插值和差分进化改进灰狼优化支持向量回归机(Newton-HGWOSVR)的定位算法。该算法采用高斯滤波对参考标签的采样数据进行预处理;通过牛顿插值法对其它位置的信号强度进行估值来扩充信号数据库;通过HGWOSVR构建物理位置和信号接收强度值之间的非线性关系来完成定位。实验结果表明,该算法提升了室内定位的精度,并且能减少布置参考标签的工作量,提高了室内定位方法的工作效率。

Since the traditional LANDMARC location algorithms have poor positioning accuracy and cost laborious efforts to decorate reference tags in the RFID positioning system, a novel positioning algorithm proposed is based on the Newton interpolation method and support vector regression optimized by hybridizing grey wolf optimization with differential evolution （HGWOSVR）. By using the proposed algorithm, Gaussian filter was used to deal with the sampling data of the reference tags.Newton interpolation method was adopted to estimate the RSS value of other reference tags to expand database. The HGWOSVR algorithm was employed to build the nonlinear relationship between the RSS value of reference tags and their locations to predict the positioned tags. Simulation results show that the proposed algorithm performs better in terms of positioning accuracy, and reduces the workload of decorating reference tags, which improves working efficiency of the indoor location positioning method.