基于改进支持向量回归的联合指纹定位算法

Joint Fingerprint Positioning Algorithm Based on Improved SVR Model

针对复杂的室内环境下,传统的射频识别技术(radio frequency identification,RFID)室内定位技术获得的接收信号强度特征向量维数较低,不能充分描述环境信息,无法获得良好的定位效果的问题,基于联合指纹提出一种鲁棒性强的高精度室内定位算法。该算法首先从RFID阅读器接收到的信号中提取信号强度和相位差数据,建立指纹库。然后利用凹函数递减策略改进PSO算法,优化SVR模型训练样本数据,建立参考标签的指纹特征和其与阅读器距离的映射关系。最后利用改进PSO算法迭代寻优,从而提高室内定位精度和鲁棒性。在仿真中,将该算法与GA-SVR和PSO-SVR算法进行比较,分析了不同指纹数据集和噪声对定位性能的影响。仿真结果表明,在相同指纹数据集和环境下,该算法的定位精度和系统稳定性均优于其他两种算法。

In order to solve the problem that the dimension of characteristic vector of received signal strength obtained by traditional RFID(radio frequency identification)indoor positioning technology is low,which can't fully describe the environment information and achieve good positioning effect in complex indoor environment,a robust and high precision indoor positioning algorithm based on joint fingerprint is proposed.Firstly,the algorithm extracted received signal strength indication and phase difference data from signals received by RFID readers,and established the fingerprint database.Then the PSO algorithm was improved by the concave function decreasing strategy and optimized the SVR model,trained the sample data,so the mapping relationship between the fingerprint characteristics of reference tags and their distance to the reader was established.Finally,the improved PSO algorithm was used to optimize iteratively to improve indoor positioning accuracy and robustness.In the simulation,the proposed algorithm was compared with GA-SVR and PSO-SVR algorithms,and the effects of different fingerprint data sets and the noise on positioning performance were analyzed.The simulation results show that the positioning accuracy and system stability of the proposed algorithm are better than those of the other two algorithms in the same fingerprint data set and environment.