基于改进支持向量回归的室内定位算法

发布时间：2018-03-11 15:23

本文选题：室内定位　切入点：过滤　出处：《仪器仪表学报》2017年09期 　论文类型：期刊论文

【摘要】：针对室内无线局域网环境中无线信号不稳定,以及传统支持向量回归定位算法在构建位置坐标与信号强度时的单输出导致位置坐标信息之间的关联性降低的问题,提出一种基于改进支持向量回归的室内定位方法。该算法首先对采集到的接收信号强度(RSS)指纹进行对数处理使其更符合正态分布,然后采用高斯滤波过滤掉小概率的指纹值之后构建指纹数据库;其次,为了降低单独构建x与y坐标模型的误差,提高二维位置信息与RSS之间的关联性,在训练阶段增加训练一个校正坐标z=x·y;最后,根据加权反K近邻的方法得到最优的位置坐标。实验结果表明,提出的算法可以减少室内复杂环境带来的噪声干扰,与传统的支持向量回归定位算法相比有更高的定位精度。
[Abstract]:Aiming at the instability of wireless signal in indoor WLAN environment, and the problem that the single output of traditional support vector regression localization algorithm in constructing position coordinate and signal intensity leads to the decrease of the correlation between position coordinate information. An indoor localization method based on improved support vector regression is proposed. Firstly, logarithmic processing of received signal intensity (RSS) fingerprint is carried out to make it more suitable for normal distribution. Secondly, in order to reduce the error of constructing X and y coordinate model separately and improve the correlation between two-dimensional position information and RSS, Gao Si filter is used to filter out the fingerprint value of small probability. At last, the optimal position coordinates are obtained according to the weighted inverse K nearest neighbor method. The experimental results show that the proposed algorithm can reduce the noise disturbance caused by the complex indoor environment. Compared with the traditional support vector regression localization algorithm, the location accuracy is higher.
【作者单位】：杭州电子科技大学通信工程学院;
【基金】：国家自然科学基金(61100044,61671192) 浙江省科技计划公益技术应用研究(2017C31055) 浙江省自然科学基金(LY16F010012)项目资助
【分类号】：TN92

【相似文献】