基于微粒群优化LSSVM的室内指纹定位算法  

Indoor Fingerprint Localization Algorithm Based on Particle Swarm Optimization Using Least Square Support Vector Machine

作　　者：赵妍[1] 乐燕芬[1] 施伟斌[1] ZHAO Yan;LE Yan-fen;SHI Wei-bin(School of Optical-Electrical and Computer Engineering,University ofShanghai for Science and Technology,Shanghai 200093,China)

机构地区：[1]上海理工大学光电信息与计算机工程学院,上海200093

出　　处：《软件导刊》2019年第4期87-90,94,共5页Software Guide

摘　　要：为了降低利用最小二乘支持向量机(LSSVM)定位过程中参数选取对定位精度的影响,提出一种基于微粒群进行参数优化的室内指纹定位算法。该算法通过离线采集的RSSI数据训练最小二乘支持向量机,利用微粒群算法寻找并确定LSSVM全局最优参数,获得基于位置指纹的LSSVM定位模型。仿真结果表明,相对于传统LSSVM定位,PSO-LSSVM有效提高了定位准确度,并能在小样本情况下保持良好的定位精度。The application of least square support vector machine(LSSVM)for indoor localization suffers from instability of accuracy as the parameters are chosen based on experience. In order to reduce the influence,we use particle swarm optimization to improve posi. tioning accuracy. We obtain the training samples from offline to the LSSVM,and find the optimal LSSVM parameters in PSO. The ob. tained optimal parameters are then used in the positioning. Simulations show that compared with the original LSSVM position method, PSO-LSSVM method effectively improves localization accuracy and maintains good positioning accuracy in small sample.

关 键 词：指纹定位 微粒群算法 最小二乘支持向量机 RSSI 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]