發(fā)布時(shí)間：2018-04-06 07:12

  本文選題：室內(nèi)定位　切入點(diǎn)：位置指紋　出處：《南京郵電大學(xué)》2017年碩士論文

【摘要】：隨著無(wú)線通訊技術(shù)的高速發(fā)展和各式各樣智能終端的大范圍使用,用戶對(duì)擁有高準(zhǔn)確性的室內(nèi)定位服務(wù)的需求在急劇增長(zhǎng),在治病就醫(yī)、抗震救災(zāi)、公共社交媒體、交通工具導(dǎo)航和場(chǎng)景監(jiān)控等領(lǐng)域都展現(xiàn)出了巨大的市場(chǎng)遠(yuǎn)景�；赪LAN位置指紋室內(nèi)定位算法實(shí)現(xiàn)起來(lái)相對(duì)簡(jiǎn)單,構(gòu)建起定位環(huán)境的成本較低,符合在絕大多數(shù)室內(nèi)定位中應(yīng)用的需求,成為當(dāng)下室內(nèi)定位研究中的熱門(mén)技術(shù)。本文對(duì)基于WLAN的位置指紋定位技術(shù)進(jìn)行了研究。首先論述了基于WLAN的位置指紋定位算法中普遍存在的誤差及其產(chǎn)生原因,并總結(jié)了已有的相應(yīng)解決方案。本文從室內(nèi)定位中離線采樣和在線定位兩個(gè)階段著手,從AP定位性能差異性分析、指紋數(shù)據(jù)庫(kù)的構(gòu)建、數(shù)據(jù)庫(kù)聚類(lèi)分塊和匹配定位四個(gè)主要方面進(jìn)行了深入分析和研究。以降低建立位置指紋數(shù)據(jù)庫(kù)時(shí)的工作量和同時(shí)提高定位精度為目標(biāo),提出了一種基于AP信號(hào)方差的位置指紋室內(nèi)定位的改進(jìn)算法。由于AP的分布位置不同及AP自身的差異性,使得在定位過(guò)程中不同AP對(duì)定位效果的影響不同,該方法區(qū)分了不同AP在定位過(guò)程中的貢獻(xiàn)大小,仿真實(shí)驗(yàn)表明,基于AP信號(hào)方差的改進(jìn)定位算法不但縮短了指紋數(shù)據(jù)庫(kù)的建立時(shí)間還提高了定位的精度。其次,重點(diǎn)分析了K均值聚類(lèi)方法在基于WLAN的位置指紋室內(nèi)定位系統(tǒng)中的應(yīng)用,總結(jié)分析了該聚類(lèi)算法在定位過(guò)程中的利弊,并從理論的角度提出了相應(yīng)的改進(jìn)方案。最后,因?yàn)镵均值聚類(lèi)方法中存在類(lèi)與類(lèi)邊界處定位性能差的問(wèn)題,不能對(duì)類(lèi)與類(lèi)邊界處參考點(diǎn)進(jìn)行準(zhǔn)確的選取,這直接導(dǎo)致在類(lèi)與類(lèi)相鄰處的定位精度較低。為解決這個(gè)問(wèn)題,本文提出了基于二次K均值聚類(lèi)的位置指紋室內(nèi)定位方法,通過(guò)對(duì)第一次聚類(lèi)結(jié)果的再處理,減少了因類(lèi)與類(lèi)相鄰處參考點(diǎn)選擇不合理而帶來(lái)的誤差,提高了定位的精度。
[Abstract]:With the rapid development of wireless communication technology and the wide use of all kinds of intelligent terminals, the demand of users for indoor positioning services with high accuracy is increasing dramatically, and they are seeking medical treatment, earthquake relief, public social media,Vehicle navigation and scene monitoring and other areas have shown a huge market perspective.Based on WLAN location fingerprint indoor location algorithm is relatively simple to achieve, the cost of building a location environment is relatively low, in line with the needs of most indoor positioning applications, it has become a hot technology in the research of indoor location.In this paper, the location fingerprint location technology based on WLAN is studied.In this paper, the errors and their causes in the location fingerprint location algorithm based on WLAN are discussed, and the corresponding solutions are summarized.This paper starts with the two stages of off-line sampling and on-line positioning in indoor positioning, and makes in-depth analysis and research on four main aspects of AP location performance difference analysis, fingerprint database construction, database clustering and matching location.In order to reduce the workload of establishing location fingerprint database and improve the location accuracy, an improved location fingerprint indoor location algorithm based on AP signal variance is proposed.Because of the different distribution of AP and the difference of AP itself, the effect of different AP on the localization effect is different. The method distinguishes the contribution of different AP in the positioning process.The improved location algorithm based on AP signal variance not only shortens the time of establishing fingerprint database, but also improves the accuracy of location.Secondly, the application of K-means clustering method in the location fingerprint indoor location system based on WLAN is analyzed, the advantages and disadvantages of the clustering algorithm in the localization process are summarized and analyzed, and the corresponding improvement scheme is put forward from the theoretical point of view.Finally, because the K-means clustering method has the problem of poor localization performance at the boundary of classes and classes, it is impossible to select the reference points at the boundary of classes and classes accurately, which directly leads to the low positioning accuracy at the adjacent areas of classes and classes.In order to solve this problem, a location fingerprint indoor location method based on quadratic K-means clustering is proposed in this paper. By reprocessing the results of the first clustering, the error caused by the unreasonable selection of reference points between the cluster and the cluster is reduced.The accuracy of positioning is improved.
【學(xué)位授予單位】：南京郵電大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類(lèi)號(hào)】：TN925.93;TP311.13

【參考文獻(xiàn)】

相關(guān)期刊論文 前3條

1 張凱淵;劉佩林;錢(qián)久超;裴凌;;多傳感器融合機(jī)器人室內(nèi)定位系統(tǒng)設(shè)計(jì)與實(shí)現(xiàn)[J];信息技術(shù);2014年11期

2 何艷麗;;無(wú)線傳感器網(wǎng)絡(luò)質(zhì)心定位算法研究[J];計(jì)算機(jī)仿真;2011年05期

3 徐鳳燕;李j賓;王宗欣;;一種新的基于區(qū)域劃分的距離-損耗模型室內(nèi)WLAN定位系統(tǒng)[J];電子與信息學(xué)報(bào);2008年06期

相關(guān)博士學(xué)位論文 前1條

1 張明華;基于WLAN的室內(nèi)定位技術(shù)研究[D];上海交通大學(xué);2009年

相關(guān)碩士學(xué)位論文 前1條

1 趙聰;基于位置指紋的WLAN室內(nèi)定位算法研究[D];哈爾濱工業(yè)大學(xué);2014年

，

本文編號(hào)：1718508