【摘要】：隨著室內(nèi)定位技術(shù)的迅速發(fā)展,基于紅外線、超聲波、藍(lán)牙、超寬帶、Wi-Fi(Wireless Fidelity)等的定位技術(shù)迅速成為學(xué)術(shù)和應(yīng)用研究的熱點(diǎn)�；赪i-Fi的室內(nèi)定位在定位精度、穩(wěn)健性、安全性和復(fù)雜度等方面有著自身的優(yōu)勢(shì),該技術(shù)充分利用了現(xiàn)有的成熟硬件平臺(tái)和無線網(wǎng)絡(luò)接入點(diǎn),能在智能終端上以應(yīng)用程序的形式實(shí)現(xiàn)高精度的定位服務(wù),因而基于Wi-Fi的室內(nèi)定位研究備受關(guān)注。本文對(duì)基于Wi-Fi的室內(nèi)定位技術(shù)進(jìn)行深入探討,主要研究接收信號(hào)預(yù)處理技術(shù),離線階段的位置指紋庫構(gòu)建技術(shù)和在線階段的室內(nèi)定位算法三個(gè)方面。首先,為得到穩(wěn)定的位置指紋數(shù)據(jù),提高定位系統(tǒng)的穩(wěn)定性,根據(jù)室內(nèi)Wi-Fi信號(hào)的傳播模型和接收信號(hào)特征,研究基于卡爾曼濾波的信號(hào)強(qiáng)度預(yù)處理方法,使用對(duì)數(shù)譜域抑制信號(hào)多徑效應(yīng)的方法,對(duì)接收信號(hào)進(jìn)行預(yù)處理。其次,針對(duì)指紋庫構(gòu)建代價(jià)高的問題,提出數(shù)據(jù)插值方法,研究矩陣填充的方法,把基于SVT算法的矩陣填充應(yīng)用于低秩位置指紋庫的重建;提出使用密度峰值快速搜索聚類技術(shù)對(duì)位置指紋地圖進(jìn)行分類,并與K-means聚類和仿射傳播聚類對(duì)比,對(duì)位置指紋地圖進(jìn)行分類預(yù)處理。最后,研究在線定位方法,提出基于信號(hào)傳播模型的接收信號(hào)強(qiáng)度填充的定位方法,對(duì)基于位置指紋匹配的近鄰定位算法、貝葉斯定位算法和壓縮感知定位算法進(jìn)行詳細(xì)研究。此外,本文設(shè)計(jì)了一個(gè)基于Wi-Fi的室內(nèi)定位實(shí)驗(yàn)系統(tǒng)原型,該系統(tǒng)使用卡爾曼濾波和對(duì)數(shù)譜域抑制多徑效應(yīng)的方法對(duì)接收信號(hào)進(jìn)行預(yù)處理,采用密度峰值快速搜索聚類實(shí)現(xiàn)位置指紋庫分塊處理,使用貝葉斯算法和壓縮感知算法進(jìn)行在線定位。在實(shí)際室內(nèi)環(huán)境中完成參考點(diǎn)的布署,實(shí)現(xiàn)室內(nèi)區(qū)域的位置指紋地圖構(gòu)建,進(jìn)行定位實(shí)驗(yàn)。實(shí)驗(yàn)結(jié)果顯示,該定位系統(tǒng)具有較高的定位精度,定位誤差在2米以內(nèi)的累積概率為85%,同時(shí)能夠以較少的計(jì)算量實(shí)現(xiàn)穩(wěn)健的定位。
[Abstract]:With the rapid development of indoor positioning technology, the localization technology based on infrared, ultrasonic, Bluetooth, ultra-wideband, Wi-Fi (Wireless Fidelity) and so on has become a hot spot in academic and applied research. Indoor positioning based on Wi-Fi has its own advantages in positioning accuracy, robustness, security and complexity. This technology makes full use of existing mature hardware platforms and wireless network access points. The research of indoor location based on Wi-Fi is paid more attention to because it can realize the high precision location service in the form of application program on the intelligent terminal. In this paper, the indoor location technology based on Wi-Fi is deeply discussed, including the pre-processing technology of receiving signal, the construction technology of position fingerprint database in off-line stage and the indoor location algorithm in on-line stage. Firstly, in order to obtain the stable position fingerprint data and improve the stability of the positioning system, according to the propagation model of indoor Wi-Fi signal and the characteristics of the received signal, the signal intensity preprocessing method based on Kalman filter is studied. The received signal is preprocessed by using logarithmic spectral domain to suppress multipath effect. Secondly, aiming at the high cost of constructing fingerprint database, a data interpolation method is put forward, and the method of matrix filling is studied. The matrix filling based on SVT algorithm is applied to the reconstruction of low-rank position fingerprint database. In this paper, the fast searching clustering technique of peak density is used to classify the location fingerprint map, and compared with the K-means clustering and affine propagation clustering, the location fingerprint map is preprocessed. Finally, the on-line localization method is studied, and the location method based on signal propagation model is proposed. The nearest neighbor location algorithm based on location fingerprint matching, Bayesian location algorithm and compressed sensing location algorithm are studied in detail. In addition, a prototype of indoor positioning experiment system based on Wi-Fi is designed in this paper. The system uses Kalman filter and logarithmic spectral domain to pre-process the received signal. The location fingerprint database is partitioned by fast searching clustering with peak density, and online location is carried out by using Bayesian algorithm and compression perception algorithm. In the actual indoor environment, the reference points are deployed, the location fingerprint map of the indoor area is constructed, and the location experiment is carried out. The experimental results show that the positioning system has high positioning accuracy, the cumulative probability of positioning error within 2 meters is 85um, and the robust positioning can be achieved with less calculation.
【學(xué)位授予單位】：電子科技大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2016
【分類號(hào)】：TN92

【參考文獻(xiàn)】

相關(guān)期刊論文 前1條

1 劉云浩;楊錚;王小平;簡麗榮;;Location,Localization,and Localizability[J];Journal of Computer Science & Technology;2010年02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4條

1 秦泗明;基于位置指紋的WiFi室內(nèi)定位技術(shù)研究[D];電子科技大學(xué);2013年

2 魏雷;WIFI位置指紋定位技術(shù)研究及仿真器設(shè)計(jì)[D];西南交通大學(xué);2012年

3 楊清;基于指紋的無線室內(nèi)精確定位方法研究[D];浙江大學(xué);2011年

4 姜莉;基于WiFi室內(nèi)定位關(guān)鍵技術(shù)的研究[D];大連理工大學(xué);2010年

，

本文編號(hào)：2373098