本文選題：室內(nèi)定位 + WLAN�。� 參考：《北京工業(yè)大學(xué)》2014年碩士論文

【摘要】：近年來，隨著無線通信技術(shù)的快速發(fā)展以及移動設(shè)備的逐漸普及，各種室內(nèi)環(huán)境下基于位置服務(wù)的需求變得越來越迫切。室內(nèi)是人類活動最為密集且停留時間最長的場合，人們迫切地需要實現(xiàn)室內(nèi)環(huán)境下的定位、導(dǎo)航以及人員物資的全方位監(jiān)控等智能化泛在服務(wù)。因此，室內(nèi)環(huán)境下基于位置的服務(wù)存在著大量的應(yīng)用需求和廣闊的應(yīng)用空間�；赪LAN位置指紋的室內(nèi)定位技術(shù)因其設(shè)備簡單，定位精度高而成為近年來室內(nèi)定位技術(shù)研究的熱點。本課題對該熱點進行了研究并主要完成了以下幾點工作： 第一，針對存在著大量干擾因素的室內(nèi)環(huán)境下接收信號強度RSSI波動性較大的問題，提出了通過卡爾曼濾波算法對RSSI進行預(yù)處理，從而有效消除隨機干擾，再現(xiàn)系統(tǒng)狀態(tài)。通過實驗確定了卡爾曼濾波算法的最佳參數(shù)值。 第二，針對卡爾曼濾波算法在RSSI出現(xiàn)躍變時，收斂速度慢的問題，提出了一種改進的卡爾曼濾波算法，該算法能夠利用前幾次的RSSI觀測值迅速判斷出RSSI是否發(fā)生了躍變，并在有RSSI躍變發(fā)生時，修改卡爾曼濾波算法的相關(guān)參數(shù)，降低算法對發(fā)生躍變前的狀態(tài)估計值的認(rèn)可度，從而提高算法在發(fā)生RSSI躍變情況后的收斂速度，減小RSSI估計誤差。通過實驗驗證了改進算法的有效性。 第三，針對大范圍定位時，算法的匹配運算量較大的問題，提出了一種基于區(qū)域劃分和AP ID過濾的匹配算法。該算法是通過區(qū)域劃分和AP ID過濾將較大定位區(qū)域內(nèi)大部分參考價值較低的位置指紋過濾掉，縮小指紋匹配的范圍，然后再采用最近鄰法確定更精確的估計位置，從而極大地減少匹配定位過程中的運算量，提高系統(tǒng)的定位速度。 第四，為了提高定位算法對運動目標(biāo)的實時跟蹤定位能力，減小定位誤差，還提出了一種基于卡爾曼濾波的室內(nèi)運動目標(biāo)實時定位算法，該算法通過對匹配算法得到的估計位置進行卡爾濾波處理，使定位結(jié)果最大程度的接近真實的運動軌跡。并通過實驗驗證了該算法的有效性。 第五，采用模塊化的設(shè)計思想，設(shè)計并實現(xiàn)了一種基于WLAN位置指紋的室內(nèi)定位系統(tǒng)驗證軟件，為指紋定位算法的研究和改進提供了一種平臺，并加入了躍變自適應(yīng)卡爾曼濾波算法和基于區(qū)域劃分和AP ID過濾的匹配算法，通過該軟件進行定位實驗，，驗證了本文所提算法的有效性，同時，也證明了該軟件具有良好的定位性能。
[Abstract]:In recent years, with the rapid development of wireless communication technology and the popularity of mobile devices, the need for location-based services in various indoor environments has become more and more urgent. Indoor is the most intensive and the longest stay of human occasions, people urgently need to realize the indoor environment positioning, navigation, personnel and materials of the omnidirectional monitoring and other intelligentized ubiquitous services. Therefore, location-based services in indoor environment have a large number of application requirements and broad application space. The indoor location technology based on WLAN position fingerprint has become a hot research area in recent years because of its simple equipment and high positioning accuracy. This topic has carried on the research to this hot spot and has mainly completed the following work: Firstly, aiming at the large volatility of received signal RSSI in indoor environment with a large number of interference factors, a Kalman filter algorithm is proposed to preprocess the RSSI, which effectively eliminates the random interference and reproduces the system state. The optimal parameter value of Kalman filter algorithm is determined by experiments. Secondly, an improved Kalman filter algorithm is proposed to solve the problem of slow convergence rate of Kalman filter algorithm when the RSSI jump occurs. The algorithm can quickly determine whether the RSSI has jumped or not by using the previous RSSI observations. When the RSSI jump occurs, the relevant parameters of the Kalman filter algorithm are modified to reduce the recognition of the state estimation value before the jump, so as to improve the convergence speed of the algorithm after the RSSI jump occurs and reduce the RSSI estimation error. The effectiveness of the improved algorithm is verified by experiments. Thirdly, a matching algorithm based on region partition and AP ID filtering is proposed to solve the problem of large amount of matching operation. The algorithm filters out most of the low-reference location fingerprint in the larger location area by region partition and AP ID filtering, reduces the range of fingerprint matching, and then uses the nearest neighbor method to determine the more accurate estimated location. Thus greatly reduces the operation in the matching localization process, enhances the system localization speed. Fourthly, in order to improve the ability of real-time tracking and localization of moving targets, a real-time localization algorithm based on Kalman filter is proposed. In this algorithm, the estimated position of the matching algorithm is processed by Karl filter, so that the location results are close to the real motion trajectory as much as possible. The validity of the algorithm is verified by experiments. Fifthly, a verification software of indoor location system based on WLAN location fingerprint is designed and implemented with the modular design idea, which provides a platform for the research and improvement of fingerprint location algorithm. The adaptive Kalman filtering algorithm and the matching algorithm based on region partition and AP ID filtering are added to the algorithm. The localization experiment is carried out by the software, and the validity of the proposed algorithm is verified, at the same time, It is also proved that the software has good positioning performance.
【學(xué)位授予單位】：北京工業(yè)大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2014
【分類號】：TN925.93

【參考文獻(xiàn)】

相關(guān)期刊論文 前5條

1 倪巍,王宗欣;基于接收信號強度測量的室內(nèi)定位算法[J];復(fù)旦學(xué)報(自然科學(xué)版);2004年01期

2 林以明;羅海勇;李錦濤;趙方;;基于動態(tài)Radio Map的粒子濾波室內(nèi)無線定位算法[J];計算機研究與發(fā)展;2011年01期

3 郎昕培;許可;趙明;;基于無線局域網(wǎng)的位置定位技術(shù)研究和發(fā)展[J];計算機科學(xué);2006年06期

4 石鵬,徐鳳燕,王宗欣;基于傳播損耗模型的最大似然估計室內(nèi)定位算法[J];信號處理;2005年05期

5 ;Radio-map Establishment based on Fuzzy Clustering for WLAN Hybrid KNN/ANN Indoor Positioning[J];中國通信;2010年03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

本文編號：1821599