采用ZigBee节点网络的室内定位系统研究

发布时间：2018-08-13 14:55

【摘要】：随着数据业务和多媒体业务的快速增长,人们对定位和导航的需求日益增加,尤其是在复杂的室内环境,如机场大厅、展厅、仓库、超市、图书馆、地下停车场等,常常需要定位移动终端或其持有者在室内的位置。与室外环境相比,室内定位受定位时间、定位精度及室内复杂环境等条件的限制,导致目前比较完善的室外定位方法还无法有效地运用到室内环境中,因此寻找定位精度高、稳健性好、实现简单、成本低的室内定位系统已经成为当今业界研究的热点。在阐述室内定位的研究现状并比较常见的室内定位系统后,选择基于ZigBee的无线传感器网络来实现室内定位系统的开发。在搭建好室内定位系统的软硬件平台后,通过比较目前几种常用的室内定位方法,选择了方便易行的基于接收信号强度指示(RSSI)信息对系统进行研究。针对在RSSI测距过程中由于环境噪声的影响使定位误差在超过一定的距离后过大的问题,实现了一种分区域加权质心算法,该算法将定位区域分成多个子区域来实现近距离定位,还采用了节点距离倒数之和代替距离和的倒数作为权重,同时对权重进行系数修正以充分利用节点信息。仿真结果表明,该算法的相对定位精度比原有加权质心算法提高了25%,并且算法稳健性更好。由于室内复杂环境所造成的测距误差是不可避免的,导致采用传统定位算法的定位系统的性能受到限制,还利用多层前馈(BP)人工神经网络的容错能力和非线性映射能力来进一步改善定位性能。仿真结果表明,该网络在6m×3m的区域内平均定位精度为0.21m。此外,针对网络在训练过程中易陷入局部极小点的问题,运用粒子群算法(PSO)来优化网络的初始权值和阈值,实验结果表明该改进算法的定位精度可达0.16m,比普通BP神经网络的相对定位精度提高了23.8%。
[Abstract]:With the rapid growth of data and multimedia services, the demand for positioning and navigation is increasing, especially in complex indoor environments, such as airport halls, exhibition halls, warehouses, supermarkets, libraries, underground parking lots, etc. It is often necessary to locate the location of the mobile terminal or its holder indoors. Compared with outdoor environment, indoor positioning is limited by the conditions of positioning time, positioning precision and complex indoor environment, which results in the relatively perfect outdoor positioning methods can not be effectively applied to indoor environment, so the precision of locating is high. The indoor positioning system, which has good robustness, simple implementation and low cost, has become a hot spot in the industry. After expounding the research status of indoor positioning and the common indoor positioning system, the wireless sensor network based on ZigBee is selected to realize the development of indoor positioning system. After setting up the hardware and software platform of the indoor positioning system, by comparing several commonly used indoor positioning methods, a convenient and easy method based on the received signal strength indication (RSSI) information is selected to study the system. In order to solve the problem that the location error is too large after a certain distance in the process of RSSI ranging, a sub-region weighted centroid algorithm is implemented, which divides the location area into several sub-regions to realize the short distance location. The sum of reciprocal distance of nodes is used to replace the reciprocal of distance sum as weight, and the coefficient of weight is modified to make full use of node information. The simulation results show that the accuracy of the algorithm is 25% higher than that of the original weighted centroid algorithm, and the robustness of the algorithm is better than that of the original weighted centroid algorithm. Because the ranging error caused by indoor complex environment is inevitable, the performance of localization system using traditional localization algorithm is limited. The fault tolerance and nonlinear mapping ability of multilayer feedforward (BP) artificial neural network are also used to further improve the location performance. The simulation results show that the average positioning accuracy of the network is 0.21 m in the region of 6m 脳 3m. In addition, the particle swarm optimization algorithm (PSO) is used to optimize the initial weights and thresholds of the network. The experimental results show that the accuracy of the improved algorithm can reach 0.16m, which is 23.88m higher than that of the common BP neural network.
【学位授予单位】：上海师范大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TN92

【相似文献】