基于LED可见光通信的室内定位关键技术研究

发布时间：2018-06-29 22:22

本文选题：可见光通信 + 室内定位系统　；参考：《上海大学》2016年博士论文

【摘要】：室内定位技术在商业中心、大型公共建筑(地铁、机场、图书馆等)、高危工业区、医院及养老院等场所有着广泛的应用前景,目前如何提高定位实时性及定位精度仍是研究难点,因此研究室内定位技术具有重要的理论意义和实用价值。基于LED可见光通信(VLC)的室内定位技术可利用已有的照明LED灯具,用移动终端的摄像头接收的方式实现光通信与室内定位。该技术与其他定位技术相比,有精度高、可移植性强和实施简单等优点,受到学者们重视。本文分析了现有用单摄像头做接收端的LED VLC室内定位技术,针对其通信编码效率、定位精度、视频目标跟踪算法有效性等问题展开研究,并通过构建实物原型系统进行实际效果验证。本论文主要创新工作及研究成果如下:1)为了实现普通照明与光通信的兼容,根据混光原理设计了新型LED筒灯,该筒灯亮度、色温独立可调,可实现高性能白光照明,同时三基色独立可控,可实现高效光通信。提出基于改进人工鱼群算法的LED混光优化算法。为加快人工鱼群算法的收敛速度,引入相对基强化学习算法优化人工鱼群算法的种群;为加快收敛速度又保持解精度,引入3参数Lorentzian函数和正态分布函数作为自适应权重算子优化视域和步长。用改进算法进行LED混光优化,仿真结果表明,改进算法收敛速度及收敛率都优于原始算法。优化后的筒灯测试结果表明,LED颗粒利用效率、发光效率均高于用其他混光方法设计的灯具,并保持高的显色指数。2)为了减小照度不均对定位精度的影响,提出采用改进人工鱼群算法优化室内灯具布局。由于灯具坐标略微变化,均匀度变化很小,因此解空间比较平坦,优化算法易陷入局部最优。引入多父体交叉算子、柯西变异算子和最小代沟模型使人工鱼群算法的种群保持多样性并兼顾收敛速度,避免早熟陷入局部最优。使用前述LED筒灯作为光源,进行室内不同灯具布局条件下照度仿真,Matlab和Dialux输出结果表明改进人工鱼群算法比原算法有更快的收敛速度,更高的收敛率,优化后均匀度提高了53%。3)针对现有LED VLC通信编码方法存在编码效率低的问题,提出了变字长颜色编码(VLCC)技术。设计了编、解码方法,从理论上定义了VLCC一系列性质。和现有的几种编码方法相比,新编码技术编码效率最高。变字长颜色编码技术能够用一帧图像完成完整的数据传输,包含字节头且数据位足够长,而其他编码方法需要传输多帧图像才能完成。使用前述LED筒灯进行了实验验证,表明该算法切实可行,具有实用价值。4)研究了根据灯具位置信息对静态目标进行定位的方法。现有单摄像头目标位置计算方法是用针孔相机模型推导出来的,且未考虑灯具尺寸问题,精度较差,需要多次校准。为此采用普通相机成像原理,研究了用单摄像头进行目标定位新算法。结合灯具形状的及颜色编码特点,提出先用形态学滤波,再用轮廓搜索来寻找圆的方法来寻找灯具图像,提取包含颜色区和数据区的条纹,进行颜色头的提取与验证。依据变字长颜色编码技术的解码规则,对数据区条纹进行滤波与解码,根据推导的新定位公式最终实现定位。实测结果表明目标定位误差在5cm以内。5)研究了动态目标的跟踪定位技术。针对跟踪目标几何特征和颜色编码特点,提出了粒子目标跟踪算法。新算法用颜色代替颜色直方图作为目标特征,引入启发式搜索方法;在预测目标区多放粒子,周边区少放粒子,并根据搜索区域的大小调整粒子数目;在预测目标区如未搜索到,进行启发式全局搜索。实验测试表明,新算法在日光灯干扰、局部遮挡等情况下鲁棒性良好,在目标有加速度的情况下反应速度远优于粒子滤波跟踪算法。实测40个点的坐标位置,程序计算值的跟踪误差在5cm以内。结果表明新定位系统有很好的鲁棒性、较高的准确度、以及良好的实时性,可产生很高的社会价值和经济价值。
[Abstract]:Indoor positioning technology is widely applied in commercial center, large public building (subway, airport, library, etc.), high risk industrial area, hospital and nursing home. It is still a difficult problem to improve the positioning real-time and positioning accuracy. Therefore, it is of great theoretical significance and practical value to study indoor positioning technology. Based on L ED visible light communication (VLC) indoor positioning technology can make use of the existing lighting LED lamps and the cameras received by the mobile terminal camera to realize optical communication and indoor positioning. Compared with other positioning techniques, the technology has the advantages of high precision, strong portability and simple implementation. This paper has been paid attention to by scholars. This paper analyzes the existing single camera. In the receiver, LED VLC indoor positioning technology is used to study the communication coding efficiency, positioning accuracy, and the effectiveness of the video target tracking algorithm. The actual results are verified by building a prototype system. The main innovation work and research results in this paper are as follows: 1) in order to realize the compatibility of ordinary lighting and optical communication, based on The principle of mixing light is designed for a new LED tube lamp. The light brightness and color temperature are independent and adjustable, and high performance white light can be realized. At the same time, three basic colors are independent and controllable, and high efficiency optical communication can be realized. A LED hybrid optimization algorithm based on improved artificial fish swarm algorithm is proposed. The optimization of the algorithm is optimized to speed up the convergence speed of artificial fish algorithm, and the relative base reinforcement learning algorithm is introduced. The population of artificial fish swarm algorithm; in order to speed up the convergence speed and maintain the solution precision, the 3 parameter Lorentzian function and normal distribution function are introduced as the adaptive weighting operator to optimize the horizon and step length. The improved algorithm is used to optimize the LED mixed light. The simulation results show that the improved algorithm has better convergence rate and convergence rate than the original algorithm. The test results show that the efficiency of LED particles is higher than that of the luminaire designed by other light mixing methods, and the high color rendering index (.2) is maintained. In order to reduce the influence of uneven illumination on the positioning accuracy, an improved artificial fish swarm algorithm is proposed to optimize the layout of the interior luminaire. The solution space is flat and the optimization algorithm is easy to fall into the local optimal. The multi parent crossover operator, the Cauchy mutation operator and the minimum generation gap model make the population of the artificial fish swarm keep diversity and the convergence speed is taken into account, and the early maturing is avoided in the local optimum. The LED tube lamp is used as the light source to carry out the indoor lighting layout conditions. Degree simulation, Matlab and Dialux output results show that the improved artificial fish swarm algorithm has faster convergence speed than the original algorithm, higher convergence rate, and 53%.3) after optimization. In view of the problem of low coding efficiency in the existing LED VLC communication coding method, a variable word length color coding (VLCC) technology is proposed. A series of properties of VLCC are defined in theory. Compared with several existing coding methods, the new coding technology has the highest coding efficiency. The variable word length color coding technology can complete the data transmission with one frame, including the byte head and the long data bit, and the other coding methods need to transmit multi frame images to complete. The LED down lamp is used. Experimental verification shows that the algorithm is practical and practical and is of practical value.4). The method of locating the static target according to the position information of the lamps and lanterns is studied. The existing single camera target location calculation method is derived from the pinhole camera model, and does not consider the luminaire size problem, the precision is poor, and many calibration methods are needed. In general camera imaging principle, a new algorithm for target location using a single camera is studied. Combining the shape of the lamp and the feature of color coding, the method of morphological filtering and contour search is used to find the lamp image, and the stripes are extracted with the color area and data area, and the color head is extracted and verified. The decoding rules of the long color coding technology, filter and decode the stripes of the data area, and finally realize the positioning according to the new positioning formula derived. The measured results show that the target positioning error is within the 5cm.5) and studies the tracking and positioning technology of the dynamic target. The new algorithm uses color instead of color histogram as the target feature and introduces a heuristic search method. In predicting the target area, the particle number is placed, the surrounding area is few particles, and the number of particles is adjusted according to the size of the search area. The heuristic global search is carried out in the prediction target area if not searched. The experimental test shows that the new algorithm is in the fluorescent lamp. When the target has acceleration, the response speed is far superior to the particle filter tracking algorithm. The coordinate position of the measured 40 points and the tracking error of the calculated values are within 5cm. The results show that the new positioning system has good robustness, high accuracy, and good real-time performance. A high social and economic value.
【学位授予单位】：上海大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TN929.1

【相似文献】