基于智能手机的车辆弯道识别与相对定位的研究

发布时间：2018-01-07 15:20

本文关键词：基于智能手机的车辆弯道识别与相对定位的研究　出处：《黑龙江大学》2015年硕士论文　论文类型：学位论文

【摘要】：车辆在弯路上行驶时,由于驾驶员的技术生疏或是违规行驶等原因,极易造成车辆转弯时产生侧滑、碰撞等交通事故的现象。所以,保障弯道处的交通安全尤为重要。其中,车道识别尤其是弯道识别在智能交通领域作用也更为显著。同时,智能手机作为一种好的载体,能够为智慧交通领域提供多种功能应用,这使得利用智能手机进行智能交通领域的设备开发成为了一种新的解决问题的思路与方法。本文研究的问题正是基于智能手机的车辆弯道识别与相对定位问题。首先,本文研究了探测弯道的传感器的选择问题。本文通过实际采集车辆在道路上的行驶数据信息,将数据按照弯道的类型进行分类,利用假设检验的拟合检验处理,最终确定识别弯道所需的传感器以及其变量值。实验证明,这些被筛选出来的传感器以及变量值可以使弯道识别算法达到百分之九十以上的准确率。其次,本文研究了基于智能手机传感器的弯道识别算法。在探索弯道识别所需传感器的基础上,本文针对实验采集数据的特点,提出了一种融合距离最近判别准则和卡方统计量的弯道识别方法,探索手机传感器采集的数据对弯道识别的影响。通过大量实验证明,该算法可以很好的对于七种道路类型进行识别,准确率在90%以上。最后,本文研究了基于智能手机的车辆相对位置定位算法。该算法分为两种情景,一种是直路情景,一种是弯路情景。本文证明了直道中的车辆相对位置定位算法不适用于弯道中的车辆相对位置定位,并提出了融合弯道识别算法与手机wifi的信号强度值的弯道车辆相对位置定位算法。试验结果显示,定位的平均准确率可以达到90%以上,而且不会受到天气和汽车类型的影响。
[Abstract]:The vehicle on the detour, because the driver is rusty or cause illegal driving, easily causing the vehicle sideslip when the collision accident phenomenon. Therefore, to protect the corners of traffic safety is very important. Among them, especially the curve lane recognition recognition function in the field of intelligent transportation is more significant at the same time, the intelligent mobile phone as a good carrier, can provide a variety of functions for the application of intelligent transportation field, which makes the development of the field of intelligent transportation equipment, the use of intelligent mobile phone has become a way of thinking and a new method to solve the problem. The problem is the problem of vehicle road recognition intelligent mobile phone and relative positioning based on this study. First, the selection of sensors to detect corners. The actual running data information acquisition vehicle on the road, the data according to the type of curve Classified by hypothesis test fitting test, and ultimately determine the bend sensor identification needed and the variable value. Experiments show that these sensors are selected and the variable value can make the road recognition algorithm to achieve more than ninety percent accuracy. Secondly, this paper studies the curve recognition algorithm based on intelligent sensor based on mobile phone. Explore the road recognition required sensor, based on the characteristics of the experimental data, and proposes a new fusion distance curve recognition method recently criterion and chi square statistics, explore the impact of mobile phone collecting sensor data on the curve recognition. Through a lot of experiments show that the algorithm can be very good for the seven types of road recognition accuracy rate above 90%. Finally, this paper studies the intelligent mobile phone of the vehicle relative location algorithm based on the algorithm is divided into two. A scene, a scene is straight, a scene is detours. This paper proves that the straight vehicles in the relative location algorithm is not suitable for the corners of the vehicle relative location, and put forward the vehicle signal intensity curve curve fusion recognition algorithm and WiFi values of the relative position of the mobile phone positioning algorithm. The test results show that the the average positioning accuracy rate can reach above 90%, and not affected by the weather and the types of cars.

【学位授予单位】：黑龙江大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U495;TN929.53

【相似文献】