【作者】 周伟；

【导师】 王雷；

【作者基本信息】 大连理工大学 ， 软件工程， 2018， 硕士

【摘要】 近年来,无人机行业发展迅速,市面上出现了越来越多的无人机。一方面,无人机行业的迅速发展给我们的日常生活带来很多便利和乐趣;另一方面,无人机的非法的使用也给我们带来很多困扰和损失,尤其是最近几年屡见不鲜的无人机“黑飞”事件严重扰乱了社会治安和人民生活。所以,在机场、私人场所等安全性要求较高的地方对无人机进行检测,及时的发现非法无人机的闯入极为重要。实现对无人机的定位也有助于解决这一问题,实际上无人机的定位本质上是3D定位。此外,很多情况下平面定位已经无法满足我们的需求,3D定位的需求日益迫切,可以想象在一栋高楼里面,如果可以进行3D定位,将会大大降低我们搜寻的成本。面对这些问题,本文提出了基于信道状态信息的无人机检测系统和基于AoA的特征点约束粒子群优化3D定位算法。无人机检测系统收集无人机和控制器的通信信号,从无人机的移动性、空间性、振动性3个方面对无人机的运动特性进行了分析,结果表明可以从无人机和控制器的通信信号中得出无人机这3个物理特性,进一步利用这3个物理特性来判断是否有无人机出现。对于3D定位,本文首先对3D定位的时间复杂度进行分析,结果表明如果直接将2D定位算法应用于3D定位,时间复杂度将会随着参与定位的AP节点的增多而指数增加。对此,本文提出了基于AoA的特征点约束粒子群优化3D定位算法,该算法首先将3D定位问题转化为优化问题,然后根据AoA测量信号在3D环境中的分布特点,计算出有代表性的特征点对解搜索空间进行约束,最后使用PSO算法查找全局最优解,即目标节点的位置。本文对无人机检测系统进行了实测实验来验证系统的性能,实验结果表明我们的无人机检测系统可以达到87.3%的准确率和85.8%的召回率。对于3D定位算法,我们使用Matlab工具进行了仿真实验,实验结果表明我们提出的3D定位算法性能良好,平均定位精度可以达到0.7米。更多还原

【Abstract】 In recent years,the drone industry has developed rapidly and there are more and more drones on the market.On the one hand,the rapid rise of the drone industry has brought a lot of convenience and pleasure to our daily lives;on the other hand,the illegal use of drones has also brought us many problems and losses,which seriously disrupts social security and people’s lives.Therefore,it is extremely important to detect drones in places with high security requirements such as airports and private places.In addition,in many cases,2D plane localization can no longer meet our needs,and the demand for 3D positioning is becoming increasingly urgent.It can be imagined that in a tall building,if 3D localization can be performed,the cost of our search will be greatly reduced.In order to solve these problems,this paper proposes a drone detection system based on RF signal physical layer information and an AoA-based feature-points constrained PSO Localization algorithm.The drone detection system collects the communication signals of the drone and the controller,and then analyzes the drone’s movement characteristics including drone’s mobility,spatiality,and vibration which is useful for detecting the drones.For 3D localization,this paper analyzes the time complexity of 3D localization.The results show that if the 2D localization algorithm is directly applied to 3D localization,the time complexity will increase exponentially with the number of AP nodes participating in the positioning.This paper proposes an AoA-based feature point constrained PSO 3D localization algorithm.The algorithm converts the 3D localization problem into an optimization problem,and then calculates the feature points to constrain the search space,and finally use the PSO algorithm to find the global optimal solution,i.e.,the position of the target node..This paper conducts measured experiments to verify the performance of the drone detection system.The experimental results show that our drone detection system can achieve a precision of 87.3% and a recall of 85.8%.For the 3D localization algorithm,we use the Matlab tool for simulation experiments.The experimental results show that our proposed 3D positioning algorithm performs well and the average positioning accuracy can reach 0.7 meters.更多还原