基于EKF视觉组合导航方法的研究
发布时间：2018-09-03 10:46
【摘要】：随着导航技术不断发展,单一的导航系统已不能满足载体导航的精度和稳定性,组合导航成为目前发展的主流方向。GPS/SINS组合是当前流行的导航方式之一,能较好的提供载体的速度、姿态和位置信息,但是当载体处于GPS信号受到干扰的区域(如街道、隧道、峡谷等),GPS/SINS组合导航的精度会受到干扰,稳定性降低。为解决GPS信号缺失而带来的影响,本文引入光流视觉导航系统,将GPS/SINS/光流进行组合,并进行了相关研究。首先根据捷联惯导解算原理和惯导元件的数学模型,设计了载体的轨迹发生器。利用轨迹发生器提供载体的速度、姿态和位置信息进行捷联解算matlab实验仿真,得出纯捷联惯导的不足。为解决捷联惯导误差发散的问题,提出GPS/SINS组合方法,考虑载体为微型四旋翼飞行器,系统为非线性系统,故采用EKF代替KF进行数据融合滤波,并进行实验仿真,表明GPS/SINS组合要优于纯捷联惯导。然后,引入光流视觉导航系统,选用Hom-Schunck法进行光流计算载体的速度信息。为了提高导航系统的容错能力,选用联邦滤波组合结构,设计了GPS/SINS和SINS/光流两个子滤波器,组成了 GPS/SINS/光流的组合导航系统,并采用EKF进行数据滤波。通过四旋翼实际飞行试验对比,结果表明GPS/SINS/光流组合的导航效果要优于GPS/SINS组合。
[Abstract]:With the continuous development of navigation technology, the single navigation system can not meet the accuracy and stability of carrier navigation. Integrated navigation has become the mainstream direction of development. GPS / sins integration is one of the most popular navigation methods. It can provide the velocity, attitude and position information of the carrier, but the accuracy of GPS / sins integrated navigation will be disturbed and the stability will be reduced when the carrier is in the area where the GPS signal is interfered (such as streets, tunnels, canyons, etc.). In order to solve the influence caused by the lack of GPS signal, the optical flow visual navigation system is introduced in this paper, the GPS/SINS/ optical flow is integrated, and the related research is carried out. Firstly, the track generator of carrier is designed according to the principle of sins solution and the mathematical model of inertial navigation element. The matlab simulation of strapdown solution is carried out by using the velocity, attitude and position information of carrier provided by the locus generator, and the shortage of pure strapdown inertial navigation is obtained. In order to solve the problem of error divergence of strapdown inertial navigation system, a GPS/SINS combination method is proposed. Considering that the carrier is a micro four-rotor aircraft and the system is a nonlinear system, EKF is used instead of KF for data fusion filtering, and the experimental simulation is carried out. It shows that the GPS/SINS combination is superior to the pure strapdown inertial navigation system. Then, the optical flow visual navigation system is introduced, and the Hom-Schunck method is used to calculate the velocity information of the carrier. In order to improve the fault-tolerant ability of the navigation system, the federated filter integrated structure is selected. Two sub-filters, GPS/SINS and SINS/, are designed to form the integrated navigation system of the GPS/SINS/ optical flow, and the EKF is used to filter the data. The results show that the navigation effect of GPS/SINS/ optical flow combination is better than that of GPS/SINS combination.
【学位授予单位】：天津工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN967.2
本文编号：2219760
[Abstract]:With the continuous development of navigation technology, the single navigation system can not meet the accuracy and stability of carrier navigation. Integrated navigation has become the mainstream direction of development. GPS / sins integration is one of the most popular navigation methods. It can provide the velocity, attitude and position information of the carrier, but the accuracy of GPS / sins integrated navigation will be disturbed and the stability will be reduced when the carrier is in the area where the GPS signal is interfered (such as streets, tunnels, canyons, etc.). In order to solve the influence caused by the lack of GPS signal, the optical flow visual navigation system is introduced in this paper, the GPS/SINS/ optical flow is integrated, and the related research is carried out. Firstly, the track generator of carrier is designed according to the principle of sins solution and the mathematical model of inertial navigation element. The matlab simulation of strapdown solution is carried out by using the velocity, attitude and position information of carrier provided by the locus generator, and the shortage of pure strapdown inertial navigation is obtained. In order to solve the problem of error divergence of strapdown inertial navigation system, a GPS/SINS combination method is proposed. Considering that the carrier is a micro four-rotor aircraft and the system is a nonlinear system, EKF is used instead of KF for data fusion filtering, and the experimental simulation is carried out. It shows that the GPS/SINS combination is superior to the pure strapdown inertial navigation system. Then, the optical flow visual navigation system is introduced, and the Hom-Schunck method is used to calculate the velocity information of the carrier. In order to improve the fault-tolerant ability of the navigation system, the federated filter integrated structure is selected. Two sub-filters, GPS/SINS and SINS/, are designed to form the integrated navigation system of the GPS/SINS/ optical flow, and the EKF is used to filter the data. The results show that the navigation effect of GPS/SINS/ optical flow combination is better than that of GPS/SINS combination.
【学位授予单位】：天津工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN967.2
【参考文献】
相关期刊论文 前9条
1 杨天雨;贾文峰;赖际舟;邓一民;;惯性/光流/磁组合导航技术在四旋翼飞行器中的应用[J];传感器与微系统;2016年01期
2 赵海;陈星池;王家亮;曾若凡;;基于四轴飞行器的单目视觉避障算法[J];光学精密工程;2014年08期
3 沈宁;唐大全;李飞;;无人机视觉导航中的图像处理及位姿解算[J];光电技术应用;2012年06期
4 刘小明;陈万春;邢晓岚;殷兴良;;光流/惯导多传感器信息融合方法[J];北京航空航天大学学报;2012年05期
5 李建;李小民;钱克昌;;无人机GPS/SINS/Vision组合导航技术[J];中国惯性技术学报;2011年04期
6 邵玮;祝小平;赵刚;;基于视觉/GPS/MEMS-INS数据融合的MUAV导航方案[J];系统仿真技术;2011年01期
7 王小刚;郭继峰;崔乃刚;;INS/Vision相对导航系统在无人机上的应用[J];哈尔滨工业大学学报;2010年07期
8 李耀军;潘泉;赵春晖;于盈;李军伟;;基于INS/SMNS紧耦合的无人机导航[J];中国惯性技术学报;2010年03期
9 黄显林;姜肖楠;卢鸿谦;李明明;;自主视觉导航方法综述[J];吉林大学学报(信息科学版);2010年02期
，本文编号：2219760
本文链接：https://www.wllwen.com/shoufeilunwen/xixikjs/2219760.html
