基于卡尔曼滤波和改进DBSCAN聚类组合的GPS定位算法

GPS positioning algorithm based on Kalman filter and improved DBSCAN clustering combination

实时获取智能移动终端的地理位置信息是增强现实(AR)实景智能导航系统实现的关键,为了提高智能终端GPS定位的精度,提出了一种基于卡尔曼滤波与改进的具有噪声的基于密度的聚类方法(DBSCAN)结合的GPS组合定位优化方法.通过对GPS系统采集到的位置坐标数据进行卡尔曼滤波,去除较大的数据波动,控制定位误差范围,采用DBSCAN聚类算法进行分类去噪和二次聚类,对类中数据求得算术均值和类间数据总数进行加权求重心,确定位置坐标.实验结果表明,提出的算法能有效提高GPS单点定位精度,减少定位误差,同时很好地满足了AR实景智能导航系统实时性和鲁棒性的要求.

Real-time acquisition of geographic location information of smart mobile terminals is the key to the realization of an augmented reality(AR)real-scene smart navigation system.In order to improve the accuracy of GPS positioning for smart terminals,a GPS combined positioning optimization method based on Kalman filtering and improved DBSCAN clustering algorithm is proposed.Kalman filtering is performed on the position coordinate data collected by the GPS system to remove large data fluctuations and control the positioning error range.Using DBSCAN clustering algorithm for classification denoising and secondary clustering,the arithmetic mean value of the data in the class and the total number of data between the classes are weighted to find the center of gravity,and the position coordinates are determined.The experimental results show that the proposed algorithm can effectively improve the GPS single-point positioning accuracy,reduce positioning errors,and at the same time well meet the real-time and robustness requirements of the AR real-world intelligent navigation system.