基于因子图优化的DBSCAN聚类组合导航算法

DBSCAN Cluster Combinatorial Navigation Algorithm Based on Factorial Graph Optimization

在传感器工作中,无人车发动机震动、道路颠簸等外部因素都会引起传感器定位数据的误差,针对这一现象文中提出了一种基于因子图优化的DBSCAN聚类算法,在震动环境中收集传感器定位数据集,使用因子图优化对数据进行降噪,可有效缩减噪音点,再通过DBSCAN聚类算法进行簇划分,对类中数据求得重心,从而确定最终测算坐标位置,可以在震动环境中有效提升定位数据精度。

In response to this phenomenon,a DBSCAN clustering algorithm based on factor graph optimization is proposed,which collects sensor positioning data set in the vibration environment,uses factor map optimization to reduce the noise of the data,which can effectively reduce the noise point,and then divide the cluster by DBSCAN clustering algorithm to find the center of gravity of the data in the class,so as to determine the final measured coordinate position.It can effectively improve the accuracy of positioning data in vibration environment.