应急管理巡护轨迹的优化方法及实验仿真

Optimization method and experimental simulation of emergency management patrol trajectory

为了解决应急管理系统APP中巡护轨迹的漂移问题,该文提出基于MapReduce并行化编程模型对轨迹进行优化处理的方法,首先基于GeoHash和加权汉明距离的DBSCAN聚类算法(轨迹漂移修正算法)筛选漂移点,然后采用三点一次平滑算法(轨迹平滑算法)对漂移点进行平滑处理。通过仿真实验展现轨迹优化效果表明,轨迹漂移修正算法能有效去除漂移点,轨迹平滑算法能减少轨迹的锯齿状,通过分析轨迹漂移修正算法和轨迹平滑优化算法的时间复杂度表明,基于加权汉明距离的DBSCAN算法效率高,900个坐标的数据集运行时间小于0.1 s,单服务器可满足轨迹优化应用。

In order to solve the drift problem of patrol trajectory in the emergency management system APP, this paper proposes a method to optimize the trajectory based on MapReduce parallel programming model. Firstly, the DBSCAN clustering algorithm based on GeoHash and weighted Hamming distance(trajectory drift correction algorithm) is used to screen the drift points, and then the three-point smoothing algorithm(trajectory smoothing algorithm) is used to smooth the drift points. The simulation experiment shows that the trajectory drift correction algorithm can effectively remove the drift points, and the trajectory smoothing algorithm can reduce the sawtooth of the trajectory. By analyzing the time complexity of the trajectory drift correction algorithm and the trajectory smoothing optimization algorithm, it shows that the DBSCAN algorithm based on weighted Hamming distance has high efficiency, and the running time of data sets of 900 coordinates is less than 0.1 s, which indicates that a single server can meet the trajectory optimization application.