The invention relates to a road anomaly detection method based on window division and deformation clustering, and the method comprises the steps that 1, a mobile terminal carries out the threshold detection and sliding window processing of z-axis acceleration data, and screens a to-be-determined segment; 2, the mobile terminal judges whether the road segment is an abnormal segment or not through arandom forest algorithm; and step 3, the cloud deforms the data of the road abnormal segments to a uniform length through a deformation clustering method. The method has the beneficial effects that apossible road anomaly window is determined by utilizing threshold detection and a sliding window, whether the position is a road anomaly is determined according to a random forest algorithm, the anomaly type of the road is determined according to a deformation clustering and support vector machine algorithm, and the anomaly type is returned; according to the method, the road abnormal section canbe completely intercepted, and the abnormal condition of the road can be more accurately detected on different data sets; the method provided by the invention is superior to some existing methods in three indexes of energy consumption, network delay and network consumption.
本发明涉及一种基于窗口划分以及形变聚类的道路异常检测方法,包括:步骤1、移动端对z轴加速度数据进行阈值检测和滑动窗口处理,筛选待确定的片段;步骤2、移动端通过随机森林算法判断道路片段是否为异常片段;步骤3、云端通过形变聚类的方法将道路异常片段的数据变形至统一长度。本发明的有益效果是:本发明利用阈值检测与滑动窗口确定可能的道路异常窗口,根据随机森林算法确定该处是否为道路异常,根据形变聚类与支持向量机的算法确定道路的异常类型并返回异常类型;本发明能够较为完整地截取道路异常段落,在不同的数据集上可以更加精确地检测出道路的异常状况;本发明在能源消耗、网络延迟和网络用量三个指标上优于现有的一些方法。
Road anomaly detection method based on window division and deformation clustering
一种基于窗口划分以及形变聚类的道路异常检测方法
13.10.2020
Patent
Elektronische Ressource
Chinesisch
Similarity Based Vehicle Trajectory Clustering and Anomaly Detection
British Library Conference Proceedings | 2005
|Road network brittleness evaluation method based on anomaly detection
Europäisches Patentamt | 2020
|