基于S-G滤波的交通流故障数据识别与修复算法

Detection and repair algorithm of traffic erroneous data based on S-G filtering

为了提高检测器采集到的交通流数据的有效性,基于S-G滤波法和数据驱动方法,提出一套交通流故障数据识别与修复的算法。以北京市某路段上单个交通流检测器所收集的数据为基础,结合实时数据与历史数据,完成故障数据识别与修复过程。与传统方法相比,该方法整合了交通流故障数据识别与故障数据修复两部分功能,运算更加高效而可靠。通过对北京市某路段实测数据的处理实例对算法进行验证。结果表明该方法实用性较高、运算速度快,为异常数据处理提出了新的思路。

In order to improve the effectiveness of the detector to collect traffic flow data,an algorithm was proposed to identify and repair the erroneous data based on S-G filtering and data-driven method.Based on the data of a single traffic flow detector collected from a certain road in Beijing,the real-time data was combined with historical data,and then the detection and repair processes of erroneous data were completed.Compared with the traditional method,the identification and the repair of traffic erroneous data are integrated in the algorithm so that the implementation is more efficient and reliable.By taking the data collected from a certain road in Beijing as example,it was shown that the algorithm has good practical application and fast computational speed,which provides a new idea for erroneous data processing.