基于车牌数据的行程速度特性及交通状态评估

Characteristics of Travel Speed and Traffic Status Assessment Based on License Plate Data

针对目前基于车牌的数据信息量大、覆盖范围广,但数据利用率低、资源浪费严重等问题,进行基于车牌的数据挖掘研究。分析已有数据处理方法及可能存在的问题,提出基于聚类思想的时间间隔获取方法,并对提取的行程时间数据进行异常判别和修复。利用相关方法计算平均行程速度,采用统计学分析方法研究了行程速度的时变特性,得到不同天相同时段平均行程速度的稳态变化规律。通过平均行程速度与速度标准差的量化关系研究,获得一天中不同时段行程速度的波动情况,最后设计以速度阈值为判断准则的交通状态评估算法,实例分析结果与实际交通状况一致,说明算法的科学性和有效性,可为交通管理部门提供可靠的决策依据。

Aiming at the problems of large amount of data information based on license plate,wide coverage,but low data utilization rate and serious waste of resources,conducting data mining research based on license plate.Analyzing existing data processing methods and possible problems,a time interval acquisition method based on clustering ideas is proposed,and abnormality discrimination and repair is performed on the extracted travel time data.Using the relevant method to calculate the average travel speed,studying the time-varying characteristics of the travel speed by statistical analysis method,obtaining the steady-state variation of the average travel speed in the same period of different days.Through the quantitative relationship between the average travel speed and the standard deviation of speed,the fluctuation of the travel speed at different times of the day is obtained.Finally,the traffic state evaluation algorithm based on the speed threshold value is designed.The results of the example analysis are consistent with the actual traffic operation,which shows the scientificity and effectiveness of the algorithm and provides real-time and reliable decision-making basis for traffic management departments.