Bidirectional detection and restoration method for traffic flow abnormal data based on KNN algorithm

Abstract

The invention relates to a bidirectional detection and restoration method for traffic flow abnormal data based on a KNN algorithm, and the method comprises the following steps: 1), obtaining normal traffic flow historical data, taking each five pieces of continuous data as a group, and building a historical data state vector library; 2), obtaining an abnormal value in a to-be-restored traffic flowdata, and marking the abnormal value as v(w); 3), constructing an abnormal data state vector X according to the abnormal value; 4), calculating the Euclidean distances d between the abnormal data state vector X and all historical data state vectors Xn, carrying out the optimization, and obtaining k groups of historical data state vectors after optimization and the corresponding Euclidean distances di (i=1, 2, ...k) 5), calculating a restoration value v(w)' according to the k groups of historical data state vectors after optimization and the corresponding Euclidean distances di; 6), carrying out the deletion, filling and restoration of the abnormal value. Compared with the prior art, the method is high in restoration precision, is high in applicability, improves the restoration precision,and improves the traffic data quality.

Description

technical field [0001] The invention relates to the field of intelligent traffic systems, in particular to a method for bidirectional detection and repair of abnormal traffic flow data based on KNN algorithm. Background technique [0002] Complete traffic flow data is the basis of traffic management and control. The vehicle detector is a device that detects road traffic flow operating parameters and is an important component of the intelligent transportation system. In actual traffic operation, due to abnormal road traffic environment, the detector Failure, communication failure and other reasons lead to abnormalities in the collected traffic data, which affects the quality of traffic data. The quality of traffic flow data directly affects the effect of estimation, prediction and evaluation of traffic status. Therefore, it is necessary to restore traffic anomaly data. [0003] The current restoration methods for abnormal traffic flow data mainly include historical average m...

AI Technical Summary

Problems solved by technology

The historical average method and the moving average method mainly use data to calculate the average value, and the repair accuracy is not high; the interpolation method is mainly used for repairing serious data loss, which has limitations; the data repair method based on time series, in the case of continuous In abnormal cases, the repair accuracy is greatly reduced

Images