Automatic incident detection method based on under-sampling and used for unbalanced data set

Abstract

The invention discloses an automatic incident detection method based on under-sampling and used for an unbalanced data set. The automatic incident detection method comprises the steps of (1) using a maximum and minimum normalization method to carry out normalization processing on actually-measured traffic flow data, carrying out under-sampling processing on a majority class in a training set on the basis of a neighborhood cleaning rule to obtain a new training set which is relatively balanced, (2) selecting a radial basis function as a kernel function of a support vector machine, using an improved grid search algorithm to optimize a penalty factor C and a kernel parameter g of the support vector machine, and (3) training the support vector machine through the training set which is relatively balanced so as to obtain an automatic incident detection model used for the unbalanced data set. According to the automatic incident detection method based on under-sampling and used for the unbalanced data set, the problem that an existing traffic incident detection algorithm is not applicable to unbalanced traffic data in reality is solved, detection performance of the traffic incident detection algorithm is remarkably improved, the average detection time is shortened, and the requirement of traffic incident detection for real-time performance is met.

Description

technical field [0001] The invention belongs to the technical field of traffic intelligent management and control, and relates to an automatic detection method for traffic events based on under-sampling and facing unbalanced data sets. Background technique [0002] Traffic incidents not only cause congestion and delays, but also easily lead to secondary accidents. Accurate and rapid detection of traffic incidents, timely rescue and treatment of incidents can effectively reduce traffic congestion and delays caused by traffic incidents, and avoid secondary accidents. Automatic Incident Detection (AID) is an important part of modern traffic monitoring system and the basis of advanced traffic management system and traveler information system. Road traffic safety and service levels are of paramount importance. [0003] In recent years, the research of AID algorithm mainly focuses on the application of new technologies such as neural network, fuzzy theory, wavelet analysis and s...

AI Technical Summary

Problems solved by technology

However, in the real world, there are far more traffic normal operating states than traffic event states, and traffic event detection is essentially an imbalanced classification problem, which is rarely considered in previous automatic traffic event detection algorithms.

Most of the above-mentioned traffic incident detection algorithms are classified based on balanced data sets, which often lead to high false alarm rate, low detection rate and long average detection time when used in traffic incident detection, and the detection effect is disappointing.

[0004] Support vector machine (SupportVectorMachine, SVM) has been used in traffic incident detection, but it shows obvious "bias" when dealing with imbalanced classification problems, which is not conducive to the learning of minority samples

Images