Traffic Incident Detection for Urban Arterial Road Based on Data Fusion and Learning Vector Quantization

It is important to make the urban arterial road unobstructed, so a useful method is needed to detect the occurrence of the traffic incident before the road becomes congested. On the basis of analyzing the features of the traffic flow on urban arterial road, the model of learning vector quantization (LVQ) neural network was introduced to identify the occurrence of a traffic incident on the arterial road. The input variables came from the loop detectors on the section and the probe vehicles with GPS. Then the data fusion technology was used with LVQ. With microscopic traffic simulation software VISSIM, the simulation experiment was applied to prove the validity of this model for traffic incident detection. The results showed that the data fusion could reduce the false alarm rate. Meanwhile, the results also showed that the LVQ model had a good detection capability in traffic incident detection for urban arterial road.