Study On Freeway Automatic Incident Detection Algorithm Based On Information Fusion

With the rapid development of highway, the number of motor vehicles is sharply increasing, as a result, traffic accidents happen frequently. On the one hand, people enjoy rapidity, comfort and convenience provided by highway, on the other hand, they feel inconvenient on account of traffic accidents and traffic jam. It is an urgent demand to establish effective highway management information system under serious problem of traffic safety. Effective traffic detection is the first step of management information, simultaneously, it's a key. However, the current practical application effect of traffic incident detection is not ideal, so research on effective AID algorithms is the most problem need to be solved.As one recent information integration processing method, the multi-sensor information fusion technology has provided a new way for AID. Using the information fusion technical design highly effective AID method, not only can expand the application scope of the information fusion technology, but also can impel the realization of the highway safe, highly effective management goal. It has the high theory significance and the practical application value.After analyzing the current situation of freeway Automatic Incident Detection (AID) technic, combined with the hiberarchy and basic principle of multi-sensor information fusion technology, this article has discussed the traffic incident information fusion theory, and then designed the Automatic Incident Detection(AID) accord to the hiberarchy of information fusion and base on the multi-sensor information fusion technology which aim at the problems of simplex detection information.At the stage of data pretreatment, wavelet analysis method is adopted to make noise canceling of the data traffic flow in accordance with the processing steps of the method called AID based on information fusion. This paper takes wavelet neural network algorithm as the realized tool to the incident detection fusion algorithm of the module at the stage of multiplel information fusion method processing phase. Combined with wavelet analysis theory and the neural network theory, adopting highway dynamic traffic flow model, a traffic accident automatic detection algorithm based on wavelet neural network is established, which fully exert the wavelet analysis theory and the advantage of neural network theory. Then, take simulation of highway traffic incident detection technology based on information fusion by using MATLAB.Finally, in order to verification the effectiveness of the algorithm, test the AID algorithm of this article using the actual data of South Ring Road Xi'an. The results show that the algorithm based on information fusion have better detection rate, false alarm rate and mean time to detection, experiments show that this algorithm is better than other classical algorithms AID algorithm.