Research on technical model and method of urban road traffic accident and traffic conflict based on artificial intelligence

With the rapid development of artificial intelligence, automatic control technology, computer technology and communication technology, various traffic models and comprehensive analysis methods are emerging one after another, and new theories and research results are constantly appearing, which have shown great power and development potential in practical engineering applications. In this paper, based on artificial intelligence, the accident-prone points are analyzed by using spatial clustering method and cumulative frequency curve method of DBSCAN (density-based spatial clustering of applications with noise), and the accident-prone points (or sections) in this road are found. Then, using the theory of neural network technology, the real-time decentralized and coordinated intelligent control strategy of urban traffic network is studied. Finally, the paper applies the model based on traffic conflict technology to the representative urban road signalized intersection, and evaluates its safety level, so as to improve the road service level of the intersection.