With the rapid road development and the growing number of car ownership in China, road systems are facing unprecedented pressure and various issues. For example, road safety issue is severe for provincial highways due to improper design and lack of surveillance. In order to analyze the traffic safety of provincial highway, a fuzzy neural network model was established to predict various types of traffic accidents. Experiments are conducted based on accident data of a provincial highway in Hubei Province from 2010 to 2016, and the experiment result shows that the average prediction error is around 12%. The model can be applied to the micro-prediction of traffic accidents, providing an important scientific basis for similar road accident prediction and security measures. keywords: Provincial highway; traffic safety; fuzzy neural network; accident prediction

CICTP