With emerging new technologies, the vehicles in the future with connected and automated vehicles (CAVs) will be equipped with the ability to communicate with each other and aim to provide drivers with information in a timely and reliable way to improve driver safety. This study is designed to explore the application and effect of CAVs in reducing vehicle head-on and angle crashes in merging area of freeway. This study proposed an intention-oriented autonomous decision-making strategy for avoiding crashes between on-ramp vehicles and main-line vehicles in merging area of freeway. Firstly, an interactive model of driving intention was constructed based on cooperative game theory to determine the adjustment sequence of conflict vehicles. Secondly, the rule and algorithm for driving behavior were proposed based on probability reduction method. Finally, a series of scenarios were designed and simulated on the MATLAB platform to estimate the outcome of the above methods. The analysis results show that the proposed strategy can effectively reduce vehicle head-on and angle crashes, the conflict-resolution cost was reduced by 11% compared with the non-cooperative game state. In addition, we also found that the proposed strategy in this research can achieve autonomous resolution in different scenarios.

CICTP