With the development of intelligent technology, how to effectively deal with complex traffic scenarios, especially emergencies and traffic congestion under high traffic flow, has become a key problem to be solved in the construction of intelligent high-speed. Aiming at the problem of highway traffic congestion, this paper proposes a dynamic control strategy based on traffic congestion model. Firstly, the evaluation model based on congestion index is constructed, and the evaluation accuracy is improved by optimizing the weight coefficient; Secondly, SARIMA model is used to predict frequent congestion, identify congestion time and area in advance and take preventive measures; For accidental congestion, a fast response algorithm combining data fusion and event analysis is designed to deal with accidents, construction, weather and other emergencies. Based on the above methods, a dynamic management and control strategy analysis framework is established, which can adjust the management and control measures in real time to achieve accurate and efficient congestion control. Through the application of the test section of the high-speed expansion project, the effectiveness of the strategy is verified, which provides a theoretical basis and technical support for the intelligent management and control of the smart highway.