Emergency traffic control refers to the actions taken to maintain order and explores the challenges of road traffic control in emergency response scenarios, focusing on the selection of control efforts and addressing the ‘last mile’ problem of resource allocation. We propose a decision model that integrates Hierarchical Task Network (HTN) planning and Support Vector Machine (SVM) models for formulating comprehensive emergency traffic control plans. The SVM model, achieving a classification accuracy of 92.5%, is employed to analyze historical data and make decisions based on real-time conditions, environmental changes, and resource needs. The control objectives are encoded into the HTN planning framework, generating executable action plans. Experimental validation using emergency traffic control scenarios during the 2020 COVID-19 pandemic shows that the model improves emergency response efficiency by 10%, demonstrating its efficiency and flexibility in developing emergency traffic control plans.