Model and Optimal Control for Urban Traffic Network under Incident Condition

Incident-induced traffic congestion is a critical issue in formulation urban traffic congestion problems. This paper presents a method which serves to alleviate incident-induced traffic congestion for urban traffic network. The proposed approach includes two major parts: models and optimal control algorithm. Based on the typical phase types in Chinese cities, calculation models for controlled flow at intersection exit lanes, for movement offsets, for delay and queue of each movement, and for delay and queue of links adjacent to incident location were formulated respectively. Movement was the basic unit when delay and queue were calculated. Then a real-code genetic algorithm was presented to implement the optimal control under incident condition, in which we used a penalty function to ensure constraints. In the preliminary test, eight optimization strategies were chosen for the network control. The result showed that the total delay could decrease approximately 20% even if no vehicle diversion was considered under incident condition.