Abstract
Urban Traffic Flow Guidance System (UTFGS) may cause the overreaction and concentration of travelers and lead to the phenomenon of Braess Paradox. The paper develops a constrained K-shortest paths algorithm within a dynamic restricted searching area in consideration of realistic road network spatial distribution features. Through the test with the microscopic transportation simulation tool VISSIM, the algorithm could not only decrease the searching scale but also efficiently balance the traffic flow in the net and avoid the phenomenon of Braess Paradox.