Bi-level Programming Model And Algorithm Of Urban Roads Convertible Lanes

Abstract:A kind of tidal phenomenon occurs in the congested traffic flow in some cities, it happens during the morning hours commuting from the suburbs to the city center, and the rush hours from the city center to the suburbs. On the symmetrical distributed urban roads, the asymmetrical traffic tidal results in unbalanced circumstance traffic flow, even leads to traffic congestion on single side of the road. So, setting up variable lane is an effective method of rationally use of road resources and tackling the problem of unbalanced bidirectional traffic capacity on both sides of the road.But, as people’s commuting trips usually contains some other destinations, the travel demand during morning and evening peak does not match precisely, that is to say, the congested roads differ between morning and evening peak. So this phenomenon directly result to the asymmetry of the optimal variable lane establishment during morning and evening peak.Considering the daily operating costs and how well drivers know about the road conditions, this paper comprehensively overlook the setting of variable lane during morning and evening peak, as to minimize the total establishment costs. The variable lane settings option is provided for the interaction between traffic operators and travelers. I use a primary-secondary method with bi-level programming model to describe the issues of variable lane optimization design. The upper level plan is designed to optimize the variable lane setting program, aiming to minimize the management cost of variable lanes and the total travel cost; the two lower level plans are designed to identify the balanced statue of traveling during morning and evening peak, through the given set of variable lane planning. Based on the genetic algorithm, this paper further designed the algorithm of variable lane optimization design using primary-secondary method with bi-level programming model.Numerical examples indicate that the optimized variable lane scheme shows strong overall rationality, and can effectively reduce the total cost of travel within a certain cost of the integrated management.