Zonal Reserve Capacity Model Based On Variable Demand And Its Application

The essence of congestion in urban road is the unbalance between transportation supply and demand. Research on how to coordinate the development between transportation system and land use is the key to solve the problem of congestion. The NDP based on the concept of network reserve capacity provides an approach to analysis the constraint relationship among transportation network, transportation planning and land use development from the view of network capacity. Therefore, it is useful to propose efficient policy for transportation planning and land use management. This thesis has conducted the following researches.Firstly, the constraint mechanism of network reserve capacity to land use is analyzed through the approach of system dynamics. The deficiency of network reserve capacity on revealing the different effect of the land use development of each zone on transportation system is discussed. And then the concept of zonal reserve capacity is proposed. Moreover, the effect of variable demand on zonal reserve capacity is also analyzed. Secondly, the zonal reserve capacity bi-level model based on singly constraint variable demand is constructed. The lower level program describes the traveler's choice behavior of path and destination when the trip production rate of each zone is determined. The upper level objective function is created to maximize network total demand, which is represented by zonal reserve capacity. And the signal control and link capacity constraints are considered. Thirdly, in order to analyze the effect of transportation planning policy on zonal reserve capacity, the NDP of link capacity improvement is embedded into the above reserve capacity model by considering the discrete link capacity expansion variable. Fourthly, both the traveler's choice behavior of path and destination is considered on the situation when both the trip production and attraction rate of each zone is determined. And based on the above reserve capacity model, the zonal reserve capacity bi-level model based on doubly constraint variable demand can be constructed. Lastly, the genetic algorithm and double stage algorithm or convex combination algorithm are applied to solve these models. The model applications in coordinating land use and transportation planning are illustrated by a simple network.The models proposed in this thesis can be useful as a planning tool for land use management. It can predict how much additional trip generation for each zone can be accommodated by road network, as well as the corresponding signal control strategy or link capacity expansion scheme, and hence efficient policy for land use development control can be proposed.