Pareto-Optimal Sustainable Transportation Network Design under Spatial Queuing.

Huang, Wei; Xu, Guangming; Lo, Hong K.

This paper presents a multi-objective network design problem with environmental considerations for urban networks with queues. A spatial queuing link model is introduced to take account of the spatial effect of queuing. With this more realistic link performance function capturing spatial queuing, the network equilibrium flow patterns can be more accurately identified. Furthermore, to better estimate vehicle emissions, this paper proposes a refined emission estimation model, which distinguishes between travel speeds in free-running state and queue-forming state over a link. A multi-objective bi-level programming is then developed, in which the upper-level problem optimizes the investment decisions, whereas the lower-level problem characterizes the user equilibrium with spatial queuing delays. The metaheuristic of non-dominated sorting genetic algorithm II (NSGA-II) is adopted to solve the multi-objective network design problem. Numerical tests on the Sioux Falls network and the Barcelona network confirm the effectiveness of our proposed model and algorithm in identifying queuing equilibrium flows and Pareto optimal solutions. The refined models and valuable information about trade-offs among objectives are particularly helpful for environmentally sustainable transport network planning.

BARCELONA (Spain); SIOUX Falls (S.D.); SUSTAINABLE transportation; BILEVEL programming; METAHEURISTIC algorithms; GENETIC algorithms; ALGORITHMS; INFORMATION modeling; EQUILIBRIUM

Networks & Spatial Economics, 2020, Vol 20, Issue 3, p637

1566-113X

Academic Journal

10.1007/s11067-020-09494-6