Bicriterion traffic assignment: Efficient algorithms plus examples

This paper describes the mathematical properties and solution algorithm of a stochastic bicriterion equilibrium traffic assignment model in which both criteria may be flow-dependent. Formulated as a variational inequality, the model can reflect that different trip makers – due to habit, taste, or information – react differently to various factors such as travel time, out-of-pocket cost, and traffic congestion. Combining a novel tree builder/trip loader that obviates path enumeration together with Newton's method and linear programming, its reduced simplicial decomposition solution algorithm is fast enough to make the model a practical planning tool.