Household Residential Location Choice Equilibrium Model Based on Reference-Dependent Theory

We focus on the location choice behavior of new residents in the urban system using the reference-dependent theory approach. Reference-dependent theory offers an alternative analytical framework for the study of residential location choice behavior. This problem is formulated as bilevel programming in which the upper-level programming model is of household residential choice behavior with a combination of the reference-dependent theory and a discrete choice model, whereas the lower-level is a multiclass stochastic user equilibrium model. Here, the heterogeneity of households/travelers and the stochastic character of traffic networks are carefully taken into account. The multiclass stochastic user equilibrium condition is formulated into an equivalent variational inequality problem. Furthermore, an iterative algorithm combined method of successive averages has been adopted to solve this bilevel model. In numerical experiments, we illustrate the effect of different highway toll strategies and different reference points on the urban spatial structure. Finally, some interesting conclusions are obtained.