Bi-level Programming Model for Shared Parking Considering Residential Parking Resources

In order to relieve the parking problem caused by limited parking resources, a shared parking guidance and service process with the participation of residential parking spaces was proposed, which was based on the sharing parking theory and intelligent traffic guidance strategy. First, the basic conditions of sharing parking with residential areas were discussed. Then, a bi-level programming model of sharing parking guidance was established. In the model, the goal of the upper-level programming is to minimize the mean value of unoccupied parking difference index (MUPDI), which is the index to measure whether the parking resources are available for balanced use; while the goal of the lower-level is to minimize the average walking distance after parking, which is the basis to evaluate the feasibility of a shared parking. Finally, based on the solving principle of the bi-level programming model and the particle swarm optimization algorithm, a nested optimization algorithm was proposed to solve the programming model, and a simulation-based case study was made to verify their validity. The simulation shows that although the minimum real value of the parking unoccupied rate is 0.301 in the Hexiangyuan residential area, the maximum ratio of the residential parking lots that need to be shared for balanced use in peak time is only 0.093. This means that using the sharing scheme, the parking problem of the Ocean City can be relieved by using only 1/3 the unoccupied parking lots in the residential area. In addition, the average walking distance of 160.59 m shows the feasibility of the sharing scheme.