Multi-scenarios Behavior Choice Model of Shared Parking in Residential Area

With the development of economics, the parking contradiction has become more serious in residential area. For optimization of the distribution of parking resource in residential area with limited land resources, the strategy of shared parking has been studied and adopted by more and more countries. The aim of this research is to determine the influence factors of selecting shared parking facilities in residential area and to increase the probability of choosing shared parking can increase with changing the characteristics of people or parking facilities. It will help to relieve parking contradiction and optimize parking resources in residential area. In order to acquire people’s parking behavior characteristics, the paper designed questionnaire with three scenarios which contain shared parking and non-shared parking facilities. By questionnaire data introduced, the discrete choice modeling was set up to investigate the variability of probability of choosing shared parking across individual characteristics, socioeconomic attributes, trip and parking attributes, desire of accepting or providing shared parking and parking attributes in scenarios. For simulating the nonlinear effects of variables on the target variable more accurately, the BP neural network was used to filter redundant attributes. According to the model estimation results, charging had a significant impact on the selecting shared parking facilities. It found that people prefer selecting shared parking, which is same as the analysis of questionnaire. This indicates that the BP neural network is used to filter factors optimally. Finally, for appealing to people parking in shared parking facilities, it is suggested that the strategy of decreasing charging in shared parking facilities should be adopted. And the paper explored details areas for future research.