Modeling and optimization of intelligent decision-making for shared parking based on evolutionary game theory

Shared parking is an intelligent solution to the urban parking problem. This paper constructs an evolutionary game model involving three parties: the shared parking platform, travelers, and parking space owners. Based on bounded rationality and prospect theory, the dynamic evolutionary mechanism of multi-agent strategy selection is analyzed. MATLAB simulation results show that the optimal strategies for the system are platform subsidies, travelers using shared parking spaces, and parking space owners renting out their spaces. The level of subsidies is positively correlated with the willingness of participants (the platform's subsidy ratio to travelers/parking space owners should be controlled around 0.2/0.6). Time costs, rental prices, and potential losses are negatively correlated with participation willingness (the potential loss for parking space owners should be ≤7). The revenue distribution coefficient is positively correlated with rental willingness, and the platform is most sensitive to perceived value (the sensitivity to perceived gain/loss should be controlled around 0.6/2.25). This research provides a theoretical basis for the intelligent decision-making system of shared parking based on computational intelligence.