共享停车泊位的分配-定价-收益分配机制

A mechanism that consists of allocation/pricing/revenue distribution was established to manage shared parking resources. First, an optimal allocation/pricing model of parking permits for shared parking lots from multiple lands and based on the users' preferences for different areas and parking time periods was established; especially, the optimal allocation model was determined to be a winner determination problem in a combinatorial auction (NP-hard problem). Next, using the Benders decomposition principle, duality theory, and new ascending proxy agent system, an evolutionary method to conduct the auction of parking permits was proposed. Finally, to arouse the enthusiasm of shared parking lot owners, a revenue distribution model based on the contributions for social surplus was determined using the Shapley value method. Through theoretical analysis and a numerical experiment, it was found that the proposed mechanisms can achieve the exact maximal social surplus, minimal competitive equilibrium price, and fair revenue distribution. The results show the following: ①Compared to a Vickrey-Clark-Groves (VCG) mechanism and a Leonard mechanism, the proposed evolutionary method not only solves the optimal allocation/pricing of parking permits exactly but also has a high computational efficiency in algorithm time complexity in the worst-case scenario; ②For each step of the proposed evolutionary method, the permits prices also have the strategy-proofness, although under the non-optimal allocation of parking permits, which lets the users report their preferences truthfully; ③The evolutionary method can achieve a high social surplus within a number of iterations; ④The proposed new ascending proxy agent system has polynomial-type time complexity in the worst-case scenario; ⑤Through the proposed allocation/pricing mechanism, the manager obtains the users' preferences and the social surplus under any combination of shared parking lot owners can be calculated, thus achieving the calculability of Shapley value method in practical applications.