Spatial urban land-use planning is a complex process, through which we aim to allocate suitable land-uses while taking into consideration multiple and conflicting objectives and constraints under certain spatial contexts. Landowners should be modeled as players that are able to interact with each other so as to seek their best land-uses while considering multiple objectives and constraints simultaneously. Game theory provides a tool for land-use planners to model and analyze such interactions.In this paper, spatial urban land-use planning is considered as a game to model local competitions between landowners, whose players (i.e. landowners) of which play to pick the most suitable land-use for themselves. The game is defined based on the objectives of consistency, dependency, suitability, compactness of land-uses, and land-use per capita demand. In this paper, three different scenarios are designed for the players. In the first scenario, the players are greedy and only accept the most compatible land-use. In the second scenario, conversely, the players are fully collaborative and care about other players’ payoff. In the third scenario, the players are first greedy, but when they cannot achieve an agreement with other players, they change their attitude to become gradually collaborative for reaching the Nash equilibrium (NE). Furthermore, the dissatisfaction index (DI), which represents the number of unsatisfied landowners with their current land-use, is defined to compare the different scenarios. The proposed model is tested in a district located in District 7 in Tehran (the capital city of Iran) with 2710 parcels.Results of the first scenario showed that, at the beginning of the game, 50 % of the landowners were not satisfied with their current land-uses, but after 50 iterations, about 100 landowners were dissatisfied with their land-use and this scenario was not able to reach the NE. Results of the second scenario indicated that, in order to reach an optimized layout, 325 parcels needed to be changed. Also, after reaching the NE in this scenario, values of the objective functions did not significantly improve. So, lowering the expectations of the players would not lead to appropriate results. The outcomes of the third scenario provided appropriate results, which could be achieved when the expectation levels of the players could be changed during the game. Furthermore, the NE was obtained among the players and the objective functions improved by 20 % on average in comparison with the other scenarios.Moreover, results of the scenarios were compared with the optimized layout obtained by a genetic algorithm (GA) using different parameter values. Results of the comparison also revealed that the urban layouts produced by game theory improved the objective function values obtained by the GA in about 10 % and improved the GA’s running time in more than 85 %, on average in this research.
Read full abstract