Design and Implementation of a Spatial Decision Support System for Land Use Suitability Assessment Based on Sugeno Integral and Imperialist Competitive Algorithm

Land use suitability assessment is a traditional problem and many researches have been undertaken to address this problem. The main reason why this problem is important is that experts want to consider all the aspects into account when they are trying to find the optimum location for a specific purpose. In other words, experts want to find the best place from all point of views such as environmental, ecological, economic and political aspects. Therefore, a decision support system is obligatory in order to facilitate decision making using mathematical models. On the other hand, due to the fact that a place is going to be chosen in this problem, GIS is a main science involved in this assessment. In this paper, a spatial decision support system is proposed using the integration of Sugeno integral and Imperialist Competitive Algorithm (ICA). Sugeno integral is able to aggregate alternative scores with respect to their interaction. In other decision making methods, it is assumed that the criteria are independent but it is against the real world situations. For example, in land use suitability assessment problem, some criteria such as land price and distance to major roads are not independent. Therefore, this study can improve spatial decision support systems by taking the impact of interaction among criteria into account. Sugeno integral operator uses fuzzy capacities instead of layer weights. Fuzzy capacities show the importance of each group of criteria for land suitability assessment. Furthermore, Sugeno integral can provide a number of numerical measures to indicate the importance of each criteria (Shapley value), the interaction among each set of criteria (interaction index) and the power of each criteria to veto the final decision (veto index). Shapley value is a parameter defined by game theory which indicates the power of each player in a game. In terms of decision making problem, Shapley index shows the importance of each criterion in the decision making process. A more important criterion has a higher impact on the results of the decision making. Interaction index shows how two players cooperate. If the two players have a positive cooperation they will make a better situation and if they have negative interaction the power of their coalition will be less than the power of each of them. In multiple criteria decision making, interaction index represents how two criteria interact. When the simultaneous satisfaction of two criteria is favourable, the interaction among them is positive and when the simultaneous satisfaction of the two criteria is not what the decision maker wants, it means that the two criteria have negative interaction. In this research imperialist competitive algorithm is applied to find the best values of fuzzy capacities that best describe the experts’ knowledge. In other words, a constrained optimization problem is solved here to compute the optimum value of fuzzy capacity for each set of criteria. ICA is selected because it is able to find the optimum value of a continuous function under constrains. The proposed SDSS is employed for land use suitability assessment for a new power plant. The results indicate that the method is highly suitable for modeling GIS-based decision making over interacting criteria. This model may be used in other areas of decision support systems with minor modifications.