Optimum Site Selection in Discrete Space Using Geospatial Information System and Fuzzy Inference Systems

Fair location for fire stations is a significant point to save and protect property and life's people. So, fair position of fire stations is one of the most important factors that could reduce damage caused by incidents. Therefore appropriate dispersal of fire stations must be created in the city. So, resolving optimum site selection problem for fire stations is the most problem that engaged municipal managers. Basically, optimum site selection problem for fire stations is a semi-structured and complex problem with great criteria and sometimes inconsistent. Geospatial Information System, as a tool for management, spatial data analyzing, is a powerful alternative to facilitate solving optimum site selection problems. GIS cant solve these problems, therefore it's fair to incorporate GIS to other method and applications to investigate and solve problems. Fuzzy system is one of the aforementioned methods that help decision makers in this condition. In fact fuzzy system has enough ability to implement expert's knowledge, and it's the most prominent property. Therefore combination of GIS and Fuzzy Inference Systems is able to solve spread extent of optimum site selection problems. In this paper, combination of GIS and two FIS, Takagi-Sugeno and Assilan-Mamdani, was implemented to determine optimum sites for fire stations and compare these systems together. Also efficiency of different Defuzzification methods and implication in Takagi-Sugeno and Assilan-Mamdani Fuzzy systems compared through TOPSIS and SAW (MCDM methods). Outputs implied that it is possible to use fuzzy methods instead of MCDMs and traditional methods in spatial-driven decisions and Takagi-Sugeno has more efficiency in compare with Assilan-Mamdani inference system (Coefficient correlation was 1 for Takagi-Sugeno and 0.96 for TOPSIS and SAW).