Multi-Objective evolutionary algorithm for modeling of site suitability for health-care facilities

Introduction: To meet ever-increasing demands for health-care services, it is necessary to determine optimal locations for new health-care facilities. Method: A multi-objective model based on the genetic algorithm (GA) has been applied to evaluate site suitability for new clinics in part of Tehran urban areas. A multi-objective GA has been combined with a single GA to solve the location-allocation problem composed of the three objective functions. Geo-spatial Information System (GIS) has been used to prepare, analyze and visualize spatial data. Results: The results showed that optimizing of a single objective may result in unacceptable solutions with respect to other objectives. For example, the best solution resulting from optimization of the second objective function (proximity of the sites to the streets), was the worst one according to the first (travel cost) and third (land-use compatibility) objective functions. Therefore, 10 alternative solutions as the Pareto front in the objective area were indentified and investigated. Conclusion: Visualization of the best solutions for each objective and compromise between different objectives provide valuable possibilities for selection of the best alternative for decision makers.