Land Use Allocation Optimization Using Advanced Geographic Information Analyzes

Space allocation of urban facilities is a critical field of decision making in urban planning. Supposing that Geographic Information (GI) provide us with more objective parameters for space allocation, Geographic Information Systems (GIS) is used by urban planners. But usage of GIS in urban planning is limited to simple analyses believing that complex spatial analyzes complicate the currently established approaches in urban planning. We propose that sticking to the basic rules of decision making in urban planning referring to the reality and providing the GI analyzes as syntactic sugar which complement decisions made by urban planners, enable us to use more complex GI analyzes. In this paper, we study the application of two complexes GI analyzes in urban space allocation. These analyzes are first population density surface generation using Pycnophylactic Estimation-Maximization (PEM) areal interpolation and second Adaptive Multiplicatively Weighted Voronoi Diagram (AMWVD) to derive service areas of urban facilities. The results show that these two GI analyzes optimize the urban planners' space allocation, while they do not suppress the established rules of urban planning. These analyzes are used to define the overpopulation level of primary schools estimating their current service area and comparing their estimated and expected level of service. The service areas of schools are estimated minimizing the proportional difference of the estimated and expected level of service, based on the population, proximity and safety of access rules that are used for space allocation of the primary schools. The minimization is carried out using linear and Simulated Annealing (SA) approaches. It is observed that linear optimization generates better results. This means that using the two complexes analyzes (PEM and AMWVD), the complexity of the space allocation of primary schools is not changed dramatically.