A Simulation of Land Use Change in Hamadan in 2040 Using Cellular Automation and Data Mining Method Support Vector Vectorization and Fuzzy Automated Cells

Introduction The development of the physical growth and population growth of cities has been on a steep rise over the past three decades. With new developments, cities were quickly adapted to changes, and the expansion of the service sector, and the concentration of industries and factories in cities have attracted people from many villages and small cities to major cities. Not only has these changes caused the disparity and dissonance of the physical expansion of the cities, but it also has greatly impacted the spatial development of cities. Cities may vary from a variety of perspectives, but contrary to these differences, cities have several features that make them similar. The dynamics and growth are two bases that exist in most cities. However, modeling of dynamism and growth, and models that involve the complexity of cities are very difficult. Therefore, these complexities make it harder to use old models for modeling because they are static, linear, cumulative, interconnected, and based on the theory of simple systems. So, in order to model urban systems, new methods should be employed that are dynamic, nonlinear, non-cumulative, discrete, and low-end. Therefore, this research aims to predict the future development of Hamadan city by employing a method that evaluates urban variables in a nonlinear way.
Methodology Cities are among the most complex structures which are built through human societies. A cellular automaton is widely used for the simulation of complex nonlinear systems. In this article, we first explain the structural and the general principles of CA and then use fuzzy logic as a method of transforming rules of the CA model. These logic rules are desirable in dealing with complex non-linear relationships. An urban cellular automation is designed to model urban and regional systems that include simulation of urban projects such as urban development simulation, simulation of land use change, and so on. This is a proposed model for using simulation of user variations. The space-spatial variables used in the simulation of the geographic database are extracted and the variables of the change of use of satellite imagery will be obtained. Once the cell's position is determined in the next period, the cells are compared one by one. The process of performing this method consists of two main stages of classification and simulation. In the first step, linear categorizers are used to make decision making using vector regression. In the simulation stage, according to different variables, the probability of land use changes in the city will be predicted.
Results The present study showed the extent, and quality of land use in the city by 2040, and then the future development of the city of Hamedan was simulated and predicted for the years to come. The simulation outcomes in the form of raster maps as well as the calculation of the city's physical development in 2040 using statistical methods revealed that if land use changes continued with this trend, the area of land built up by 2040 would increase and this area would probably approximately reach 6350 hectares, and the area of the lands of the garden and the tree will be reduced further. According to the calculations, the physical development of the city of Hamedan has been doubled over the past thirty years.
Conclusion According to available statistics, the city of Hamadan has experienced a growing trend over the last 50 years. The city of Hamedan, due to its location, has many restrictions, however, the population of the city is increasing as a result of the urban area of Hamedan. The simulation of land use change in Hamadan city was used and the results indicate a good correlation between real and simulated models. Therefore, it can be concluded that the cellular automation model, based on the rules of transmission of regression vector support, is a useful tool for simulating urban systems. The strengths of this approach can be seen in terms of the physical outlook, the flexibility of the program, real-life prospects, the lack of involvement of non-programmers in the program preparation process, the attention to urban problems, and the lack of emphasis on the production of planning standards, etc. Given that most of the factors affecting the formation and dynamics of cities are dynamic and gradual, changes are affected by the areas under study and may even change from city to city. Therefore, it is suggested that different parts of the city be considered for optimal parameters of cellular automation and compared with each other. This modeling helps designers, planners, managers and other researchers to be able to predict the status of cities and other land use changes in the future.