Urban growth modeling with cellular automata and genetic algorithm

With the ever increasing of unplanned urbanization، the quality of citizen`s life are lessening. Understanding the process of urban growths facilitates Sustainable Urban Planning and Urban Management. The spatial patterns of urban modeling can make a useful perspective on how cities develop under different social، economic and environmental factors. So far، many efforts have been carried out in urban growth modeling using a cellular automat، but to create a reliable and credible model، some issues still remain unsolved. The complexity of the urban development process، the numerous variables and the variety of rules have caused the calibration of cellular automation models to be one of the challenging issues in the simulation of urban growth. In this study a new model for urban growth modeling is devised. The model integrates the advantages of cellular automation and genetic algorithm to predict Shiraz urban development between 1990 and 2000. Effective parameters in this process are urban neighborhoods، slope، distance to main roads and undeveloped areas. The chromosome`s contribution of genes to each parameter is then estimated. Optimum results have been exploiting، and finally the overall accuracy and Kappa index were used to evaluate the model. The overall accuracy and the final kappa index obtained were equal to 0. 91837 and 0. 68406 which indicate that 91. 837% of the cells were simulated true and the simulation results were 68. 406 % better than the random simulation.