Research On Land Use Simulation Based On Cellular Automaton

With the rapid expansion of the city,great changes have taken place in the urban land use pattern.Accurate land use data and reasonable land use planning can provide useful reference for urban planning and land development.Meanwhile,simulating land use can predict urban land use status and make reasonable allocation to land use in the future,so it is necessary to simulate the land use in the future.Cellular Automata is a dynamic model that the time,space and state are discrete.The model can simulate the land use through the regional land use change to the global in the way of bottom to top from the angle of time and space.The key of the simulation is the determination of the transformation rules.The classification of rule mining based on Ant Colony Algorithm excavates the more objective transformation by searching many data.Therefore,it is reasonable and feasible to simulate land use using the space-time dynamics model established by combining the idea of Ant Colony Algorithm and the core of Cellular Automata and taking the GIS as a tool.This dissertation improves the classification of rule mining based on Ant Colony Algorithm and processes the acquired basic data by GIS.Then the improved algorithm and the processed data are used to excavate the transformation rules of land use automatically,and the transformation rule of mining is used as the transformation rule of Cellular Automaton to simulate land use.The precision is verified by the point-to-point test method.Finally,simulating the land use in 2020 and putting forward the planning proposal according to the simulation results.