Simulation on optimized allocation of land resource based on DE-CA model

Optimization of land resource allocation is a research hot spot in land science. In this study, a new differential evolution-cellular automata (DE-CA) model was established. With this model, the quantitative structure of land-use was optimized with differential evolution algorithm (DEA). The results obtained were then used as the quantitative restraint condition of CA model to optimize the land-use spatial patterns. The application of this model successfully achieved the proper combination of land-use quantitative structure with the land-use spatial pattern, which overcame the problem in previous studies with particular emphasis on either optimization of land-use quantitative structure or optimization of spatial patterns in the field of optimization of land resource allocation. Finally, we applied the established DE-CA model to optimize the allocation of land resources for the year 2010 and 2020 based on 2005 and 2010, respectively, in Dawa County and Liaoning Province in northeast of China. The accuracy and reasonability of the optimized results were analyzed and assessed. The results showed that the overall accuracy of the optimized results was 84.56% with Kappa coefficient of 0.7860, indicating the good performance of the established DE-CA model. Furthermore, the simulated scheme was shown to be consistent with the real situation. Thus, this model can provide the references for formulation of the regional land-use planning and provide scientific basis for the substantial utilization of land.