Search

In geographical analysis, spatial simulation and optimization are usually separate processes tackling different problems. It is, however, increasingly necessary to integrate them. Particularly in a fast developing area, the development to be simulated is seldom inertial (i.e., strictly following the historical trend); instead, it is likely to be interfered by new planning measures. Meanwhile, in such an area an optimization plan might not be even meaningful if it only addresses a snapshot of a highly dynamic landscape. In this study, we explored the possibility of integrating cellular automata (CA), a widely used method for simulating urban development and land use changes, and ant colony optimization (ACO), an advanced technique for solving complex path optimization problems. We named the resulting integrated system the geographical simulation and optimization system (GeoSOS) and applied it to a case study concerning finding the optimal path for a planned expressway in Dongguan, a fast-growing city in one of the most economically active regions of China. In the case study, the CA component of the GeoSOS generated simulations of the industrial land use changes for some years in the next decade. The ACO component of the GeoSOS, which had been revised from the conventional ACO to work on raster surfaces, took the simulations as input and completed raster-based path optimizations. In terms of the cumulative utility, a measurement used to evaluate the performance of the optimization, the coupling method surpasses the noncoupling method by 10.3 percent. [ABSTRACT FROM AUTHOR]