利用生物地理学优化算法获取城市扩展元胞自动机模型参数

提出了一种基于生物地理学优化算法寻找城市扩展元胞自动机（cellular automata，CA）模型最佳参数的方法。转换规则制定及相应权重参数获取是构建城市扩展CA的核心和难点。生物地理学优化算法（biogeography-based optimization，BBO）通过模拟生物物种在栖息地的分布、迁移和灭绝来求解优化问题。利用BBO算法自动获取城市扩展CA模型参数值，构建BBO-CA模型进行城市扩展模拟实验，并与粒子群算法（particle swarm optimization，PSO）、蚁群算法（ant colony optimization，ACO）、遗传算法（genetic algorithm，GA）及逻辑回归（logistic regression，LR）等方法相比较。结果表明，BBO算法具有较好的收敛性，可有效地快速自动寻找城市扩展CA模型最佳参数组合，获取的空间变量权重参数较为合理；BBO-CA模型明显提升了城市用地模拟精度，城市用地模拟精度为72.5%，相对PSO、ACO、GA、LR各算法分别提升了1.1%、1.2%、2.7%和4.0%，Kappa系数达到0.700，分别提升了0.015、0.016、0.034和0.046，且整体空间布局与实际情况更为接近，验证了应用BBO算法的可行性与优势。

Calibrating Urban Expansion Cellular Automata Using Biogeography-Based Optimization

A new method is presented in this paper using biogeography-based optimization to calibrate urban expansion cellular automata (CA). Determining the transition rules and corresponding parameters is the key to a CA model. Biogeography-based optimization (BBO), is a new intelligent bionic optimization algorithm, solving problems by simulating the distribution, migration, and extinction of biological species. In this paper, a BBO algorithm is used to obtain transition rules and parameter values, and construct a BBO-CA model to simulate urban expansion. Compared with particle swarm optimization (PSO), the ant colony algorithm (ACO), genetic algorithm (GA), and logistic regression (LR), the BBO algorithm can effectively and quickly yield optimal and reasonable parameters. BBO performs effectively in terms of convergence and stability, with greater accuracy for urban cells and visual spatial layouts of simulation results. This paper illustrates the novel capabilities of the BBO algorithm for acquisition of variable parameters for urban cellular automata and has potential for simulations of other urban geographic phenomena.