复杂环境下多梯度中点位移中性景观模型

针对传统方法生成的景观模型无法对复杂环境下的真实景观进行全面模拟的问题,提出一种基于交互式遗传算法的多梯度中点位移中性景观模型。利用中点位移法形成分形布朗运动曲线,根据曲线衡量设置等间隔变化的空间自相关参数生成多梯度中性景观模型;对中点位移中性景观模型进行正规化处理,在处理的基础之上,利用适应度分享法避免种群过早收敛,最大程度保证种群多样性;选用比例选择,交叉和变异操作设计遗传算子,利用交互遗传算法获取符合中点位移中性景观模型条件的最优解。实验结果表明,所提模型能够全面模拟出复杂环境下的真实景观,其应用性较强。

Neutral landscape model with multi gradient neutral point under complex environment

In view of the problem that the landscape model generated by traditional methods can not fully simulate the real landscape in complex environment, a multi gradient neutral point displacement neutral landscape model based on interactive genetic algorithm is proposed. The fractal Brown motion curve is formed by using the middle point displacement method, and the multi gradient neutral landscape model is generated according to the spatial autocorrelation parameters set by the curve. The neutral landscape model is normalized and the fitness sharing method is used to avoid the premature convergence and the most premature convergence on the basis of the result. To ensure population diversity to a large extent, genetic operators are designed with proportional selection, crossover and mutation operation, and the optimal solution is obtained by using interactive genetic algorithm. The experimental results show that the proposed model can fully simulate the real landscape under complex environment, and its application is strong.