基于复杂系统演化优化的实码多目标嵌套加速遗传算法

复杂性科学研究已经成为系统科学领域研究的重要方向,然而由于复杂系统的多层次、非线性等特点, 加之含有大量随机不确定因素,给系统的演化仿真工作带来巨大的困难,提出基于实数编码的多目标嵌套加速遗传算法(RMONAGA)来对复杂系统进行演化仿真,利用复杂适应系统自身特点与遗传算法特点的相似性,采用嵌套、多目标等方法较好的解决了复杂系统中不同层次交互、不同部门间适应性矛盾等问题.将该方法应用于三江平原区域农业水土资源复杂适应性优化配置演化研究中, 验证了该方法的科学合理性,以期对CAS理论在各领域的建模仿真工作提供一定的帮助.

Real coded multi-objective nested accelerating genetic algorithm based on complex system evolutionary optimization

The research of complex system science has became an very essential part of the system science. However, due to it's multi-level, nonlinear, and coupled with a large number of random uncertainties in the evolution of simulation work to the system caused immense difficulties. On this background, the real coded multi-objective nested accelerating genetic algorithm (RMONAGA) has been proposed to evolve and simulate the complex system. Based on the similarity between the features of system and characteristics of this algorithm, the problem of interaction of different levels and the adaptive conflict between different sectors has been solved by using the nested and multi-objective method. The method is applied to do the research of regional agricultural soil and water resources complex adaptive system optimization of Sanjiang Plain. The result shows that this method is scientifically rational, in order to provide some help in the field of modeling and simulation when CAS theory have been applied.