基于粗糙集理论的仿人控制器设计

Design of imitating human controller based on rough set theory

针对如何模拟人的控制行为以及提取控制行为特性问题,提出了基于粗糙集理论的仿人控制器的规则提取方法和仿人控制器的设计方法,包括用人控制行为的样本集建立初始决策信息表、信息表离散化、属性约简、规则提取和决策表完备化等。实验结果表明:将所设计的仿人控制器应用到原有的实际系统中,能顺利实现规则校验,提出的仿人控制器的设计方法是可行的。

To model the behavior of human operator and to extract the characteristics of control-behavior of human operator, a rule extraction and design method for imitating human controller （IHC） is proposed in this paper based on rough set theory （RST）. It includes the construction of the primal decision information table based on the sample data from the control behavior of a human operator. The experimental results demonstrate that the rule extraction can smoothly be achieved when the IHC designed is applied in the intrinsic real systems. Therefore, the design method for IHC proposed is feasible and of high value to the study on how to imitate human operator＇s behavior.