一种自学习专家控制器

对一类系统提出了一种两级产生式自学习专家控制器,该控制器依据一种较为完善的特征模型,将控制理论的定性知识应用于专家控制和自学习控制之中,从而使该自学习专家控制器具有良好的控制性能和适应性,仿真结果证实了这一点.文中还对两级控制分别给出了两种高效的推理方法,增强了控制的实时性。

A Self-Learning Expert Controller

Presents a two-level production, self-learning expert controller for a certain kind of systems. Based on a feature model that is closer to perfectness, the controller applies the control theory to intelligent and self-learning controls. The resulting self-learning expert controller possesses good control performance and adaptivity, as has been confirmed by simulation results. Two kinds of effective inference are proposed for each level, to improve the real-time properties of the controller.