摘要
对传统模糊控制的双重特性进行了深入探讨．定性和定量分析都显示出模糊控制 集中了相对应的线性控制和变结构滑模控制的特性．该双重特性可以通过改变属性函数和增 益来调整．高的增益和多个属性函数会导致强的变结构控制和弱的线性控制，反之亦然．仿 真结果证实了以上分析．困变结构控制更多地用于较复杂过程，而线性控制则用于较简单过 程，故独特的双重特性对模糊控制的设计和调整影响重大．对于简单过程，模糊控制的初始 增益可以用已调好的线性控制参数来设计；对于复杂过程，可用变结构控制理论来设计．模 糊控制的双重特性有助于解释其对复杂过程有较佳的鲁棒性．
Fuzzy logic control (FLC) based on tracking error feedback is called conventional FLC in this paper. A thorough study of the conventional FLC on its dual features is presented. Both linguistic and numerical analyses show that conventional FLC combines properties of its linear control and variable structure control (VSC) components. These features can be adjusted by tuning the scaling gains and membership functions (MFs). Using large gains or more MFs result in stronger VSC effect but less linear control effect, and vice versa. Simulation results have confirmed this analysis. As VSC is usually advantageous for complex processes while linear control is better for simpler ones, the influence of these dual features on FLC design and tuning is significant. Thus, initial gains of FLC are suggested to be designed based on the well-tuned linear control structure for a simple process, and they can be determined based on the VSC theory for a complex process. These dual features can also help understand the robustness of the conventional FLC systems.
出处
《自动化学报》
EI
CSCD
北大核心
2001年第4期447-459,共13页
Acta Automatica Sinica
基金
the RGC grant (9040507) from the Hong Kong Goverment.
