摘要
基于很多条模糊控制规则的模糊逻辑推理软件化实现方法有时难以满足快时变过程控制的要求,而基于部分规则的模糊推理硬件化实现方法则具有非常高的推理速度.提出了一种基于作用模糊子集推理的数/模混合推理结构,讨论了其实现原理,并用HSPICE完成了电路的仿真试验.结果表明,当规则数增加时,具有此结构的模糊控制器的器件数目增加很少,大大节省芯片面积,而且具有纳秒级的推理速度.
In real-time fuzzy control system with fast variant time, it is sometimes very difficult for the fuzzy logic reasoning by software, which is based on a large number of fuzzy control rules, to meet the demand of the controlled process with fast variant time. On the contrary, fuzzy logic reasoning by hardware possesses very high reasoning speed and maybe requires only a few fuzzy control rules. An analogue/digital hybrid inference construction based on functioning-fuzzy-subset inference has been developed and its principle of hardware implementation has been discussed as well. In the end, the circuit diagram in principle of the analogue/digital hybrid inference construction has been presented. This construction is more befitting for the multi-input system and the system with a large number of fuzzy control rules, and it not only has ns level reasoning speed, but can save chip area largely.
出处
《哈尔滨工业大学学报》
EI
CAS
CSCD
北大核心
2006年第4期563-566,共4页
Journal of Harbin Institute of Technology
基金
国家自然科学基金资助项目(59908001)
哈尔滨工业大学跨学科交叉性研究基金资助项目(HIT.MD200030)
关键词
模糊推理
数/模混合推理结构
模糊芯片
电路设计
fuzzy inference
analogue/digital hybrid inference construction
fuzzy chip
circuit design
