Minimum rule-based parallel structure fuzzy controller

One of the most important issues in fuzzy logic controller design is the number of fuzzy subsets for inputs and outputs. The number of fuzzy terms should be big enough to achieve sufficient approximation and must be small enough to save the space of memory. There is a trade off between approximation and memory space (algorithm cycle time). This paper proposes a simple parallel structure of fuzzy controllers where simulation is performed on a real temperature model. It can be shown that the proposed parallel structure fuzzy controllers has the same performance and stability compared with the conventional fuzzy controller, with an advantage of significantly lower number of rules. This can save memory space and reduces cycle time which is an important element in practical implementation problems