| Popis: |
Presents an effective and systematic design methodology for automatic synthesis of fuzzy systems, which are characterized by trapezoid-shaped membership junctions. Given some training patterns, the proposed design methodology is to satisfy the control performance requirement by iteratively adding new fuzzy rules and tuning their associated membership functions. The main advantage of our approach is that it exploits the inheritances of trapezoid-shaped membership functions. As a result, the new rules can be accurately determined to improve the control performance. The major contribution of our work is that it greatly shortens the design time of fuzzy systems. Benchmark data consistently shows that the proposed approach may automatically synthesize a reliable fuzzy system within only several hours. |
