Abstract
Fuzzy Model Predictive Control (FMPC) algorithms presented here are model-based control schemes in which the models used for prediction are Takagi-Sugeno fuzzy systems (TSFS). Three approaches to FMPC design are discussed. The fuzzy model in the first approach can be represented as a time-varying affine model that is used for control. In the second approach, the fuzzy system is a convex combination of multiple affine models, where the control is a convex combination of multiple controllers. Lastly, the control of the third algorithm is obtained when only the model with the highest certainty is used in the design. Also, we extend the idea to have an adaptive controller for the first algorithm, where the parameters of the fuzzy model are updated online.
| Original language | English |
|---|---|
| Pages | 423-428 |
| Number of pages | 6 |
| DOIs | |
| Publication status | Published - 1999 |
| Externally published | Yes |
| Event | Proceedings of the 1999 IEEE International Symposium on Intelligent Control - Intelligent Systems and Semiotics - Cambridge, MA, USA Duration: 15 Sept 1999 → 17 Sept 1999 |
Conference
| Conference | Proceedings of the 1999 IEEE International Symposium on Intelligent Control - Intelligent Systems and Semiotics |
|---|---|
| City | Cambridge, MA, USA |
| Period | 15/09/99 → 17/09/99 |