Abstract | ||
---|---|---|
This paper presents a satisfactory modeling method for data-driven fuzzy modeling problem based on support vector regression and Kalman filter algorithm. Support vector learning mechanism has been utilized to partition input data space to accomplish structure identification, then the complex model can be constructed by local linearization represented as T-S fuzzy model. For the ensuing parameter identification, we proceed with Kalman filter algorithm. Compared with previous works, the proposed approach guarantees the good accuracy and generalization capability especially in the few observations case. Numerical simulation results and comparisons with neuro-fuzzy method are discussed in order to assess the efficiency of the proposed approach. |
Year | DOI | Venue |
---|---|---|
2006 | 10.1007/11816157_162 | ICIC (1) |
Keywords | Field | DocType |
support vector learning,complex model,support vector,t-s fuzzy modeling,satisfactory modeling method,parameter identification,neuro-fuzzy method,structure identification,t-s fuzzy model,kalman filter algorithm,data-driven fuzzy modeling problem,numerical simulation,neuro fuzzy,kalman filter,support vector regression | Computer science,Support vector machine,Fuzzy logic,Kalman filter,Artificial intelligence,Fuzzy control system,System identification,Artificial neural network,Machine learning,Linearization,Fuzzy rule | Conference |
Volume | ISSN | ISBN |
4113 | 0302-9743 | 3-540-37271-7 |
Citations | PageRank | References |
1 | 0.35 | 9 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Wei Li | 1 | 28 | 5.33 |
Yupu Yang | 2 | 332 | 25.20 |
Zhong Yang | 3 | 5 | 1.06 |