Neural-Network-Based Fuzzy Logic Control and Decision System
IEEE Transactions on Computers - Special issue on artificial neural networks
Adaptive fuzzy systems and control: design and stability analysis
Adaptive fuzzy systems and control: design and stability analysis
Based on embedded database greenhouse temperature and humidity intelligent control system
WSEAS Transactions on Circuits and Systems
| Hi-index | 0.01 |
A hybrid neuro-fuzzy approach called the NUFZYsystem, which embeds fuzzy reasoning into atriple-layered network structure, has been developedto identify nonlinear systems. A set of membershipfunctions at the input layer is partially linked witha layer of rules, using pre-set parameters. By meansof a simplified centroid of gravity defuzzificationmethod, the output becomes linear in the weights.Therefore, very fast estimation of the weightparameters can be achieved by using the orthogonalleast squares (OLS) method, which also provides amethod to efficiently remove the redundant fuzzy rulesfrom the prototype rule base of the NUFZY system. Inthis paper, the NUFZY system is applied to identifylettuce growth and greenhouse temperature from realexperimental data.Results show that the NUFZY model with the fast OLStraining can perform quite well in predicting bothlettuce growth and greenhouse temperature. In contrastto the mechanistic modeling procedures, theneuro-fuzzy approach offers an easier route and a fastway to build the nonlinear mapping of inputs andoutputs. In addition, the resulting internal networkstructure of the NUFZY system is a self-explanatoryrepresentation of fuzzy rules. Under this frame, it isa perspective that one is able to incorporate thehuman knowledge in this approach, and, hopefully, todeduce any interpretable rules that describe thesystems‘ behavior.