Incremental fuzzy modelling of dependencies in complex physical systems

Considers the fuzzy modelling of dependencies in complex physical systems. Recently, we have been working on a SISO (single-input, single-output) case and developed a successful model. In reality, we face more general cases than SISOs. The present step is to extend the SISO model to MISO (multiple-input, single-output) cases. In the SISO model, we consider the fact that a linear relation can be approximated by just two fuzzy rules. To implement this approach, the output domain of the function is partitioned into a finite number of regions with boundaries defined by local maxima and minima. In these regions, the function is monotonic and continuous. Each region is approximated by two fuzzy rules. The direct extension of the model to the MISO case is to consider that, for instance, a 3D linear relation can be approximated by three rules and, more generally, an n-dimensional relation is approximated by n rules.