A genetic programming/NARMAX approach to nonlinear system identification
Katya Rodríguez‐Vázquez · 1997
The tree-based coding structure of genetic programming is explored as an alternative to the more usual linear coding structure of conventional genetic algorithms for use in nonlinear system identification. Based on nonlinear polynomial model structures and the concept of the NARMAX model, genetic programming is shown to be effective in the selection of appropriate model structures.