Naim Ajlouni
Faculty of Computer Science and Information, Applied Science University, 11931, Amman-Jordan
Sadeq Al-Hamouz
Faculty of Computer Science and Information, Applied Science University, 11931, Amman-Jordan
ABSTRACT
Evolutionary techniques are proposed in a new and novel paradigm to solve the problem of designing a robust neural PID plus feed forward controller for a plant with prescribed plant parameter uncertainties. The evolutionary scheme used, involves generating two separate populations, one representing the controller and the other the plant. The controller population is then evolved against a fixed population of plants representing the uncertainty space, such that the controller can control all these plants effectively. A cost function involving time-domain performance is then deployed, subject to a frequency domain stability constraint. The resulting paradigm results in a robust controller design with excellent time-domain performance. This evolutionary approach is illustrated by evolving a neural PID plus feed forward controller for a linear plant, which has a set of prescribed uncertainties.
PDF References Citation
How to cite this article
Naim Ajlouni and Sadeq Al-Hamouz, 2004. Genetic Design of Neural PID plus Feed Forward Controllers. Information Technology Journal, 3: 6-11.
DOI: 10.3923/itj.2004.6.11
URL: https://scialert.net/abstract/?doi=itj.2004.6.11
DOI: 10.3923/itj.2004.6.11
URL: https://scialert.net/abstract/?doi=itj.2004.6.11
REFERENCES
- Zames, G., 1981. Feedback and optimal sensitivity: Model reference transformation. Multiplicative Seminorms Control, 26: 301-302.
Direct Link - Hills, W.D., 1990. Co-evolving parasites improve simulated evolution as an optimization procedure. Phys. D, 42: 228-234.
Direct Link