Título: Synthesis and Tuning of Non -Conventional Structures for FOPID Controllers Using Genetic Programming
Autores: Humberto Ícaro Pinto Fontinele, Guilherme de Alencar Barreto & David Ciarlini Chagas Freitas
Resumo: This paper presents a method for self-configuration and optimal tuning of parameters for PID controllers, including fractional-order PID controllers (FOPID), based on the evolutionary mechanism of Genetic Programming (GP). This method enables the synthesis of non-conventional structures and the optimization of the tuning of these controllers by automatically adjusting the gains and fractional exponents of the derivative and integral terms, based on time-domain performance indices. The results obtained demonstrate that the high flexibility in gain and phase margin adjustments, an inherent characteristic of controllers with high-order and fractional structures, when properly tuned, allows for effective performance in the control of complex systems, such as high-order plants and unstable systems. A comprehensive evaluation over various types of plants is carried out, with the proposed approach exhibiting overall superior performance compared to the two SOTA methods for all analyzed plants.
Palavras-chave: genetic programming; non-conventional PID; optimal tuning; self-configurable PID.
Páginas: 8
Código DOI: 10.21528/CBIC2025-1190012
Artigo em PDF: CBIC_2025_paper1190012.pdf
Arquivo BibTeX:
CBIC_2025_1190012.bib