From single to many-objective PID controller design using particle swarm optimization

Proportional, integrative and derivative (PID) controllers are among the most used in industrial control applications. Classical PID controller design methodologies can be significantly improved by incorporating recent computational intelligence techniques. Two techniques based on particle swarm opt...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:International journal of control, automation, and systems 2017, Automation, and Systems, 15(2), , pp.918-932
Hauptverfasser: Freire, Hélio, Moura Oliveira, P. B., Solteiro Pires, E. J.
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Proportional, integrative and derivative (PID) controllers are among the most used in industrial control applications. Classical PID controller design methodologies can be significantly improved by incorporating recent computational intelligence techniques. Two techniques based on particle swarm optimization (PSO) algorithms are proposed to design PI-PID controllers. Both control design methodologies are directed to optimize PI-PID controller gains using two degrees-of-freedom control configurations, subjected to frequency domain robustness constraints. The first technique proposes a single-objective PSO algorithm, to sequentially design a two degrees-of-freedom control structure, considering the optimization of load disturbance rejection followed by set-point tracking optimization. The second technique proposes a many-objective PSO algorithm, to design a two degrees-of-freedom control structure, considering simultaneously, the optimization of four different design criteria. In the many-objective case, the control engineer may select the most adequate solution among the resulting optimal Pareto set. Simulation results are presented showing the effectiveness of the proposed PI-PID design techniques, in comparison with both classic and optimization based methods.
ISSN:1598-6446
2005-4092
DOI:10.1007/s12555-015-0271-0