This paper introduces an electrical drives control architecture combining a fractional-order controller and a set-point pre-filter. The former is based on a fractional-order proportional-integral (PI) unit, with a non-integer order integral action, while the latter can be of integer or non-integer type. To satisfy robustness and dynamic performance specifications, the feedback controller is designed by a loop-shaping technique in the frequency domain. In particular, optimality of the feedback system is pursued to achieve input-output tracking. The set-point pre-filter is designed by a dynamic inversion technique minimizing the difference between the ideal synthesized command signal (i.e. a smooth monotonic response) and the pre-filter step response. Experimental tests validate the methodology and compare the performance of the proposed architecture with well-established control schemes that employ the classical PI-based symmetrical optimum method with a smoothing pre-filter.
Synthesis of fractional-order PI controllers and fractional-order filters for industrial electrical drives / Lino, Paolo; Maione, Guido; Stasi, Silvio; Padula, F.; Visioli, A.. - In: IEEE/CAA JOURNAL OF AUTOMATICA SINICA. - ISSN 2329-9266. - STAMPA. - 4:1(2017), pp. 58-69. [10.1109/JAS.2017.7510325]
Synthesis of fractional-order PI controllers and fractional-order filters for industrial electrical drives
LINO, Paolo;MAIONE, Guido;STASI, Silvio;
2017-01-01
Abstract
This paper introduces an electrical drives control architecture combining a fractional-order controller and a set-point pre-filter. The former is based on a fractional-order proportional-integral (PI) unit, with a non-integer order integral action, while the latter can be of integer or non-integer type. To satisfy robustness and dynamic performance specifications, the feedback controller is designed by a loop-shaping technique in the frequency domain. In particular, optimality of the feedback system is pursued to achieve input-output tracking. The set-point pre-filter is designed by a dynamic inversion technique minimizing the difference between the ideal synthesized command signal (i.e. a smooth monotonic response) and the pre-filter step response. Experimental tests validate the methodology and compare the performance of the proposed architecture with well-established control schemes that employ the classical PI-based symmetrical optimum method with a smoothing pre-filter.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.