Datenbestand vom 10. Dezember 2024

Impressum Warenkorb Datenschutzhinweis Dissertationsdruck Dissertationsverlag Institutsreihen     Preisrechner

aktualisiert am 10. Dezember 2024

ISBN 9783843937382

72,00 € inkl. MwSt, zzgl. Versand


978-3-8439-3738-2, Reihe Regelungstechnik

Christoph Krimpmann
Sliding Mode Control of Mechatronic Systems - From robust control towards self-tuning control with experimental investigations on hydraulic valves

151 Seiten, Dissertation Technische Universität Dortmund (2018), Softcover, A5

Zusammenfassung / Abstract

The class of mechatronic systems bears many challenges in the design of control algorithms due to its interdisciplinary development processes because, e.g., mechanical friction, as well as magnetic hysteresis in actuators, have to be considered.

Conventionally used sliding mode controllers as a member of commonly used robust controllers generate linear closed loop dynamics in an ideal case. However, practical applications demand increased reaction rates which results in nonlinear closed-loop dynamics. To enhance the dynamics, this work discusses different approaches to formulate nonlinear closed-loop dynamics for sliding mode controllers and introduces the novel “successive pole adjustment” sliding surface that adjusts the dynamics at runtime. By applying Lyapunov’s direct method, the stability properties of the control law are investigated and extended, such that adaption mechanisms are deduced to adjust the control gains at runtime.

The evaluation of the discussed control strategies follows a two-stage process. To investigate the control structure, hardware-in-the-loop parameter optimizations are performed to optimize the piston position step responses of high-response directional valves as well as the pressure step responses of a pressure relief valve. In the second step, the properties of the parameter adaption are investigated to highlight the ease of applicability and that the long-lasting manual design- or optimization processes can be replaced by self-tuning approaches.