Národní úložiště šedé literatury Nalezeno 1 záznamů.  Hledání trvalo 0.01 vteřin. 
Optimization of a Parallel Mechanism Design with Respect to a Stewart Platform Control Design
Březina, Lukáš ; Palčák,, František (oponent) ; Čech, Vladimír (oponent) ; Singule, Vladislav (oponent) ; Malenovský, Eduard (vedoucí práce)
The proposed work is dealing with an optimal model of a parallel manipulator dynamics for a control design purposes. The approach is based on modeling of the system dynamics in Matlab Simmechanics followed by the model linearization. The obtained linear model may be simply inspected from the controllability and observability point of view. It is also computational modest thanks to its simplicity. This is demonstrated on designing of a two – layer control for a model of a Stewart platform. The control based on such a linear model was successfully tested on the original nonlinear model. The essential part of the the work is dealing with modeling of uncertain parameters in the dynamic model of the Stewart platform and DC motor Maxon RE 35. The proposed approach is based on modeling of a parametric uncertainty where the uncertainty is defined individually for particular elements of the model state matrices. The uncertainty itself is set by the difference between parameters of corresponding matrices of the nominal linear model and model with maximally perturbed parameters. The obtained uncertain model is for its form suitable for the robust control design methods, for example via minimizing an H-infinity norm. The method was used for a compensation of the shifting of the linearization operating points in the Stewart platform and for compensation of the modeling inaccuracy of selected parameters in the Stewart platform and the DC motor model. The obtained models (SimMechanics and uncertain state - space) were compared with the single linear actuator of the Stewart platform. The difference between the simulated SimMechanics model and measured data was almost completely covered by the uncertain model. The method is highly versatile and applicable on any state-space model.

Chcete být upozorněni, pokud se objeví nové záznamy odpovídající tomuto dotazu?
Přihlásit se k odběru RSS.