Volume 69, Issue 9 - Schwerpunktheft: Ausgewählte Beiträge des GMA-Fachausschusses 1.40, Teil 2 / Thomas Meurer, Alexander Schaum

Adaptive structures where actuators are incorporated into a building structure have the potential to reduce resource consumption in construction industry drastically. However, the performance of static load compensation depends to a large extend on the actuator placement. This paper presents optimal actuator placement for systems with distributed parameters based on the Gramian compensability matrix. To provide a general framework for different kind of loads, static loads are discretized as Dirac impacts. The resulting optimal actuator placement is robust against unknown load amplitudes, as load profiles are only considered qualitatively in the cost function. Further, the optimal control input for a given load results directly from the optimization problem. The procedure is illustrated for a Kirchhoff-Love plate and integrated fluidic actuators.

Windup effects can be subdivided into controller windup and plant windup. Controller windup can be prevented by stabilizing the compensator during saturation and plant windup by an additional dynamic element. When using a compensating design, i. e., the zeros and poles of the plant are compensated by the poles and zeros of the controller, plant windup does not occur. The compensating control is parametrized by one parameter allowing nearly arbitrary disturbance attenuation. This type of control is restricted to minimum-phase systems. But it has a number of advantages. It simplifies the SISO and especially the MIMO design of compensators with integral action considerably, it has good robustness properties and it allows a diagonal decoupling of the reference behavior for arbitrary MIMO system. Two examples demonstrate the results achievable.

Detailed prediction models with robust constraints and small sampling times in Model Predictive Control yield conservative behavior and large computational effort, especially for longer prediction horizons. Here, we extend and combine previous Model Predictive Control methods that account for prediction uncertainty and reduce computational complexity. The proposed method uses robust constraints on a detailed model for short-term predictions, while probabilistic constraints are employed on a simplified model with increased sampling time for long-term predictions. The underlying methods are introduced before presenting the proposed Model Predictive Control approach. The advantages of the proposed method are shown in a mobile robot simulation example.

This paper proposes a scheme of model predictive control for single-loop networked control system (NCS) with probabilistically modeled communication channel and disturbances. Uncertainties of the communication network are projected onto a tailored probability for the satisfaction of state and input constraints. The proposed receding horizon control scheme uses a probabilistic terminal state and set to establish a balance between control performance and state probability distribution, while satisfying the given constraints. In addition to describing the control approach, its properties are discussed, and it is illustrated by an example.

Zusammenfassung Für den sicheren Betrieb von mobilen Großraummanipulatoren wie beispielsweise Autobetonpumpen ist eine bestmögliche Abstützung essentiell. Zurzeit erfolgt diese Aufgabe manuell und hängt damit stark von der Erfahrung und Qualifikation des Bedienpersonals ab. In dieser Arbeit werden für eine Autobetonpumpe auf Basis eines detaillierten mathematischen Modells mögliche optimale Abstützkonfigurationen untersucht. Darauf aufbauend wird eine Regelungsstrategie für die autonome horizontale und vertikale Ausrichtung der Betonpumpe bei gleichzeitig optimaler Standsicherheit entworfen. Die Funktionsweise des vorgestellten Konzepts wird anhand von Simulationsstudien nachgewiesen.

Zusammenfassung Der Beitrag behandelt die Folgeregelung von Flachwasserwellen in einer partiell gefüllten Röhre, in der der Flüssigkeitspegel durch einen beweglichen Kolben variiert werden kann. Zur Beschreibung der Flachwasserwellen werden die Saint-Venant-Gleichungen in Euler-Koordinaten verwendet, ein quasilineares hyperbolisches Differentialgleichungssystem zweiter Ordnung. Der Beitrag fasst bereits in Vorarbeiten erzielte Ergebnisse zur Flachheit und zum darauf basierenden Steuerungsentwurf zusammen und erweitert diese Resultate um eine einfache auf den linearisierten Modellgleichungen beruhende flachheitsbasierten Folgeregelung. Die Leistungsfähigkeit der entworfene Folgeregelung wird anhand von Simulationen und experimentellen Studien nachgewiesen.

This contribution is concerned with the design of observers for a single mast stacker crane, which is used, e. g., for storage and removal of loads in automated warehouses. As the mast of such stacker cranes is typically a lightweight construction, the system under consideration is described by ordinary as well as partial differential equations, i. e., the system exhibits a mixed finite-/infinite-dimensional character. We will present two different observer designs, an Extended Kalman Filter based on a finite-dimensional system approximation, using the Rayleigh-Ritz method and an approach exploiting the port-Hamiltonian system representation for the mixed finite-/infinite-dimensional scenario where in particular the observer-error system should be formulated in the port-Hamiltonian framework. The mixed-dimensional observer and the Kalman Filter are employed to estimate the deflection of the beam based on signals acquired by an inertial measurement unit at the beam tip. Such an approach considerably simplifies mechatronic integration as it renders strain-gauges at the base of the mast obsolete. Finally, measurement results demonstrate the capability of these approaches for monitoring and vibration-rejection purposes.