Improved control design variable linking for optimization of structural/control systems
Explore the source record for details and available documents.
Jin, Ik M.↗
Engineering topics
Schmit, Lucien A.
Publications and source records attributed to Schmit, Lucien A..
Control design variable linking for optimization of structural/control systems
A method is presented to integrate the design space of structural/control system optimization problems in the case of linear state feedback control. Conventional structural sizing variables and elements of the feedback gain matrix are both treated as strictly independent design variables in optimization by extending design variable linking concepts to the control gains. Several approximation concepts including new control design variable linking schemes are used to formulate the integrated structural/control optimization problem as a sequence of explicit nonlinear mathematical programming problems. Examples which involve a variety of behavior constraints, including constraints on dynamic stability, damped frequencies, control effort, peak transient displacement, acceleration, and control force limits, are effectively solved by using the method presented.
Jin, Ik Min↗
Control design variable linking for optimization of structural/control systems
Explore the source record for details and available documents.
Jin, Ik M.↗
Improved calculation of optimum design sensitivity
Optimum design parameter sensitivity analysis has become an important topic in recent years. The principal reasons for obtaining optimum design sensitivity information with respect to various problem parameters are (l) to predict revised optimum designs, associated with specified perturbations of the problem parameters, without re-optimizing the problem and, (2) to provide sensitivity information in multilevel optimization strategies. Methods for calculating these derivatives were proposed by several authors. The most important drawback in estimating parameter sensitivities by the current methods is that they do not allow for changes in the active constraint set. Changes in the active constraint set produce discontinuities in the optimum design sensitivities, and therefore, a correct formulation has to be cast in terms of directional derivatives. Optimum design parameter sensitivity analysis is addressed in terms of directional derivatives so as to include possible discontinuities and it also presents a method for estimating optimum design sensitivities using finite differences.
Sepulveda, Abdon E.↗
Improved approximations for dynamic displacements using intermediate response quantities
An approximation for dynamic displacements, which captures the nonlinearities associated with resonance, is presented. This approximation is constructed using approximate intermediate response quantities. When dynamic displacements are constrained using this high quality approximation, frequency constraints are no longer needed to keep the design away from resonance.
Thomas, Harold L.↗
Control Augmented Structural Synthesis
A methodology for control augmented structural synthesis is proposed for a class of structures which can be modeled as an assemblage of frame and/or truss elements. It is assumed that both the plant (structure) and the active control system dynamics can be adequately represented with a linear model. The structural sizing variables, active control system feedback gains and nonstructural lumped masses are treated simultaneously as independent design variables. Design constraints are imposed on static and dynamic displacements, static stresses, actuator forces and natural frequencies to ensure acceptable system behavior. Multiple static and dynamic loading conditions are considered. Side constraints imposed on the design variables protect against the generation of unrealizable designs. While the proposed approach is fundamentally more general, here the methodology is developed and demonstrated for the case where: (1) the dynamic loading is harmonic and thus the steady state response is of primary interest; (2) direct output feedback is used for the control system model; and (3) the actuators and sensors are collocated.
Lust, Robert V.↗