| Summary: | Tensegrity structures are a type of structural systems that consist of a given set of cables connected to a configuration of rigid bodies and stabilized by internal forces of the cables in the absence of external forces. Such structures provide an important platform for exploring advanced active control technologies. This thesis is, thus, a research on tensegrity structures' related problems across a wide range of engineering disciplines and from a control system's viewpoint. It proposes a new algorithm for the form-finding of tensegrity structures. This is a process that involves using the mathematical properties of these structures to search and/or define a configuration that makes the structures to satisfy the conditions of static equilibrium while being pre-stressed. The dynamic model of tensegrity structures is derived using the Finite Element Method (FEM), and the static and dynamic analyses of tensegrity structures are carried-out. Furthermore, the effect of including additional structural members (than strictly necessary) on the dynamics of n-stage tensegrity structures is also investigated and how the resulting change in their geometric properties can be explored for self-diagnosis and self-repair in the event of structural failure is examined. Also, the procedures for model reduction and optimal placement of actuators and sensors for tensegrity structures to facilitate further analysis and design of control systems are described. A new design approach towards the physical realization of these structures using novel concepts that have not been hitherto investigated in the available literature on this subject is proposed. In particular, the proposed realization approach makes it possible to combine the control of the cable and bar lengths simultaneously, thereby combining together the advantages of both bar control and cable control techniques for the active control of tensegrity structural systems. The active control of tensegrity structures in a multivariable and centralized control context is presented for the design of collocated and non-collocated control systems. A new method is presented for the determination of the feedback gain for collocated controllers to reduce the control effort as much as possible while the closed-loop stability of the system is unconditionally guaranteed. In addition, the LQG (Linear system, Quadratic cost, Gaussian noise) controllers which are suitable for both collocated and non-collocated control systems is applied to actively control tensegrity structural systems for vibration suppression and precision control.
|
|---|
