Balanced control of flexible structures (Q1906838): Difference between revisions

Control of flexible structures, a growing area of research, includes vibration suppression in bridges and tall buildings, active control of vehicle vibrations, suppression of rotor unbalanced dynamics, control of large flexible space structures, tracking control of antennas and radars, aircraft flutter suppression, and precision pointing of space born reflectors. The book provides analysis and design tools for the control of flexible structures. This is a postgraduate text on control systems analysis and design, addressed not only to students in the fields of control, mechanical, aeronautical, civil and electrical engineering, but also to scientists and professional engineers. The author assumes that readers have a basic background in control systems analysis. This background is not provided, but to make the text more tractable, each presented result is extensively illustrated with examples and numerical data. The examples are twofold simple ones that could be readily repeated by a reader, and complex ones (taken from the ``real world'') that demonstrate the effectiveness of the approach. Flexible structures are linear systems of specific properties, most often known as modal properties. These distinctive properties make their analysis and controller design a unique task. The modal approach has been most successfully applied in the analysis of open-loop dynamics. However, the modal approach is often used arbitrarily, since it lacks proper scaling properties (mode scale is a free factor, left to the designer discretion). The balanced approach combines modal properties with proper mode scaling, such that certain but crucial properties of the system (in this case controllability and observability) are reflected in the state equations. This book explains the balanced techniques, derives the properties of flexible structures in the open- and closed-loop configuration, and presents a uniform approach to open- and closed-loop analysis and design. Chapter 1 presents the basic ideas and approaches which are taken further in the book. In Chapter 2, a flexible structure is defined and its properties are specified. State-space representations and four forms of modal representations are introduced. Chapter 3 defines open-loop balanced systems and their properties. Specifically, properties of flexible structures in balanced coordinates derived in this chapter are further used for model order reduction and controller design. In Chapter 4, model reduction of an open-loop system is introduced. Optimality criteria of reduction show that in the case of flexible structures, the reduction is near-optimal. Next, reduction of systems with integrators, reduction in finite time and/or frequency intervals, reduction based on transfer function measurements, and reduction of structures equipped with sensors and actuators are presented. The method is illustrated with the reduction of a COFS-1 structure (NASA space station test item), and advanced supersonic transport, and a NASA deep space network antenna. In Chapter 5, methods for sensor and actuator placement for balanced flexible structures are presented and illustrated using the actuator placement on the CEM large flexible structure at NASA Langley Research Center. In Chapter 6, the balanced identification procedure (ERA Eigensystem Realization Algorithm) is introduced, because it is a balanced approach to the system representation, and identification is tightly connected with the design of model based controllers. Chapter 7 introduces the balanced dissipative controllers. These controllers are simple and always stable, but their simplicity also creates difficulties: the performance depends on the number of inputs and outputs (the more, the better), and for practical purposes, the collocation of inputs and outputs is required. In this chapter, a formal design procedure of dissipative controllers for flexible structures is introduced. Chapter 8 defines a balanced LQG controller to be one for which the solutions to a Riccati equation are equal and diagonal. Properties of the LQG balanced systems are derived and later specified for flexible structures. These properties allow determination of the explicit relationship between the LQG weights and system performance. A method of reduction of the balanced LQG controller is introduced and illustrated with the LQG controller design for the NASA deep space network antenna. In Chapter 9 a balanced \(H_\infty\) controller is introduced, and its properties that show its connection with the balanced LQG controller and balanced open-loop system are derived. A method of reduction of the \(H_\infty\) balanced controller gives stable reduced-order controllers with performance close to the full-order controller, as illustrated with controllers for truss structures and for the NASA deep space network antenna.