Modified LQG/LTR Control Methodology in Active Structural Control
Sheng Yan, Chao Wang, Ying Pan, Xinhua Zhang
Abstract
Sheng Yan, Chao Wang, Ying Pan, Xinhua Zhang
Abstract
This paper presents a new active structural control design methodology comparing the conventional linear-quadratic-Gaussian synthesis with a loop-transfer-recovery (LQG/LTR) control approach for structures subjected to ground excitations. It results in an open-loop stable controller. Also the closed-loop stability can be guaranteed. More importantly, the value of the controller's gain required for a given degree of LTR is orders of magnitude less than what is required in the conventional LQG/LTR approach. Additionally, for the same value of gain, the proposed controller achieves a much better degree of recovery than the LQG/LTR-based controller. Once this controller is obtained, the problems of control force saturation are either eliminated or at least dampened, and the controller band-width is reduced and consequently the control signal to noise ratio at the input point of the dynamic system is increased. Finally, numerical examples illustrate the above advantages.
OpenAlex reports 2 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
This paper presents a new active structural control design methodology comparing the conventional linear-quadratic-Gaussian synthesis with a loop-transfer-recovery (LQG/LTR) control approach for structures subjected to ground excitations. It results in an open-loop stable controller. Also the closed-loop stability can be guaranteed. More importantly, the value of the controller's gain required for a given degree of LTR is orders of magnitude less than what is required in the conventional LQG/LTR approach. Additionally, for the same value of gain, the proposed controller achieves a much better degree of recovery than the LQG/LTR-based controller. Once this controller is obtained, the problems of control force saturation are either eliminated or at least dampened, and the controller band-width is reduced and consequently the control signal to noise ratio at the input point of the dynamic system is increased. Finally, numerical examples illustrate the above advantages.
Key concepts: Linear-quadratic-Gaussian control, Control theory (sociology), Optimal projection equations, Controller (irrigation), Linear-quadratic regulator, Open-loop controller, Optimal control, Engineering