Feedback control of acoustic disturbance transient growth in triggering thermoacoustic instability

摘要

Transient growth of acoustic disturbances could trigger thermoacoustic instability in a combustion system with non-orthogonal eigenmodes, even with stable eigenvalues. In this work, feedback control of transient growth of flow perturbations in a Rijke-type combustion system is considered. For this, a generalized thermoacoustic model with distributed monopole-like actuators is developed. The model is formulated in state-space to gain insights on the interaction between various eigenmodes and the dynamic response of the system to the actuators. Three critical parameters are identified: (1) the mode number, (2) the number of actuators, and (3) the locations of the actuators. It is shown that in general the number of the actuators K is related to the mode number N as K = N~2. For simplicity in illustrating the main results of the paper, two different thermoacoustic systems are considered: system (a) with one mode and system (b) that involves two modes. The actuator location effect is studied in system (a) and it is found that the actuator location plays an important role in determining the control effort. In addition, sensitivity analysis of pressure- and velocity-related control parameters is conducted. In system (b), when the actuators are turned off (i.e., open-loop configuration), it is observed that acoustic energy transfers from the high frequency mode to the lower frequency mode. After some time, the energy is transferred back. Moreover, the high frequency oscillation grows into nonlinear limit cycle with the low frequency oscillation amplified. As a linear-quadratic regulator (LQR) is implemented to tune the actuators, both systems become asymptotically stable. However, the LQR controller fails in eliminating the transient growth, which may potentially trigger thermoacoustic instability. In order to achieve strict dissipativity (i.e., unity maximum transient growth), a transient growth controller is systematically designed and tested in both systems. Comparison is then made between the performance of the LQR controller and that of the transient growth controller. It is found in both systems that the transient growth controller achieves both exponential decay of the flow disturbance energy and unity maximum transient growth.