Bayesian Optimization Based Adaptive Control of Thermoacoustic Instabilities

Abstract

Thermoacoustics instability is one of the factors that limits the operational range of gas turbines. Both active and passive control strategies have been developed to suppress thermoacoustic instability. Active control strategies can adapt to the gas turbine's operational condition and hence provide a wider effectivity than the passive control strategies. However, it remains a challenge to best optimize the control parameters. This work proposes an adaptive control strategy by using Gaussian Process Regression with a modified Safe-Optimisation (SafeOpt) algorithm. The main benefit of the algorithm is it can safely explore the parameter space and minimize the cost function simultaneously. The performance of the algorithm in suppressing the thermoacoustic instability is experimentally demonstrated by stabilizing a turbulent combustor with loudspeakers actuation.