ACTIVE NOISE CONTROL IN DUCT USING A CONSTRAINED STATE-SPACE MODEL PREDICTIVE CONTROL

Renato Vilela Lopes¹; Maria Alzira de Araújo Nunes¹; André Murillo de Almeida Pinto¹

¹ Universidade de Brasília, Faculdade UnB Gama

doi:10.20906/CPS/COB-2015-0053

Resumo

In this paper, a state-space MPC formulation with constraint handling is proposed for design of Active Noise Control (ANC) in an acoustic duct. The proposed methodology uses as a prediction model a special formulation that uses plant input and output variables as its state variables. The constraint-handling ability of the predictive controller was exploited to ensure that the maximum variation of the control signal is within an acceptable range. The performance of the proposed control loop is demonstrated via numerical simulations and comparison of the results with the classic Filter-x Least Mean Square (FXLMS) algorithm.

Palavras-chave: MPC; ANC; Duct; State-Space