Abstract
Boundary element methods in conjunction with optimization algorithms are employed for noise control of sound fields in three-dimensional enclosures. Total time averaged acoustic potential energy is chosen as the cost function for the optimization procedure. Boundary element method is used to model the monochromatic sound field generated by noise sources in the enclosure. In active noise control, the constrained steepest descend method proves to be effective when both positions and amplitudes of the secondary sources are to be optimized, while the direct matrix inversion, the steepest descent method, and the conjugate gradient method prove to be useful when only the amplitudes of secondary sources are to be optimized. In passive noise control, the optimal positions to apply sound-absorbing material with prescribed normal specific acoustic impedance can be determined by the complete search algorithm, the direct branching algorithm, and the branch-and-descend algorithm. The developed BEM-based optimization techniques are applied to control the noise in a rectangular enclosure. In addition, a vehicle cabin is chosen as a more realistic case to verify the optimal active and passive noise control techniques. Physical interpretations are derived from the simulation and considerations involved in practical implementation are also addressed.