Abstract:

Genetic algorithms have proven to be helpful for solving complex optimization problems where more traditional gradient-based methods have failed. The optimal design of vibroacoustic systems often leads to complex optimization problems requiring a method able to find a globally optimal solution through a large number of locally optimal solutions. The vibroacoustic behavior of a simply supported plate carrying one or more point masses has been investigated through a genetic algorithm optimization method, in order to determine optimal placement of the masses for maximal sound or vibration attenuation. It has been shown that the use of an optimization method able to search through the whole optimization domain can give unexpected solutions resulting in a higher level of reduction in radiated noise or vibration response, when compared to intuitive solutions like placing the point masses on excitation points. In particular, when a well-defined frequency range is targeted, genetic algorithms can be used for tailoring a special design giving highest reduction in the specified range. [Work supported by the IRSST.]