AN OPTIMIZATION PROCEDURE IN A SIROCCO FAN USING EXPERIMENTAL DATA

摘要

The flow rate and the specific noise level of 18 siroccofans were measured and these data were analyzed by theTaguchi method and the neural network. The optimal designvalue obtained by the neural network generally showed goodagreement with that by the Taguchi method. The effects ofeight important design variables on the flow rate and thespecific noise level were discussed.