System considerations for adaptive modelling applied to active noise control

Active noise-control systems using adaptive modeling must function in a complex electroacoustic environment that introduces many problems into the basic system identification problem. These include a complex primary plant, an acoustic feedback plant, high-order filters in the input and error paths, nonideal transducers, and a complex error path plant. In addition, the primary acoustic noise source contains a very broad range of frequencies that often exceeds the capabilities of the acoustic transducers and includes nonpropagating pseudosound due to turbulent flow. The authors discuss the application of adaptive digital signal processing to active noise control in this complex system. A complete active noise control system is described that uses the TMS320 family of digital signal processing microprocessors, and results are presented from an application using a centrifugal fan.<>