The principle of the active noise control is the cancellation of an acoustic disturbance based on the destructive interference with another noise produced by the controller with the same amplitude but opposite phase. While the literature is mainly concerned with linear controllers, there is evidence that nonlinear effects may influence the behavior of an active control system. Moreover, it is widely recognized that multichannel schemes allow the spatial extension of the silenced area, in spite of their drawbacks mainly related to their computational complexity and some other critical behaviors. Such considerations thus motivate a research on nonlinear multichannel controllers.
The research is developed according to the following three main lines. Theoretical and methodological studies. This research line aims to study behaviors that are not yet fully understood and efficiently dealt with in a nonlinear multichannel environment, as those related to the correlations existing among the various signals involved and the necessity of using polynomial inverse or pseudo-inverse filters in presence of nonlinearities with memory. Derivation of efficient adaptation algorithms. The aims of this part of the research is the derivation of efficient adaptation algorithms for nonlinear multichannel active noise controllers. We will specifically consider two classes of widely used nonlinear models, i. e. the models based on the discrete Volterra series and those based on the fuzzy system theory. Derivation of efficient implementations. The controllers derived according to the models above mentioned will be compared in order to find out the realizations most convenient for the actual implementation. Traditional solutions based on versatile Digital Signal Processors will be considered. Implementations based on programmable logic devices (CPLDs or FPGAs) or Application Specific Integrated Circuits (ASIC) will be investigated, too.
The final part of the research will be devoted to the accurate and critical evaluation of the results obtained. The output of this analysis is expected to be a more complete knowledge of the various aspects involved in the problem of nonlinear multichannel active noise control. It is also expected that all the information derived in this final phase be exploited in a subsequent experimental and application-oriented research.