A comparative study of classic and modern time-domain nonlinear modeling techniques is presented for the case of loudspeakers. NARMAX modeling, as well as some modern methods, employing artificial neural networks, fuzzy logic, and their combinations, are applied using measured time-series input-output data. The comparison is performed on the basis of their ability to produce accurate but also adequate and parsimonious nonlinear loudspeaker models, as well as of the simplicity of their application. Both stationary random and real musical signals are used in the evaluation.
Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.