Acoustic Signal Processing in Subject and Object in the Loop Configurations

Abstract

Acoustic signal processing is a discipline dealing with the information extracted from sound signals with applications in industry, entertainment areas etc. Two experimental configurations concerning real-time acoustic signal processing are developed for: voice manipulation and machine fault diagnosis, which share the same experimental equipment and signal processing software. First experimental study is based on the voice manipulation technique in a subject-in-the-loop configuration. A human voice was altered and fed back to the human subject in a signal manipulation process. The experimental setup validity was verified and real-time applications were achieved. Second experiment was conducted on a drive-motor-system to detect the fixture fault conditions due to loose bolts. An object-in-the-loop fault diagnosis system was developed using the acoustic signals acquired from loose/healthy conditions. The result presents that the loose versus healthy conditions can be identified in real-time. The numerical result illustrates the pattern of natural frequencies change under both conditions.