Analysis and parameter estimation of time-varying signals: theory and methods
Ahmad Zuri Sha’ameri · Journal of Physics Conference Series · 2019
Abstract The need to capture signals, perform classification and make decisions are components of an Industry 4.0 manufacturing system. Since signals in practice are corrupted by noise, it is essential to optimize the signal processing and classification to ensure correct decision making. The time-varying nature of signals requires the use of time-frequency analysis over conventional methods such as spectrum estimation to ensure accurate estimation of signal parameters. The methods presented cover the main classes of time-frequency distributions – linear and quadratic- and compare their strengths and weaknesses for the appropriate choice of application.