An Overview of System Identification Procedures and Perturbation Signal

System identification is a fundamental process in engineering and science that involves modeling and understanding the behavior of complex systems. This paper provides a comprehensive synopsis of system identification techniques, with a focus on parametric and non-parametric approaches, along with the role of perturbation signals in enhancing the analysis. The paper explores the process of estimating system parameters by fitting mathematical models to observed input-output data. The methods of parameter estimation, including least squares, Auto-Regressive with eXogenous input (ARX), Auto-Regressive Moving Average with eXogenous input (ARMAX), Output Error (OE) and Box Jenkins (BJ), are discussed in depth. Challenges such as noisy data, model complexity, and overfitting are also examined. Perturbation Signals delves into the significance of controlled inputs, or perturbation signals, in system analysis. Different types of perturbation signals, such as fixed spectrum signals, computer-optimised signals, are discussed. The application of these signals to study system responses, extract dynamic characteristics, and estimate transfer functions is elaborated. Considerations in selecting appropriate frequencies and amplitudes for perturbation signals are presented. This paper provides a holistic understanding of how system identification techniques, in conjunction with perturbation signals, contribute to unraveling the complexities of various systems. It serves as a valuable resource for researchers and practitioners seeking insights into accurately characterizing system dynamics and behavior.