Nonlinear Model Identification of a Ball and Beam Mechanism using Experimental Data


Abedinifar M., Ertuǧrul Ş., Argüz S. H.

9th International Conference on Control, Decision and Information Technologies, CoDIT 2023, Rome, Italy, 3 - 06 July 2023, pp.105-110 identifier

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/codit58514.2023.10284105
  • City: Rome
  • Country: Italy
  • Page Numbers: pp.105-110
  • Istanbul Technical University Affiliated: Yes

Abstract

A ball and beam mechanism is widely utilized in laboratory experiments to demonstrate the behavior of more complex systems. In this research, the phenomena such as nonlinear frictions, dead-zone and time-delay in the ball and beam mechanism's mathematical model is investigated. The following procedures are taken to construct a credible mathematical model of the system for this purpose. Firstly, the ball and beam mechanism's mathematical model, which includes different probable physically meaningful nonlinearities, is simulated using MATLAB\Simulink. Then, the Particle Swarm Optimization (PSO) algorithm is coded to determine the exact nonlinear model of a ball and beam system using the experimental data. Third, the accuracy of the results obtained from the PSO algorithm is tested using the hypothesis test and the confidence interval test. According to the statistical tests, the PSO algorithm is highly accurate in determining the parameters of the actual model of the system.