Skip to Main Content
This paper is aiming at presenting parameter estimation using the Differential Evolution algorithm. Previously, evolution based algorithms has been applied to frequency domain data. Within this study, the Differential Evolution algorithm has been applied to time-domain data. Convergence of this estimation procedure, presented within this paper, is validated by using synthetic data. Results obtained are presented, and especially, the effect of the length of the excitation signal and convergence differences between distinct runs are compared. Due to the stochastic nature of the DE algorithm, it can not be said exactly how many iteration round are needed. However, based on the results obtained within this study, it can be said that the parameters were found in 300 iterations.
Date of Conference: 3-6 May 2009