P. Havel and P. Horáček (Czech Republic)
parameter estimation, global optimization, nonlinear model, glass feeder, simulation
This paper presents application of three selected global optimization algorithms that were used for parameter estimation. Parameters of a first-principle based nonlinear model of a real industrial glass feeder were identified based on historical data and least-squares criterion. Convergence rates of the algorithms are compared and simulations with the optimized parameters are shown both for train and test data set.
Important Links:
Go Back
