Genetic Algorithm Based Automatic Input Parameter Calibration Method for the Discrete Element Modeling of Vibration Feeders

Using the genetic algorithm, this paper developed an automatic method for calibrating the used four input parameters (Young's modulus, Poisson's ratio, coefficient of restitution, and friction coefficient) in the discrete element modeling (DEM) of vibration feeders. Two objective functions were used: (1) the difference in the results between the DEM simulation and the experiment and (2) the computation time of the DEM simulation. We conducted preliminary experiments to obtain an average experimental response of the vibration feeding and found that single-item-clogging usually arises near the feeder tray outlet. The proposed method managed to find the required set of parameters for minimizing the two objective functions by multi-objective optimization using the genetic algorithm. Overall, we confirmed that the obtained parameter values were physically reasonable in comparison with those of previous studies.