Abstract
We apply the approximate Bayesian computation (ABC) method to estimate aerodynamic coefficients of table tennis ball from its flight trajectory. The ABC method replaces the calculation of likelihood function to the comparison between measured and simulated data. The drag and lift coefficients are employed as parameters, and ball trajectories are predicted with the parameters by integrating the equation of motion for a projectile. When the Euclidian distance between the measured and simulated trajectories is smaller than a tolerance, then the parameters are accepted. The obtained results are compared with those with another estimating method (interpolation method) which is formulated by the polynomial interpolation of measured trajectory data. The results of ten trajectory data are compared, and it leads to the conclusion that obtained drag coefficient shows good agreement in both the methods while the values in the lift coefficient with the ABC method are smaller than those with the interpolation method. The ABC method can successfully estimate the aerodynamic coefficients of a projectile in flight from its trajectory data. The determination of adequate tolerance in the distance function, however, remains as one of important problems.
