Basis set sampling in the method of coupled coherent states: Coherent state swarms, trains, and pancakes

The paper provides a systematic account of simple sampling techniques used in the multidimensional quantum dynamical method of coupled coherent states. For the sampling techniques based on a Gaussian distribution, it is noticed that faster convergence is achieved if compression of the basis set decreases as the basis size is increased. Good results are obtained for the autocorrelation functions of wave packets propagated in Henon-Heiles potentials with up to 32 degrees of freedom. Further test calculations are performed by employing trains of coherent states sampled on the same classical trajectory with successive time delays. (c) 2008 American Institute of Physics.