Adaptive umbrella sampling: Self-consistent determination of the non- Boltzmann bias (Q1102067)

When using Monte Carlo or molecular dynamics method in statistical mechanics it is sometimes necessary to use umbrella sampling [\textit{G. Torrie} and \textit{J. P. Valleau}, ibid. 23, 187 (1977)] to sample a specific region proving bounds for the inverse matrix. First we give a different computation scheme for the iteration formula which is more efficient especially for the higher-order formulas. Next, we derive a modification of the methods which has the same properties as the original ones but compares favourably for the higher-order cases. For both versions presented here some efficiency indices are listed and compared with those of the original formulas.