Bogoliubov variational bound (Peierls–Bogoliubov inequality)
The Bogoliubov variational bound is a widely used inequality that turns free-energy estimation into an optimization problem over tractable reference models. It underlies many “mean-field-like” approximations and provides controlled upper bounds on the canonical free energy .
Statement
Let be the true Hamiltonian and fix inverse temperature $\beta$ . Choose a reference Hamiltonian with partition function and free energy
Let denote expectation in the reference canonical ensemble defined by .
Then the Bogoliubov (Peierls–Bogoliubov) inequality states
where is the true free energy.
Derivation idea (Jensen / convexity)
Write the partition function using a trace (for classical systems, this trace is the sum/integral over microstates):
Because the exponential is convex, Jensen's inequality implies
Taking of both sides gives the stated upper bound on .
The same inequality can also be viewed as a consequence of nonnegativity of relative entropy between the trial Gibbs state generated by and the true Gibbs state.
Variational use
If depends on variational parameters (e.g. effective fields), then
Optimizing the right-hand side yields the tightest bound within the chosen class of reference Hamiltonians.
A standard choice is to take to be a sum of independent one-site terms; optimizing then produces self-consistency equations that coincide with the mean-field variational construction .
Physical interpretation
The bound says: “an interacting system is no less stable (no lower free energy) than a solvable trial system plus the average interaction correction measured in the trial ensemble.” Practically, it converts difficult interacting thermodynamics into an optimization problem, while guaranteeing that the resulting approximate free energy is an upper bound on the exact one.