Nonlinear generalized master equations and accounting for initial correlations (Q1034518)

The problem of the derivation of equations that allow describing the evolution of the measurable values that characterize nonequilibrium states of a many-particle systems still remains one of the basic tasks of statistical physics. These equations generally should be the nonlinear evolution equations converting into kinetic equations on some time scale. Several approaches are usually used to solve this problem, commonly starting with the linear Liouville-von Neumann equation for a distribution function of the many-body system under consideration. Another approach leads to the chain of coupled first-order differential equations for many-particle distribution functions known as the Bogoliubov-Born-Green-Kirkwood-Yvon (BBGKY) hierarchy. One of the possible attempts was given by Bogoliubov (see the Introduction in the paper for a short explanation). Bogoliubov successfully derived the kinetic (Markovian) equations for a one-particle distribution function, the (classical and quantum) Boltzmann equation in particular. However, on the other hand it should be noted that there is not yet a way to take initial correlations into account exactly when dealing with the BBGKY chain. An another approach leads to the so-called generalized master equation (GME) where an \(N\)-particle distribution function (or statistical operator) is divided into the essential (relevant) and inessential (irrelevant) parts using time-independent projection operators. But generally it is impossible to obtain a nonlinear master equation (like the Boltzmann equation) in the framework of the linear time-independent projection operator formalism without additional approximations and the proper interpretation of the initial correlations is crucial in obtaining the closed evolution equations. The initial correlations can be important for considering the ultrafast relaxation and non-Markovian processes and also for describing the decoherence phenomenon and to take the initial correlations into account, a method (based on the conventional time-independent projection operator technique) that allows converting the linear inhomogeneous time-convolution GME (TC-GME) and time-convolutionless GME (TCL-GME) into the homogeneous form was proposed [see \textit{V. F. Los}, J. Phys. A, 34, 6389--6403 (2001; Zbl 0982.82018); J. Stat. Phys., 119, 241--271 (2005; Zbl 1125.82027)]. This approach resulted in exact linear time-convolution and time-convolutionless homogeneous GMEs (TC-HGME and TCL-HGME), which take the dynamics of the initial correlations into account via modified memory kernels governing the evolution of the relevant part of the distribution function of a many-particle system. In the paper the author analyzed how the obtained linear TC-HGME works for a spatially homogeneous dilute gas of classical and quantum particles. A new class of evolution equations for the relevant part of a distribution function that are suitable for deriving both the nonlinear and linear evolution equations for the reduced distribution functions of interest and that can hold on all time scales was obtained. In Sec. 2, the exact inhomogeneous nonlinear TC-GME for the relevant part of a distribution function containing an irrelevant part (including the initial correlations) is given. This equation is a generalization of the Nakajima-Zwanzig linear GME and can be an alternative to the BBGKY chain. In Sec. 3, the obtained inhomogeneous nonlinear GMEs was applied to the spatially inhomogeneous nonideal dilute gas of classical particles. In the linear approximation with only two-particle correlations (collisions) taken into account this inhomogeneous nonlinear GME results in the new inhomogeneous nonlinear equation for a one-particle distribution function containing a self-consistent Vlasov term. In Sec. 4, taking into account the initial correlations the inhomogeneous nonlinear GME obtained in Sec. 2 was transformed into the nonlinear TC-HGME. In the next section the derived homogeneous nonlinear GME was applied to the same system as in Sec. 3. In the linear approximation of the density parameter, a new homogeneous nonlinear equation for a one-particle distribution function that holds on all time scales and takes the influence of the dynamics of initial correlations into account in all stages of the evolution process was obtained. If the gas dynamics have the necessary property (ergodic mixing flow in the phase space), then all initial correlations vanish on the kinetic time scale, and this equation can be converted into the irreversible nonlinear Boltzmann equation with a selfconsistent nonlinear Vlasov term.