糖心TV

Will there ever be a complete theory of nonequilibrium systems having 
the same predictive power as that of equilibrium statistical 
mechanics? Though most researchers in statistical mechanics would hope 
and conjecture that the answer is yes, we do not have at present such 
a theory.

We know, of course, a lot about nonequilibrium systems. Results such 
as Onsager's reciprocity relations, Onsager and Machlup's principle of 
minimum dissipation, and the more recent results that are Jarzynski's 
equality and the Gallavotti-Cohen fluctuation theorem, among others, 
all point to many general and interesting properties of nonequilibrium 
systems and their fluctuations. However, no one has come up so far 
with a common thread -- a common theory -- accounting for or 
integrating all of these results. Does such a theory exist? Can these 
results be viewed as consequences of a small number of general 
principles? Taking equilibrium statistical mechanics as a model, is 
there a way to devise a theory based on something generalizing the 
concept of a statistical-mechanical ensemble that would be applicable 
to nonequilibrium systems at large?

This mini-workshop will aim at summarizing part of our current 
understanding of these questions and at advancing some possible 
answers. The speakers will provide a background for the discussion by 
presenting some recent results related to the foundations of 
nonequilibrium statistical mechanics. An open discussion will follow, 
at which participants are expected to ask questions and advance ideas 
as to where we stand in terms of research on nonequilibrium systems, 
and how we might pursue that research. Everyone is welcome.