Relations entre l’entropie physique, le codage de l’information et l’énergie de simulation

Abstract

It is shown that an isomorphism exists between the mixing entropy and the size of a computer program that simulates the physical system. This isomorphism must be constructed with respect to some theorems, and it is shown that the composition of two isomorphisms, one based on a run length encoding and another by encoding sequences in a dictionary allows us to quantify the entropy of binary and ternary systems at the equilibrium. Finally, it is shown that the energy consumed by a physical system encoded by this system and executed on a Turing machine is proportional to the free energy of the thermodynamic system.