Incomplete information and fractal phase space (Q1432953)

This paper focuses on the statistical mechanics of fractal phase space where the non-trivial fractal dimension of unaccessible points lead to the concept of an incomplete information hypothesis. It explains first why with such a state-space the conventional probability normalization cannot be satisfied, due to unaccessible states, and indicates how to include them as elements of a physical system in a theory. A recent method building a statistical mechanics based upon this notion of incomplete information is developed through the introduction of an incompleteness parameter \(\omega\), whose physical signification is based on the fractal structure. The so-called incomplete information hypothesis, simply admitting that a part of the ignorance about complex systems may not be accessible, yields the notion of incompleteness \(\Omega\) to be the weight of the whole accessible space (not necessarily equal to one in this fractal structure). The incompleteness parameter \(\Omega\) is introduced as the power at which the probabilities of the accessible states have to be put to get a sum equal to 1. Generalized information notions are then developed in the same way, leading to the generalized notions of incompleteness entropies. Using a known connection between information and topological and fractal dimension of the phase space, the author relates then the incompleteness parameter \(\omega\) with the fractal dimension of unaccessible points in the special case of self-similar structures. This parameter is afterwards interpreted as an indicator of the phase space expansion, describing how much accessible points increase or decrease at each step of the refinement of the state space. Together with an interpretation in terms of the information growth, it describes explicitly how this incompleteness parameter \(\omega\) may be considered as a measure of the degree of chaos or fractal in a physical system.