Abstract:In the study of non-equilibrium statistical mechanics, Ruelle derived explicit formulae for entropy production of smooth dynamical systems. The vanishing or strictly positivity of entropy production is determined by the entropy formula of folding type \[h_{\mu}(f)= F_{\mu}(f)-\displaystyle\int\sum\nolimits_{\lambda_i(f, x)<0} \lambda_i(f, x)d\mu(x), \] which relates the metric entropy, folding entropy and negative Lyapunov exponents. This paper establishes the formula for all $C^{1+\alpha}$ maps, including those with degeneracy. More specifically, the entropy formula of folding type holds if and, under a condition on the Jacobian of the map, only if $\mu$ has absolutely continuous conditional measures on the stable manifolds. To overcome the degeneracy, the Pesin Theory for general $C^{1+\alpha}$ maps is developed.
What problem does this paper attempt to address?
### What problems does this paper attempt to solve?
This paper aims to explore and prove a new entropy formula, namely the folding - type entropy formula, which is applicable to all \(C^{1 + \alpha}\) mappings, including those with degenerate points. Specifically, the formula is:
\[h_{\mu}(f)=F_{\mu}(f)-\int\sum_{\lambda_{i}(f, x)<0}\lambda_{i}(f, x)d\mu(x)\]
where:
- \(h_{\mu}(f)\) represents the metric entropy,
- \(F_{\mu}(f)\) represents the folding entropy,
- \(\lambda_{i}(f, x)\) represents the Lyapunov exponent.
To achieve this goal, the paper has solved the following key problems:
1. **Extension of Ruelle's inequality**: Ruelle's inequality was originally only applicable to smooth dynamical systems, while this paper generalizes it to \(C^{1+\alpha}\) mappings and takes into account the existence of degenerate points.
2. **Absolutely continuous condition**: It is proved that a necessary and sufficient condition for the above - mentioned entropy formula to hold is that the measure \(\mu\) has an absolutely continuous conditional measure on the stable manifold. Specifically, for almost all \(x\), the conditional measure \(\mu_{x}\) on the stable manifold \(W^{s}(x)\) must be absolutely continuous with respect to the Lebesgue measure \(\lambda_{s}^{x}\).
3. **Overcoming degeneracy**: For degenerate points (i.e., points where \(\det(D_{x}f) = 0\)), the paper introduces a generalized version of Pesin's theory to ensure the validity of certain local properties and estimates even in degenerate cases.
4. **Establishing a measurable partition**: By constructing a measurable partition compatible with the stable manifold, the author is able to decompose the global dynamical system behavior into local behaviors, thus better analyzing the entropy properties of the system.
### Markdown format of formulas
The following are the formulas involved presented in Markdown format:
\[h_{\mu}(f)=F_{\mu}(f)-\int\sum_{\lambda_{i}(f, x)<0}\lambda_{i}(f, x)d\mu(x)\]
\[F_{\mu}(f)=H_{\mu}(\epsilon|f^{- 1}\epsilon)\]
\[\sum_{z: fz = y}J_{f}(z)=1\]
\[\prod_{n = 0}^{\infty}\frac{J_{f}(f^{n}y)}{J_{f}(f^{n}x)}<\infty\]
\[\left|\log D_{s}(f^{n}y)-\log D_{s}(f^{n}x)\right|\leq4\tilde{C}_{0}(e^{k_{0}\epsilon}C'_{0})^{\tilde{\alpha}}e^{(k_{0}+n - n_{0})\epsilon(\tilde{\alpha}+s)}d^{\tilde{\alpha}}_{Ws}(f^{n}x,f^{n}y)\]
These formulas show how to understand and calculate the entropy properties of dynamical systems from different perspectives (such as folding entropy, Lyapunov exponents, etc.).