Few islands approximation of Hamiltonian system with divided phase space
Leonid Bunimovich, Giulio Casati, Tomaž Prosen, Gregor Vidmar
Abstract
Leonid Bunimovich, Giulio Casati, Tomaž Prosen, Gregor Vidmar
Abstract
It is well known that typical Hamiltonian systems have divided phase space consisting of regions with regular dynamics on KAM tori and region(s) with chaotic dynamics called chaotic sea(s). This complex structure makes rigorous analysis of such systems virtually impossible and significantly complicates numerical exploration of their dynamical properties. In this paper we outline a new approach for the analysis of Hamiltonian systems with divided phase space. These systems are approximated by a sequence of Hamiltonian systems having an increasing (but finite) number of KAM islands. The islands in the approximating systems are sub-islands of the islands in the initial system with an infinite number of KAM-islands. We apply this approach to two-dimensional billiards and demonstrate that it works. In particular the statistical characteristics of the approximating systems tend to the ones for the whole system when the number of islands in the approximating systems grows. Therefore our approach opens up a new way for numerical and analytical studies of the dynamics of Hamiltonian systems with divided phase space.
OpenAlex reports 5 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
It is well known that typical Hamiltonian systems have divided phase space consisting of regions with regular dynamics on KAM tori and region(s) with chaotic dynamics called chaotic sea(s). This complex structure makes rigorous analysis of such systems virtually impossible and significantly complicates numerical exploration of their dynamical properties. In this paper we outline a new approach for the analysis of Hamiltonian systems with divided phase space. These systems are approximated by a sequence of Hamiltonian systems having an increasing (but finite) number of KAM islands. The islands in the approximating systems are sub-islands of the islands in the initial system with an infinite number of KAM-islands. We apply this approach to two-dimensional billiards and demonstrate that it works. In particular the statistical characteristics of the approximating systems tend to the ones for the whole system when the number of islands in the approximating systems grows. Therefore our approach opens up a new way for numerical and analytical studies of the dynamics of Hamiltonian systems with divided phase space.
Key concepts: Phase space, Hamiltonian system, Chaotic, Mathematics, Torus, Hamiltonian (control theory), Standard map, Dynamical systems theory