Stochastic Chaos with Its Control and Synchronization
Zhang Ying,Xu Wei,Zhang Tianshu,Yang Xiaoli,Wu Cunli,Fang Tong
2008-01-01
Abstract:The discovery of chaos in the sixties of last century was a breakthrough in concept,revealing the truth that some disorder behavior, called chaos, could happen even in a deterministic nonlinear system under barely deterministic disturbance. After a series of serious studies, people begin to acknowledge that chaos is a specific type of steady state motion other than the conventional periodic and quasi-periodic ones, featuring a sensitive dependence on initial conditions, resulting from the intrinsic randomness of a nonlinear system itself. In fact, chaos is a collective phenomenon consisting of massive individual chaotic responses, corresponding to different initial conditions in phase space. Any two adjacent individual chaotic responses repel each other, thus causing not only the sensitive dependence on initial conditions but also the existence of at least one positive top Lyapunov exponent (TLE) for chaos. Meanwhile, all the sample responses share one common invariant set on the Poincaré map, called chaotic attractor,which every sample response visits from time to time ergodically. So far, the existence of at least one positive TLE is a commonly acknowledged remarkable feature of chaos. We know that there are various forms of uncertainties in the real world. In theoretical studies, people often use stochastic models to describe these uncertainties, such as random variables or random processes.Systems with random variables as their parameters or with random processes as their excitations are often called stochastic systems. No doubt, chaotic phenomena also exist in stochastic systems, which we call stochastic chaos to distinguish it from deterministic chaos in the deterministic system. Stochastic chaos reflects not only the intrinsic randomness of the nonlinear system but also the external random effects of the random parameter or the random excitation.Hence, stochastic chaos is also a collective massive phenomenon, corresponding not only to different initial conditions but also to different samples of the random parameter or the random excitation. Thus, the unique common feature of deterministic chaos and stochastic chaos is that they all have at least one positive top Lyapunov exponent for their chaotic motion. For analysis of random phenomena, one used to look for the PDFs (Probability Density Functions) of the ensemble random responses. However, it is a pity that PDF information is not favorable to studying repellency of the neighboring chaotic responses nor to calculating the related TLE, so we would rather study stochastic chaos through its sample responses. Moreover, since any sample of stochastic chaos is a deterministic one, we need not supplement any additional definition on stochastic chaos, just mentioning that every sample of stochastic chaos should be deterministic chaos.We are mainly concerned with the following two basic kinds of nonlinear stochastie systems, i.e. one with random variables as its parameters and one with ergodical random processes as its excitations. To solve the stoehastie chaos problems of these two kinds of systems, we first transform the original stochastie system into their equivalent deterministic ones. Namely, we can transform the former stochastic system into an equivalent deterministic system in the sense of mean square approximation with respect to the random parameter space by the orthogonal polynomial approximation, and transform the latter one simply through replacing its ergodical random excitations by their representative deterministic samples.Having transformed the original stochastic chaos problem into the deterministic chaos problem of equivalent systems, we can use all the available effective methods for further ehaos analysis. In this paper, we aim to review the state of art of studying stochastic chaos with its control and synchronization by the above-mentioned strategy.