We have studied the temporal characteristics of different nonlinear stellar pulsation models in the nonlinear framework. In general the temporal behavior i.e., time series of amplitude variation are non-stationary and therefore the linear framework becomes inadequate for further characterization. Therefore, we have used wavelet analysis to investigate their characteristics in the time-frequency domain. Further the time series obtained in most of these model suggests period doubling and chaotic behavior for certain control parameters, in the phase space. As usually only one observational parameter is measured in a given setting, it is used to reconstruct an attractor in the phase space using phase space reconstruction method. Further characterization of the attractor and hence the time series ,obtained for various models, have been made using the recurrence quantification analysis (RQA).