Relaxation Processes and Time Scale Transformation.
NAVAL RESEARCH LAB WASHINGTON DC
Pagination or Media Count:
Stochastic processes with a special class of non-stationary transition rates NSTR that can be related to stationary rates STR by time scale transformations are considered. Calculations are carried out on the time scale on which the process has STR and then related to the time scale of measurements by a time scale transformation. Application is made to Gaussian Markoff processes with a single transition rate. The resulting equilibrium autocorrelation function departs from weak stationarity. A response function is related to the Autocorrelation function and specialized to relaxation processes to yield a model consistent with Ngais recent model of relaxation. The model is therefore widely applicable in the description of measured relaxation behavior and leads naturally to Ngais renormalization relation for excitation energies. On the basis of the success of the model, consideration is given to the possibility that the process should be viewed as one with STR and an intrinsic time scale so that the process only appears to have NSTR on the time scale of measurements. Author
- Statistics and Probability