In this study

we tried to deal with the experimental data of stress relaxation of silicone rubber foam. Based on the second-order Prony series

the nonliner regression analysis method (curve fitting method)

the Bayesian method based on the Wiener process and the Bayesian method based on the Gamma process were used to estimate the parameters

and the prediction models are established respectively. By comparing the deviation between the predicted values and experimental values of the three modeling methods

the effects of the three modeling methods are compared. The results showed that the prediction deviation of the model based on the regression analysis was larger than the ones established by the Bayesian method based on the Wiener and Gamma processes

which may indicate that the stochastic process models could better predict the stress relaxation tendency of silicone rubber. The non-negative Gamma process

suitable for the nature of monotone increasing of a stress relaxation process

may contribute to its less deviation than the Wiener process. Our work provided a better way for the parameters estimation of degradation model based on the performance degradation data of polymer materials

contributing to better prediction.