在現今，產品大多具備高可靠度性質，生產者為了要提升產品競爭力，需要不斷地提供顧客關於可靠度的資訊，如產品的壽命推論，而衰變試驗(degradationtest)常被用來推估壽命相關資訊。而文獻上分析衰變資料往往使用特定隨機過程模型，例如Wiener、Gamma、Inverse Gaussian過程之衰變模型，而上述常見之衰變模型又為Tweedie衰變模型之特例，在收集資料後，對假設的模型進行估計，以得到產品壽命之推論。本文假設衰變資料來自Tweedie衰變模型，因其機率密度函數沒有封閉解，有時容易得到較不準確的參數估計值，進而影響產品壽命資訊的推估。因此，本文將基於Tweedie衰變模型之概似函數，透過模擬研究，比較最大概似估計方法以及文獻上三種變形之likelihood估計方法（quasi-likelihood、pseudo-likelihood、saddle-point approximation）之點估計準確度，以及根據點估計結果使用Inverse Laplace Transformation計算壽命分配的百分位數進行比較。本研究亦考量並比較各種估計方法在不同的樣本數、觀察數配置下之表現。

Degradation test is an efficient way to obtain high reliability products’ lifetime information. In this study, we use Tweedie model to describe the degradation path of the product’s physical or chemical characteristics. Estimation and inference of Tweedie model based on the maximum likelihood method are challenged by the presence of an infinity sum in the probability function. Some researchers proposed several approaches to fit Tweedie degradation model, like quasi-likelihood 、pseudo-likelihood 、saddle-point approximation. We use simulation study to compare these likelihood-based methods. Simulation results include parameter estimations and inference of lifetime of products.

Balakrishnan, N., & Qin, C. (2018). Nonparametric evaluation of the first passage time of degradation processes. Applied Stochastic Models in Business and Industry.
Bonat, W. H., & Kokonendji, C. C. (2017). Flexible tweedie regression models for continuous data. Journal of Statistical Computation and Simulation, 87(11), 21382152.
Chiu, W.T. (2019). Lifetime distribution and model selection for degradation data.
Duan, F., & Wang, G. (2018). Exponentialdispersion degradation process models with random effects and covariates. IEEE Transactions on Reliability, 67(3), 11281142.
Dunn, P. K. (2017). Evaluation of tweedie exponential family models [Computer software manual]. (R package version 2.3.2)
Jørgensen, B. (1997). The theory of dispersion models. Chapman and Hall/CRC.
Jørgensen, B., Demétrio, C. G., Kristensen, E., Banta, G. T., Petersen, H. C., & Delefosse, M. (2011). Biascorrected pearson estimating functions for taylor＇s power law applied to benthic macrofauna data. Statistics Probability Letters, 81(7), 749 758. (Statistics in Biological and Medical Sciences)
Meeker, W., & Escobar, L. (1998). Statistical methods for reliability data. Wiley.
Tseng, S.T.,& Lee, I.C. (2016). Optimum allocation rule for accelerated degradation tests with a class of exponentialdispersion degradation models. Technometrics, 58(2), 244254.
Yang, G. (2007). Life cycle reliability engineering.
Zhou, S., & Xu, A. (2019). Exponential dispersion process for degradation analysis. IEEE Transactions on Reliability, 68(2), 398409.