在可靠度試驗中，加速壽命試驗為一利用產品在較正常環境嚴厲的工作應力水平下退化速度較快的特性，達到減少試驗時間目的的一種方法；而步進加速壽命試驗則是將同一批物件置於逐漸增加的應力水準下的加速試驗。而在壽命試驗中資料又可分為紀錄各物件失時間的完整資料與只觀測失效的個數群組資料兩種。在提高試驗效率、降低成本的前提下；試驗數據往往並不完整而需要依賴符合產品可靠度退化機制之假設進行分析評估以取得其試驗結果。因此對於加速試驗數據的分析方法便成一重要的探討方向。 本文討論試驗資料之壽命分佈呈指數分佈下多階段應力、I型截尾的群組資料。由於資料數過少導致在統計上分析的困難。因此利用樣本所受之應力水平不斷提升的特性提出新的解決方法，並與過去之研究分析方式進行比較。進一步提出合適之分析方法與試驗規範，以取得產品之壽命評估結果；以供後續產品之改良與其維修保養策略訂定之依據。 In reliability test, accelerate life test, a method to achieve the goal of reducing test period utilizes products degradated under higher working stress level faster then ones under normal working stress level. Step stress testing is a special kind of ALT in which the stress on each specimen is increased step-by-step over time. Complete data, which record all failure times and grouped data, which only include the numbers of failure items are part types of data in life test. In this thesis, assume each product of lifetime follows an exponential distribution under the step stress accelerate life test Type-I censoring with grouped data. The few data cause Statistical analysis is very difficult. Therefore utilizes the characteristic that the product under stress level promotes unceasingly to propose that the new solution, and compare with the way of researching and analysing in the past. Finally, provide the suitable analytical method and experimental norm to predict products of lifetime. Provides the basis of the tactics of products improved and maintained.
