Life prediction of multiple performance parameters ADT based on multidimensional time series analysis

For long lifetime and high reliability products, it is difficult to obtain failure data in a short time period. Hence, Accelerated Degradation Testing (ADT) is presented to deal with the cases where no failure time data could be obtained but degradation data of parameters of the product are available. At present, the ADT life prediction method is utilized primarily with feedback from a single performance parameter ADT dataset. However, for most products, multiple performance parameters of these products will degrade with time, leading to failure. It is important to note that often the products various performance parameters will interact with each other as the performance degrades. Hence, a correct life prediction based on ADT data must take into account the integrated effect of a product's multiple performance parameters and the random effect of environmental variables. In the literature, such as in the noted references [1-5], ADT life prediction is studied using time series methods due to its excellent capability of stochastic and periodic information mining. However, life predictions using the time series method in present literature are all based upon a one-dimensional time series analysis. To take into account multiple dimensions of product performance degradation, it is important to study these parameters using an ADT life prediction based on a multidimensional time series analysis method.