Multiple change point detection seeks to identify potential shifts in the data properties. While existing detection methods primarily focus on univariate and multivariate data, they often fall short in detecting variations in profile data, which represent the functional relationships between explanatory and response variables. This paper introduces a novel change point detection method tailored for profile data, employing a smooth profile decomposition (SPD) strategy that accommodates arbitrary designs and heteroscedasticity. This approach facilitates a comprehensive representation of both overall trends and fluctuations within the profiles. Furthermore, we propose an order-splitting screening estimator (OSE) to construct a detection statistic, allowing for precise estimation of change points while ensuring a significant theoretical guarantee regarding the false discovery rate (FDR). We validate the performance and robustness of the proposed method through numerical experiments and a real case study involving wind turbines.

Fault diagnosis based on data-driven methods are widely investigated when enough supervised samples of the target machine are available to build a reliable model. However, the labeled samples in practical operated machine are usually scarce and difficult to collect. If the model is built based on the sufficient labeled samples from different source machines, the diagnosis performance will degenerate owing to the domain discrepancy. To solve this issue, in this article, transfer learning (TL) is proposed by leveraging knowledge learned from source domain to target domain. While TL methods for fault diagnosis have been actively studied, most of them focus on learning from a single source. Since the labeled samples can come from multiple domains, more general diagnosis knowledge can be learned, which is beneficial to the prediction for the target domain. Therefore, a new TL approach based on multisource domain adaptation is proposed. A multiadversarial learning strategy is utilized for obtaining feature representations, which are invariant to the multiple domain shifts and discriminative for the learning goal at the same time. Extensive experimental analysis on four different bearing datasets is performed to illustrate the effectiveness and advantage of the proposed method.

Owing to distribution discrepancy between source training and target testing data, the performance of fault diagnosis by traditional supervised learning models will degenerate. Though domain adaptation methods for diagnosis have been actively investigated recently, most of them are devoted to learning from a single source. However, in reality, the supervised samples can be collected from different sources such as various working conditions. These sources are not only different from target but also from each other. The way of effectively fusing these sources to contribute the prediction of target remains a challenge. In this work, a new framework of multi-source domain adaptation is proposed for cross-domain fault diagnosis under different working conditions. Specially, this framework is realized by two alignment stages. At the first stage, multiple specific feature spaces are obtained, then the distributions of each pair of source and target domain are aligned since it is difficult to extract the common domain-invariant features for all domains. At the second stage, by considering the domain specific decision boundaries, the probabilistic outputs of classifiers are also aligned. Various experimental analysis on four different bearing working conditions is conducted to show the effectiveness of the proposed method. The performance of the proposed method is superior to state-the-art cross-domain fault diagnosis methods.

Stochastic processes and filtering methods are popular tools for degradation modeling and online remaining useful life (RUL) prediction. However, most models are for one-dimensional degradation and various filtering methods can only handle observations from a single system. This paper studies joint online RUL prediction of multideteriorating systems with multisystem observations and measurement errors. A multivariate degradation model equipped with a batch particle filter is developed and built for characterizing multiple dependent performance deteriorations with measurement errors in each system. The batch particle filter is developed for simultaneous online parameter estimation and degradation state identification by leveraging multisystem observations. A numerical example and a case study are provided to demonstrate the proposed method. The results show that homogeneous multisystem observations from a population of multideteriorating systems can be jointly processed on-the-fly. Individualized online RUL prediction with improved precision for each system can be achieved through the joint online inference.

Condition-based maintenance (CBM) is proved to be effective in reducing the long-run operational cost for a system subject to degradation failure. Most existing research on CBM focuses on single-unit systems where the whole system is treated as a black box. However, a system usually consists of a number of components and each component has its failure behavior. When degradation of the components is observable, CBM can be applied to the component level to improve the maintenance efficiency. This paper aims to study the optimal inspection/replacement CBM strategy for a multi-unit system. Degradation of each component is assumed to follow a Wiener process and periodic inspection is considered. We cast the problem into a Markov decision framework and derive the optimal maintenance decisions that minimize the maintenance cost. To better illustrate the optimal maintenance strategy, we start from a 1-out-of-2: G system and show that the optimal maintenance policy is a two-dimensional control limit policy. The argument used in the 1-out-of-2: G system can be readily extended to general cases in a similar way. The value iteration algorithm is used to find the optimal control limits, and the optimal inspection interval is subsequently determined through a one-dimensional search. A numerical study and a comprehensive sensitivity analysis are provided to illustrate the optimal maintenance strategy.

We propose a quantile regression framework to model accelerated life tests (ALT) data. The quantile of the failure time distribution at the usage level can be easily estimated using quantile regression. Compared with traditional parametric regression methods, quantile regression is distribution-free, efficient in the presence of censoring, and more flexible in modeling ALT relations. More importantly, we show that it is able to handle ALT data with a failure-free life, which is a great challenge in the ALT literature. We use extensive simulation studies and two real ALT case studies to demonstrate the effectiveness of the proposed method.

Developing prognostics and health management (PHM) approaches for lithium-ion batteries has received increasing attention in recent years. This paper presents a new modeling framework to characterize lithium-ion battery degradation by examining detailed discharging voltage profiles in different discharging cycles. We propose a hierarchical model, combining discharging processes and degradation processes, to predict the end of discharges in different cycles and remaining useful cycles integratively. We use a real case study to demonstrate the effectiveness and promising features of the proposed framework.