•A new scope of sequential RUL prediction compared with conventional point-wise prediction.•A surrogate modeling framework for uncertainty propagation using drifted Wiener processes.•Hybrid modeling ...between machine learning and statistics•Improve the near-failure prediction accuracy by sacrificing the far-to-failure prediction accuracy.
In modern industrial systems, sensor data reflecting the system health state are commonly used for the remaining useful lifetime (RUL) prediction, which are increasingly processed by modern deep learning based approaches recently. But these deep learning models do not automatically provide uncertainty information for the RUL prediction, hence this paper is motivated to introduce a novel approach that allows to control trade-off between prediction performance and knowledge about the uncertainty of the RUL prediction. The key aspect of our approach is to use a long short-term memory (LSTM) network as an expressive black-box predictor and the Wiener process as a surrogate to model the propagation of prediction uncertainty. The uncertainty propagation model is used to interactively train the RUL predictor. Our empirical results in a turbofan engine degradation simulation use case show that the surrogate Wiener propagation model can improve the near-failure prediction accuracy by sacrificing the far-to-failure prediction accuracy.
Accurate assessment of Remaining Useful Lifetime (RUL) of power transformers plays an essential role in improving the reliability of power grid while reducing operating costs. A dynamical RUL ...prediction technique is proposed in this article through analyzing the individual difference by multi-dimensional fusion of condition monitoring data. Firstly, the on-site Dissolved Gas Analysis (DGA) measurements and the calculated Degree of Polymerization (DP) are fused with the method of Analytic Hierarchy Process (AHP) and Variable Weight Principle (VWP) to construct the power transformer Life Correlation Index (LCI). Secondly, the parameters of Wiener model are updated iteratively by Bayesian rule and Expectation Maximization (EM) algorithm, to obtain the distribution and prediction of residual transformer lifetime. The residual lifetimes of four 500 kV power transformers in operation are calculated, and the predicted remaining lifetimes are 24.73 years, 30.66 years, 2.79 years, and 0.07 years respectively. The effectiveness and reliability are demonstrated by the worst transformer case with the predicted RUL of 0.07 years, in which a short circuit fault of that transformer happened immediately after power-off maintenance. In particular, this method is able to predict the residual lifetime of transformers in a dynamic way, which provides significant practical implications for the assessment management of power transformers.
To avoid unexpected failures of units in manufacturing systems, failure mode recognition and prognostics are critically important in prognostics health management (PHM). Most existing methods either ...ignored the effects of various failure modes on remaining useful lifetime (RUL) prediction or implemented failure mode recognition and RUL prediction as two independent tasks, which failed to exploit failure mode information to obtain accurate RUL prediction. In fact, RUL highly depends on failure modes because sensor signals under different failure modes usually present different degradation patterns. To address the issue, this paper proposes a joint learning model of failure mode recognition and RUL prediction for degradation processes based on multiple sensor signals. The proposed joint learning model first extracts features by considering the degradation mechanism to ensure good interpretability for degradation modeling, and then takes the extracted features as inputs to a deep neural network. By conducting failure mode recognition and RUL prediction as a collaborative task, the proposed model can fully characterize the complex relationship among the extracted features, RUL and failure modes, and outputs the recognized failure modes and the predicted RUL of units simultaneously. A case study on the degradation of aircraft gas turbine engines is presented to evaluate the proposed model performance. Note to Practitioners -The paper aims to develop a joint learning method for failure mode recognition and RUL prediction of operating units. Specifically, the developed method addresses a challenging issue in practice, i.e., how to effectively conduct failure mode recognition and RUL prediction as a joint task based on interpretable extracted degradation features from multiple sensor signals. To implement this method in practice, four steps are included as follows: First, collect multiple sensor signals, failure time, and failure modes of historical units. Second, construct the joint learning model based on features extracted from sensor signals by considering the degradation mechanism. Third, estimate model parameters using the data of historical units. Fourth, recognize the failure mode and predict the RUL of an in-service unit. Since the proposed method is a data-driven neural network with flexible model structure that considers complex data relationships, it is expected to be applicable to many practical situations and use cases, especially for manufacturing systems with complex structures and unknown failure thresholds.
To monitor the degradation status of units and prevent unexpected failures in engineering systems, health index (HI)-based data fusion technologies have been rapidly developed by combining multiple ...sensor signals, which are helpful to understand the degradation processes of units and predict their remaining useful lifetime (RUL). Although promising, existing HI-based data fusion models for degradation modeling are still limited due to the restrictive assumptions made during the fusion or the degradation modeling processes, e.g., assuming the fusion model as a linear or kernel-based function from multiple sensor signals, or modeling the degradation process by a preselected basis function. Such assumptions are often invalid in industrial practice and may fail to accurately characterize the complicated relationships between multiple sensor signals and the underlying degradation process. To address the issue, this article proposes a generic indirect deep learning method that constructs an HI by combining multiple sensor signals to better characterize the degradation process. In particular, our innovative idea is to seamlessly integrate a deep neural network (DNN) and a long short term memory (LSTM) model to construct the HI by fusing multiple sensor signals and characterize the degradation process, which can be applied to the degradation modeling of various engineering systems. Domain knowledge including the concept of failure threshold and monotonicity of the degradation process is also considered to enhance the interpretability of the proposed method. For parameter estimation, we develop an indirect gradient descent (IGD) algorithm to train the proposed method. Simulation studies and a case study on the degradation of aircraft gas turbine engines are presented to validate the performance of the proposed method.
Abstract In various industry sectors, predicting the real-life availability of milling applications poses a significant challenge. This challenge arises from the need to prevent inefficient blade ...resource utilization and the risk of machine breakdowns due to natural wear. To ensure timely and accurate adjustments to milling processes based on the machine's cutting blade condition without disrupting ongoing production, we introduce the Fused Data Prediction Model (FDPM), a novel temporal hybrid prediction model. The FDPM combines the static and dynamic features of the machines to generate simulated outputs, including average cutting force, material removal rate, and peripheral milling machine torque. These outputs are correlated with real blade wear measurements, creating a simulation model that provides insights into predicting the wear progression in the machine when associated with real machine operational parameters. The FDPM also considers data preprocessing, reducing the dimensional space to an advanced recurrent neural network prediction algorithm for forecasting blade wear levels in milling. The validation of the physics-based simulation model indicates the highest fidelity in replicating wear progression with the average cutting force variable, demonstrating an average relative error of 2.38% when compared to the measured mean of rake wear during the milling cycle. These findings illustrate the effectiveness of the FDPM approach, showcasing an impressive prediction accuracy exceeding 93% when the model is trained with only 50% of the available data. These results highlight the potential of the FDPM model as a robust and versatile method for assessing wear levels in milling operations precisely, without disrupting ongoing production.
•A novel sequence-to-sequence sliding window approach is proposed for better mapping of the input data to the predicted Remaining Useful Lifetime Estimation.•Bidirectional LSTM networks are used as ...the based architecture for the sequence-to-sequence modelling task.•The proposed architecture is augmented with a convolutional layer and different attention techniques to improve the prediction performance.•A thorough comparison of proposed models to analyse the effect of the different architectural settings.•State of the art results on the benchmark C-MAPSS dataset.
We propose a novel sequence-to-sequence prediction approach for the estimation of the remaining useful lifetime (RUL) of technical components. The approach is based on deep recurrent neural network structures, namely bidirectional Long Short Term Memory (LSTM) networks, which we augment with an attention mechanism to allow for a more fine-grained information flow between the input and output sequence. Using the base architecture as a reference, we experiment with various forms of attention mechanisms as well as different forms of additional input embeddings. Further, we analyse the impact of the sequence length on the estimation quality. We apply our approach to the well known C-MAPSS data set previously serving as a benchmark dataset for RUL prediction. We obtain state of the art results on the data set and provide a thorough hyperparameter study that underlines, that more simple but well tuned architecture can achieve comparable or better performance than highly complex architectures.
In data-driven methods for prognostics, the remaining useful lifetime (RUL) is predicted based on the health indicator (HI). The HI detects the condition of equipment or components by monitoring ...sensor data such as vibration signals. To construct the HI, multiple features are extracted from signals using time domain, frequency domain, and time–frequency domain analyses, and which are then fused. However, the process of selecting and fusing features for the HI is very complex and labor-intensive. We propose a novel time–frequency image feature to construct HI and predict the RUL. To convert the one-dimensional vibration signals to a two-dimensional (2-D) image, the continuous wavelet transform (CWT) extracts the time–frequency image features, i.e., the wavelet power spectrum. Then, the obtained image features are fed into a 2-D convolutional neural network (CNN) to construct the HI. The estimated HI from the proposed model is used for the RUL prediction. The accuracy of the RUL prediction is improved by using the image features. The proposed method compresses the complex process including feature extraction, selection, and fusion into a single algorithm by adopting a deep learning approach. The proposed method is validated using a bearing dataset provided by PRONOSTIA. The results demonstrate that the proposed method is superior to related studies using the same dataset.
Smart transformers (STs) are challenged by high-reliability requirements. Particularly, modular STs consist of a large number of power semiconductors, and unevenly distributed thermal stress results ...in different remaining useful lifetimes of the devices in the different building blocks. This problem can be addressed with discontinuous modulation for the medium-voltage-side converter, but results in a coupled thermal stress with the isolated dc-dc cells. For overcoming this drawback, this article proposes an advanced discontinuous modulation, which enables to manipulate thermal stress in different cells independently. The algorithm is analyzed with respect to its capability in thermal stress manipulation and validated experimentally with junction temperature measurements. In addition, this article investigates the impact of the algorithm on the overall system's efficiency as well as potential lifetime of the power semiconductors in the different building blocks.
The increasing demand for customized products is a core driver of novel automation concepts in Industry 4.0. For the case of machining complex free-form workpieces, e.g., in die making and mold ...making, individualized manufacturing is already the industrial practice. The varying process conditions and demanding machining processes lead to a high relevance of machining domain experts and a low degree of manufacturing flow automation. In order to increase the degree of automation, online process monitoring and the prediction of the quality-related remaining cutting tool life is indispensable. However, the varying process conditions complicate this as the correlation between the sensor signals and tool condition is not directly apparent. Furthermore, machine learning (ML) knowledge is limited on the shop floor, preventing a manual adaption of the models to changing conditions. Therefore, this paper introduces a new method for remaining tool life prediction in individualized production using automated machine learning (AutoML). The method enables the incorporation of machining expert knowledge via the model inputs and outputs. It automatically creates end-to-end ML pipelines based on optimized ensembles of regression and forecasting models. An explainability algorithm visualizes the relevance of the model inputs for the decision making. The method is analyzed and compared to a manual state-of-the-art approach for series production in a comprehensive evaluation using a new milling dataset. The dataset represents gradual tool wear under changing workpieces and process parameters. Our AutoML method outperforms the state-of-the-art approach and the evaluation indicates that a transfer of methods designed for series production to variable process conditions is not easily possible. Overall, the new method optimizes individualized production economically and in terms of resources. Machining experts with limited ML knowledge can leverage their domain knowledge to develop, validate and adapt tool life models.
In the modern age of digitalization, electronics are fundamental to any engineering system. With the current strong focus on the Internet of Things (IoT), autonomous vehicles and Industry 4.0, ...reliable electronics are gaining crucial importance. Predicting the health of complex systems is able to avoid catastrophic failures. Prognostic and Health Monitoring (PHM) approaches are an important step toward trustable and reliable electronics. Nowadays, Artificial Intelligence (AI) and machine learning (ML) algorithms are integrated into PHM approaches, enabling complex fault diagnosis. In this contribution, we provide an overview of the application of intelligent algorithms in PHM of electronics in a systematic manner. The challenges of prognostics in electronics are provided and a detailed overview of the available PHM precursors for various electronic components and the associated selection process is given. Based on the literature review conducted, the main research challenges with ML algorithms in PHM are discussed along with performances of each model.