Search

Remaining useful life (RUL) prediction of aircraft engine (AE) is of great importance to improve its reliability and availability, and reduce its maintenance costs. This article proposes a novel deep bidirectional recurrent neural networks (DBRNNs) ensemble method for the RUL prediction of the AEs. In this method, several kinds of DBRNNs with different neuron structures are built to extract hidden features from sensory data. A new customized loss function is designed to evaluate the performance of the DBRNNs, and a series of the RUL values is obtained. Then, these RUL values are reencapsulated into a predicted RUL domain. By updating the weights of elements in the domain, multiple regression decision tree (RDT) models are trained iteratively. These models integrate the predicted results of different DBRNNs to realize the final RUL prognostics with high accuracy. The proposed method is validated by using C-MAPSS datasets from NASA. The experimental results show that the proposed method has achieved more superior performance compared with other existing methods.

Deep-learning-based methods have been successfully applied to fault diagnosis of rotating machinery. However, the domain mismatch among different operating conditions significantly deteriorates diagnostic performance of these methods in industrial applications. To solve this problem, a new fault diagnosis model based on capsule neural network (Cap-net) is constructed, and a novel online domain adaptation learning method based on deep reinforcement learning (DRL) is proposed in this article to improve the adaptivity of the fault diagnosis model. In this method, the Cap-net is first introduced into the DRL as an agent to extract representative features and diagnoses fault. Moreover, the online domain adaptation learning of the agent is conducted based on the Q-learning of the DRL so as to adapt to different operating domains that have never been experienced. Specifically, an online feature dictionary combined with cosine similarity is designed to coarsely label the online data collected from different operating domain, while a reward mechanism is defined to evaluate the obtained label. Subsequently, the online data, the corresponding label, and the reward are used to optimize the agent to obtain the desired diagnostic model. Two experiment studies are implemented to verify the effectiveness of the proposed method. The experimental results show that the proposed method has more excellent diagnostic performance and adaptivity than the existing popular methods.

Remaining useful life (RUL) prediction is a key solution to improve the reliability, availability and maintainability of engineering systems. Long short-term memory (LSTM) and convolution neural networks (CNN) are the current hotspots in the field of RUL prediction. However, the LSTM-based prognostic approach has a slow loop step to process large-scale time-series data since the dependence of the data processing process at each time on the output of the previous time limits parallelism, and the CNN-based prognostic approach is not fit for time-series data although it can process the data in parallel. In this paper, a new auto-encoder quasi-recurrent neural networks (AEQRNN) based prognostic approach is proposed for RUL prediction of the engineering systems. The AEQRNN contains convolution components that can process input data in parallel, and pooling components which has two LSTM-like gate structures to process time-series data. In addition, the AEQRNN can automatically extract hidden features from monitoring signals without manual feature design. The effectiveness of the proposed prognostic approach is validated by three prognostic benchmarking datasets, including a turbofan engine dataset, a rolling bearing dataset, and a machining tool dataset. Experimental results demonstrate that this approach has both superior prognostic performance and training speed in comparison with other kinds of recurrent neural network-based approaches and various state-of-the-art approaches in the recent literature.

Remaining useful life (RUL) prognosis is of great significance to improve the reliability, availability, and maintenance cost of an industrial equipment. Traditional machine learning method is not fit for dealing with time series signals and has low generalization and stability in prognostic. In this article, a novel ensemble long short-term memory neural network (ELSTMNN) model for RUL prediction is proposed to enhance the RUL prognosis accuracy and improve the adaptive and generalization abilities under different prognostic scenarios. The ELSTMNN contains a series of long short-term memory neural networks (LSTMNNs), each of which is trained on a unique set of historical data. A novel ensemble method is first proposed using Bayesian inference algorithm to integrate multiple predictions of the LSTMNNs for the optimal RUL estimation. The effectiveness of the ELSTMNN-based RUL prognosis method is validated using two characteristically different turbofan engine data sets. The experimental results show a competitive performance of the ELSTMNN in comparison with other prognostic methods.

Fault diagnosis is very important to ensure the efficiency and reliability of rotating machinery. Traditional fault diagnosis methods often require manual feature design and extraction, which is an exhausting work in industrial applications. Convolutional neural network (CNN)-based methods have presented a great potential to automatically extract and select representative features for fault diagnosis. In this article, a novel three-step intelligent fault diagnosis method is proposed based on CNN and Bayesian Gaussian mixture (BGM) for rotating machinery. In the fault dataset construction step, multiple binary training datasets are constructed to amplify the difference between one fault mode and the others using a defined fault labeling rule. In the fault feature extraction step, multiple binary classification tasks are, respectively, implemented to learn the representative features for each fault mode through a proposed group of CNN models based on the above-obtained training datasets. In the fault diagnosis step, multiple Gaussian mixture models (GMMs) are adopted to fit the data distributions of the learned fault features, and an BGM model based on GMM parameters and Bayesian network is designed to establish a cause-and-effect relationship between corresponding parameters and fault modes. Two case studies are used to evaluate the effectiveness of the proposed method. The experimental results show that the proposed method has a superior performance over the existing methods.

Accurate health evaluation is crucial to reliable operation of complex degradation systems. Although traditional machine learning methods such as artificial neural network (ANN) and support vector machine (SVM) have been used widely, state assessment schemes based on a single classification model still suffer from low multiclass classification efficiency, high variance, and deviation. To solve these problems, this article proposes a novel health evaluation method based on stacking ensemble learning and generalized multiclass support vector machine (GMSVM) algorithm. The proposed health evaluation framework includes three parts: 1) abnormal value elimination and missing value processing are applied for multiple sensor data; 2) statistical features are extracted from the observed data and the Pearson correlation coefficient is applied for feature selection; and 3) ensemble generalized multiclass support vector machines (EGMSVMs) are utilized to evaluate the health situation of a degradation system. Unlike the binary classifiers and deep-learning-based classifiers, EGMSVMs utilize the stacking-based method to combine several GMSVMs as submodels and random forest as a metamodel, and the metamodel ensembles the results of submodels to reach a satisfied performance. Compared to traditional SVM- and ANN-based algorithms, EGMSVMs, in processing multiclass problems, achieve high efficiency and, meanwhile, low variance and deviation. The proposed method is verified using a hydraulic test rig. The experimental results show the feasibility and applicability of the proposed health evaluation framework.

Machine health monitoring is of great importance in industrial informatics field. Recently, deep learning methods applied to machine health monitoring have been proven effective. However, the existing methods face enormous difficulties in extracting heterogeneous features indicating the variation until failure and revealing the inherent high-dimensional features of massive signals, which affect the accuracy and efficiency of machine health monitoring. In this paper, a novel data-driven machine health monitoring method is proposed using adaptive kernel spectral clustering (AKSC) and deep long short-term memory recurrent neural networks (LSTM-RNN). This method include three steps: First, features in the time domain, frequency domain, and time-frequency domain are, respectively, extracted from massive measured signals. And, an Euclidean distance based algorithm is designed to select degradation features. Second, the AKSC algorithm is introduced to adaptively identify machine anomaly behaviors from multiple degradation features. Third, a new deep learning model (LSTM-RNN) is constructed to update and predict the failure time of the machine. The effectiveness of the proposed method is validated using a set of test-to-failure experimental data. The results show that the performance of the proposed method is competitive with other existing methods.

Health degradation monitoring of rolling element bearing (REB) is of great significance to ensure safety and availability of mechanical equipment. This paper presents a new online health degradation monitoring method of REB based on growing self-organizing mapping (GSOM) and clustered support vector machine (CSVM). In the proposed method, multidimensional health degradation features of the REB are extracted to reflect health degradation process, including time-domain statistical features, frequency spectrum features, intrinsic mode function energy features, wavelet packet frequency band energy features. Multiple GSOMs are utilized to adaptively fuse each kind of the extracted features for the health indices. CSVM is constructed to achieve accurate health status identification of the REB. A health degradation experimental case of the REB is analyzed to demonstrate the effectiveness of the proposed method. The results show that the proposed method has obvious superior performance compared to other existing methods.

Time-to-failure (TTF) prognostic plays a crucial role in predicting remaining lifetime of electrical machines for improving machinery health management. This paper presents a novel three-step degradation data-driven TTF prognostics approach for rolling element bearings (REBs) in electrical machines. In the degradation feature extraction step, multiple degradation features, including statistical features, intrinsic energy features, and fault frequency features, are extracted to detect the degradation phenomenon of REBs using complete ensemble empirical mode decomposition with adaptive noise and Hilbert–Huang transform methods. In degradation feature reduction step, the degradation features, which are monotonic, robust, and correlative to the fault evolution of the REBs, are selected and fused into a principal component Mahalanobis distance health index using dynamic principal component analysis and Mahalanobis distance methods. In TTF prediction step, the degradation process and local TTF of the REBs are observed by an exponential regression-based local degradation model, and the global TTF is predicted by an empirical Bayesian algorithm with a continuous update. A practical case study involving run-to-failure experiments of REBs on PRONOSTIA platform is provided to validate the effectiveness of the proposed approach and to show a more accurate prediction of TTF than the existing major approaches.

A mixed-model two-sided assembly line is a manufacturing system designed for the production of large-sized products. In order to describe the actual condition, this paper presents a novel multiobjective programming model for balancing a mixed-model two-sided assembly line subject to multiple constraints, in which, additional constraints including zoning, synchronous, and positional constraints are considered besides the traditional constraints, e.g., the precedence constraint. Two objectives are simultaneously to be optimized, one is to minimize the combination of the weighted line efficiency and the weighted smoothness index, and the other is to minimize the weighted total relevant costs per unit of a product. A novel multiobjective hybrid imperialist competitive algorithm (MOHICA) is proposed to solve this problem. In the presented MOHICA, the sigma method is employed to quantify every individual, a novel merging method is introduced to reserve better individuals into the evolutionary population, and late acceptance hill-climbing (LAHC) algorithm is presented as a local search algorithm to achieve accurate balance between intensification and diversification. The experimental results on the selected benchmark instances and a practical case show that the proposed multiobjective algorithm outperforms nondominated sorting genetic algorithm (NSGA)-II, multiobjective improved teaching-learning-based optimization, and NSGA-III existing in the literature.