Dynamic Bayesian Networks for Feature Learning and Transfer Applications in Remaining Useful Life Estimation

Prognostics, in other words, remaining useful life (RUL) estimation is a core task of prognostics and health management (PHM). Reliable RUL predictions can reduce maintenance costs, improve production efficiency, and avoid unexpected downtime. Lots of models for RUL predictions have been proposed; however, noise and the nonlinear nature of degradation phenomena often leads to poor prognostics results, and the acquired engineered system data are usually subject to a high level of uncertainty. This makes the RUL estimation models less than satisfactory. Accurate RUL estimation and prediction not only rely on an accurate model but also depend on the adjustments of model parameters to track the variation. In this paper, an ensemble model combining the health index synthesis (HIS) approach and improved particle filtering (PF) is introduced. HIS approach was used to obtain the synthesized health index (SHI) for an engineered system with multiple sensors, which indicated the system's degradation model, while the improved PF approach was used to adjust the parameters of the degradation model obtained from the HIS approach and optimized the RUL estimation results. The performance of the prognostics approach introduced in this paper was demonstrated by using turbofan engine degradation data sets, which was supplied by NASA Ames, and results were compared with several usually used methods.

Degradation data-driven approach for remaining useful life estimation

Remaining useful life (RUL) estimation is the key of prognostics and health management (PHM) technology and is an effective way to ensure the safe and reliable operation of equipment. Aiming at the lack of historical data and prior information for the newly developed small-sample systems, an adaptive RUL estimation method based on the expectation maximization (EM) algorithm is proposed with three-source variability. First, a degradation model based on a Wiener process is established to incorporate three-source variability and dynamic sampling interval, and the analytical solution of RUL distribution is derived in the sense of the first hitting time. Second, an adaptive parameter estimation method based on the EM algorithm is proposed to update the model parameters by using the condition monitoring (CM) data from one working system running up to the current moment. Finally, a practical example of a gyroscope in an inertial navigation system is provided to substantiate the effectiveness and superiority of the proposed method. The results indicate that the proposed method can efficiently improve the accuracy of the RUL estimation.

Aircraft are highly valuable assets and large budgets are spent in preventive and predictive maintenance programs. The application of PHM (Prognostics and Health Management) technologies can be a powerful decision support tool to help maintenance planners. The RUL (Remaining Useful Life) estimations obtained from a PHM system can be used in order to plan in advance for the repair of components before a failure occurs. However, when system architecture is not taken into account, the use of PHM information may lead the operator to rush to replace a component that would not affect immediately the operation of the system under consideration. This paper presents a methodology for decision support in maintenance planning with application in aeronautical systems. The methodology combines system architecture information and RUL estimations for all components comprised in the system under study, allowing the estimation of a RUL value for the whole system. This system level RUL (S-RUL) can be used as support information for identifying the best moment to repair a component. Also, when several components present high degradation levels, the proposed methodology can be used to define a set of components that, when repaired, will bring the whole system to a safe degradation level with lowest cost. A case study is used to illustrate the application of the methodology in a simplified aircraft electrical system.

Reliability of Gallium Nitride (GaN) power semiconductor transistors are extremely important to uncover its immense possibility in high power conversion applications. Accurate remaining useful life (RUL) estimation ensures robust operational reliability of GaN FET in high power applications. Fault precursor trajectory based RUL estimation methods results in large estimation error due to both hybrid tendency of the trajectories and presence of measurement noise. Tracking and projecting fault precursor trajectory for RUL estimation is another crucial challenge when the system condition changes suddenly. In this paper, an epsilon support vector regression ( $\boldsymbol{\varepsilon}$ -SVR) aided auxiliary particle filter (APF) is applied for RUL estimation of GaN field-effect transistor (FET). On-state resistance $(\boldsymbol{R}_{\boldsymbol{DS},\boldsymbol{ON}})$ is used as the fault precursor for this study. $\boldsymbol{R}_{\boldsymbol{DS},\boldsymbol{ON}}$ of GaN FETs show dynamic characteristics during switching operation and results in false alarm in degradation identification. However, Steady state $\boldsymbol{R}_{\boldsymbol{DS},\boldsymbol{ON}}$ shows strong relationship with degradation. In this paper, a signal conditioning circuit is used to acquire $\boldsymbol{R}_{\boldsymbol{DS},\boldsymbol{ON}}$ with proper resolution. $\boldsymbol{\varepsilon}$ -SVR-assisted-APF also accurately estimates the RUL when operating condition changes suddenly. The failure region of GaN FET is modeled which shows the failure probability after $\boldsymbol{R}_{DS, ON}$ reaches the threshold value. The effectiveness of $\boldsymbol{\varepsilon}$ -SVR-assisted-APF is validated using accelerated ageing data under power cycling test.

Remaining Useful Life (RUL) estimation of rotating machinery based on their degradation data is vital for machine supervisors. Deep learning models are effective and popular methods for forecasting when rotating machinery such as bearings may malfunction and ultimately break down. During healthy functioning of the machinery, however, RUL is ill-defined. To address this issue, this study recommends using anomaly monitoring during both RUL estimator training and operation. Essential time-domain data is extracted from the raw bearing vibration data, and deep learning models are used to detect the onset of the anomaly. This further acts as a trigger for data-driven RUL estimation. The study employs an unsupervised clustering approach for anomaly trend analysis and a semi-supervised method for anomaly detection and RUL estimation. The novel combined deep learning-based anomaly-onset aware RUL estimation framework showed enhanced results on the benchmarked PRONOSTIA bearings dataset under non-varying operating conditions. The framework consisting of Autoencoder and Long Short Term Memory variants achieved an accuracy of over 90% in anomaly detection and RUL prediction. In the future, the framework can be deployed under varying operational situations using the transfer learning approach.

Computational framework for real-time diagnostics and prognostics of aircraft actuation systems

The remaining useful life (RUL) estimation is an important metric that helps in condition-based maintenance. Damage data obtained from the diagnostics techniques are often noisy, and the RUL estimated from the data is less reliable. Estimating the probabilistic RUL by quantifying the uncertainty in the predictive model parameters using the noisy data increases confidence in the predicted values. In this work, Bayesian inference is used to inversely quantify the uncertainty in the model parameters of the predictive damage model in adhesively bonded joints, and quantified parameter uncertainty is represented by a probability distribution function (PDF). It is difficult to obtain a closed-form solution for the RUL from the PDF of model parameters due to the nonlinearity of the damage models. Instead, statistical simulations are used to generate the model parameter samples from the PDF, and the RUL is estimated from the samples. In the statistical simulations, the predictive damage model is evaluated for each parameter sample. Using a physics-based model as the predictive model increases computational time, which makes the statistical techniques to be computationally expensive. It is essential to reduce the computational time to estimate the RUL in a feasible time. In this work, real-time probabilistic RUL estimation is demonstrated in adhesively bonded joints using the Sequential Monte Carlo (SMC) sampling method and cloud-based computations. The SMC sampling method is an alternative to traditional Markov Chain Monte Carlo (MCMC) methods, which enables generating statistical parameter samples in parallel. The parallel computational capabilities of the SMC methods are exploited by running the SMC simulation on multiple cloud calls. This approach is demonstrated by estimating fatigue RUL in the adhesively bonded joint. The accuracy of probabilistic RUL estimated by SMC is validated by comparing it with RUL estimated by the MCMC and the experimental values. The SMC simulation is run on the cloud and the computational speedup of the SMC is demonstrated.

Battery prognostics and health management (PHM) is essential for lithium-ion batteries in electric vehicles. In the battery PHM, accurate estimation of the battery state-of-health (SOH) and prediction of the remaining useful life (RUL) are crucial to ensure a safe and efficient battery operation. This article presents a probabilistic method for the battery degradation modeling and health prognosis based on the features extracted from the charging process using the dynamic Bayesian network (DBN). First, an aggregated feature, combining the incremental capacity analysis of constant-current (CC) charging and the time constant of constant-voltage (CV) charging, is developed to characterize the battery degradation dynamics in case some CC or CV charging information is absent. The DBN is then employed to explore the underlying correlation between the battery health indicator and the extracted features. The proposed model treats the degradation dynamics as a rich family of probability distributions to model real-world battery operation more accurately. Moreover, the battery SOH estimation and RUL prediction are carried out using the particle filtering inference algorithm. Experimental tests are conducted on two different battery cells and the results show that the proposed methods can provide an accurate and robust battery SOH estimation and reliable RUL prediction.

Remaining useful life (RUL) estimation is considered as one of the most central points in the prognostics and health management (PHM). The present paper describes a nonlinear hybrid ABC–MARS-based model for the prediction of the remaining useful life of aircraft engines. Indeed, it is well-known that an accurate RUL estimation allows failure prevention in a more controllable way so that the effective maintenance can be carried out in appropriate time to correct impending faults. The proposed hybrid model combines multivariate adaptive regression splines (MARS), which have been successfully adopted for regression problems, with the artificial bee colony (ABC) technique. This optimization technique involves parameter setting in the MARS training procedure, which significantly influences the regression accuracy. However, its use in reliability applications has not yet been widely explored. Bearing this in mind, remaining useful life values have been predicted here by using the hybrid ABC–MARS-based model from the remaining measured parameters (input variables) for aircraft engines with success. A correlation coefficient equal to 0.92 was obtained when this hybrid ABC–MARS-based model was applied to experimental data. The agreement of this model with experimental data confirmed its good performance. The main advantage of this predictive model is that it does not require information about the previous operation states of the aircraft engine.

Remaining useful life estimation based on discriminating shapelet extraction

A New Approach to Learning in Dynamic Bayesian Networks (DBNs)

Remaining useful life estimation based on nonlinear feature reduction and support vector regression

As one of the most crucial issues in the condition monitoring of mechanical systems, the remaining useful life (RUL) estimation of bearings suffers from the limitation of run-to-failure data, which makes the cross-operating condition RUL estimation of bearings a challenging subject. To address this issue, a cross-operating condition degradation knowledge learning method is proposed by constructing a shared latent feature space across different operating conditions, in which the bearing degradation features are effectively extracted and the cross-domain feature differences are weakened. Focusing on learning the cross-operating condition degradation knowledge, the time-frequency representation (TFR) samples generated by the continuous wavelet transform (CWT) method in both the source and target domains were projected onto a shared latent feature space constructed by a novel joint dictionary matrix factorization (JDMF) method. The joint-domain dictionaries solved by the JDMF method have played a core role in extracting the domain-invariant features that do not change along with the operating conditions of bearings. As a result, the convolutional neural network (CNN) model trained with the source samples in the shared latent feature space can be directly generalized to the target bearings for the degradation indicator (DI) modeling and RUL estimation. The cross-operating condition bearing RUL estimation experiment executed in the PRONOSTIA platform has demonstrated the effectiveness and superiority of the proposed JDMF-CNN framework.

A multi-model data-fusion based deep transfer learning for improved remaining useful life estimation for IIOT based systems