Search

This paper proposes hybrid methods using physics-informed (PI) lightweight Temporal Convolution Neural Network (PITCN) for bearings’ remaining useful life (RUL) prediction under stiffness degradation. It includes three PI hybrid models: a) PI Feature model (PIFM) — constructing physics-informed health indicator (PIHI) to augment the feature space, b) PI Layer model (PILM)—encoding the physics governing equations in a hidden layer, and c) PI Layer Based Loss model (PILLM)—designing PI conflict loss, taking into account the difference before and after integration of the physics input-output relations involved module to the loss function. We simulated 200 different bearing stiffness degradations, using their discrete monitored vibration signals to verify the effectiveness of the proposed method. We also investigate their inference process through feature heat map analysis to interpret how the models melt physics knowledge to assist in capturing the degradation trend. The physics knowledge considered in this paper is the dynamic relationship between vibration amplitude and stiffness in a damped forced vibration model. The results show that all three PITCN models effectively capture degradation-related trend information and perform better than the vanilla lightweight TCN. Furthermore, the visualization of the feature channels highlights the important role of physics information in model training. Channels containing physics information demonstrate higher correlation with results as they significantly dominate the heat map compared to other channels.

Sparse & noisy monitoring data leads to numerous challenges in prognostic and health management (PHM). Big data volume but poor quality with scarce healthy states information limits the performance of training machine learning (ML) and physics-based failure modeling. To address these challenges, this thesis aims to develop a new hybrid PHM framework with the ability to autonomously discover and exploit incomplete implicit physics knowledge in sparse & noisy monitoring data, providing a solution for deep physics knowledge-ML fusion by physics-informed machine learning algorithms. In addition, the developed hybrid framework also applies the self-supervised learning paradigm to significantly improve the learning performance under uncertain, sparse, and noisy data with lower requirements for specialist area knowledge. The performance of the developed algorithms will be investigated on the sparse and noise data generated by simulation data sets, public benchmark data sets, and the PHM platform to demonstrate its applicability.

Chemokines form a family of proteins with critical roles in many biological processes in health and disease conditions, including cardiovascular, autoimmune diseases, infections, and cancer. Many chemokines engage in heterophilic interactions to form heterodimers, leading to synergistic activity enhancement or reduction dependent on the nature of heterodimer-forming chemokines. In mixtures, different chemokine species with diverse activities coexist in dynamic equilibrium, leading to the observation of their combined response in biological assays. To overcome this problem, we produced a non-dissociating CXCL4–CXCL12 chemokine heterodimer OHD4–12as a new tool for studying the biological activities and mechanisms of chemokine heterodimers in biological environments. Using the OHD4–12, we show that the CXCL4–CXCL12 chemokine heterodimer inhibits the CXCL12-driven migration of triple-negative MDA-MB-231 breast cancer cells. We also show that the CXCL4–CXCL12 chemokine heterodimer binds and activates the CXCR4 receptor.

During the two last decades, the failure prognostics has been given an increasing importance by the researchers and industrials thanks to the benefits its results can induce in terms of availability, security, and maintenance costs. However, a detailed study of the literature in the domain shows that most of the published works deal with prognostics at component level, where the failure of the system is linked to the failure of its critical components. Even if this assumption can be tolerated in some cases, it cannot be generalized. This is because, in practice, the components are interconnected and, consequently, a degradation of one component can affect the others leading to a modification in the system’s health state and then in its remaining useful life (RUL). To take into account this reality, a new model for prognostics is proposed in this article, namely inoperability input–output model (IIM). This model allows representing multicomponent systems while considering the interactions between their components. More than that, the IIM enables to integrate the mission profile effects on the system degradations to make its RUL predictions more accurate. To illustrate the effectiveness of the proposed modeling approach, a mechatronic system is considered as a case study.

Failure prognostics has become a central element in predictive maintenance. In this domain, the accurate determination of the remaining useful life (RUL) allows making effective maintenance and operation decisions about the assets. However, prognostics is often approached from a component point of view, and system-level prognostics, taking into account component interactions and mission profile effects, is still an underexplored area. To address this issue, we propose an online joint estimation and prediction methodology using a modeling framework based on the inoperability input–output model (IIM). This model can consider the interactions between components and also the mission profile effects on a system’s degradation. To estimate the system’s parameters in real-time, with a minimum of prior knowledge, an online estimation process based on the gradient descend algorithm is recursively performed when acquiring new measurements. After each update, the estimated model is used to predict the system RUL. The performance of the proposed approach is highlighted through different numerical examples. In addition, these developments are applied to a real industrial application, the Tennessee Eastman Process, in order to show their effectiveness.

Modern systems, that are often constituted of different interacting components, have increasing requirements for improving availability and safety when suffering unanticipated failures. Therefore, one needs to continuously monitor them, estimate and predict their health states through the implementation of prognostics and health management processes. To make this estimation and prediction more relevant and accurate, different issues, such as the component interactions, the profile mission impacts, and the uncertainties, should be considered. Compared to component level, these issues are more challenging in system-level prognostics. Their resolution is then essential to improve reliability, accuracy, and precision of the prognostics results. In this article, the problems of component interactions, profile mission impacts, and uncertainty quantification in system-level prognostics are addressed. For this purpose, a method using the modeling framework of the inoperability input–output model and composed of three steps is proposed. First, the system inoperability is estimated using particle filtering while considering the interactions between its components. Then, the estimated inoperability is propagated for long-term predictions. Finally, the system remaining useful life is calculated depending on the system configuration. The proposed methodology is applied to a mechatronic system to show its effectiveness.

This work aims to provide useful insights into the course of action and the challenges faced by machine manufacturers when dealing with the actual application of Prognostics and Health Management procedures in industrial environments. Taking into account the computing capabilities and connectivity of the hardware available for smart manufacturing, we propose a particular solution that allows meeting one of the essential requirements of intelligent production processes, i.e., autonomous health management. Indeed, efficient and fast algorithms, that does not require a high computational cost and can be appropriately performed on machine controllers, i.e., on edge, are combined with others, which can handle large amounts of data and calculations, executed on remote powerful supervisory platforms, i.e., on the cloud. In detail, new condition monitoring algorithms based on Model-of-Signals techniques are developed and implemented on local controllers to process the raw sensor readings and extract meaningful and compact features, according to System Identification rules and guidelines. These results are then transmitted to remote supervisors, where Particle Filters are exploited to model components degradation and predict their Remaining Useful Life. Practitioners can use this information to optimise production planning and maintenance policies. The proposed architecture allows keeping the communication traffic between edge and cloud in the nowadays affordable “Big data” range, preventing the unmanageable “Huge data” scenario that would follow from the transmission of raw sensor data. Furthermore, the robustness and effectiveness of the proposed method are tested considering a meaningful benchmark, the PRONOSTIA dataset, allowing reproducibility and comparison with other approaches.

Nowadays, advanced industrial robots are increasingly used and gradually replacing human activities in smart manufacturing that requires high precision and high performance. During this process, a small deviation of a robot axis can lead to other axes drifts, and then significantly affects the product quality. Hence, this paper aims to present an effective approach to monitor and diagnose the origin position deviations of multi-axis robots. The proposed method uses the encoder measurements of each axis to extract features and build appropriate health indicators. These obtained health indicators are then injected into a Machine Learning classifier to localize the origin of the deviation, i.e which axis causes these drifts. Furthermore, the performance of this method is verified through a real industrial test bench, used for machining, that investigates various deviation severities in different axes of the robot.

In literature, a major part of the prognostic studies considers the mission profile as a static parameter when evaluating the system Remaining Useful Life (RUL). However, in practice, the way in which a system operates significantly impacts the future evolution of its degradation. Therefore, this paper aims at evaluating the impact associated with the utilization of three different methods to characterize future operating conditions within the implementation of probability-based prognostic algorithms, namely Long-short term memory (LSTM), Markov Chain and Constant (or time-invariant) usage. These three methods are compared together in terms of both prognostic accuracy and essential update times when investigating the Time-of-Discharge (ToD) of an electric bicycle Lithium-Ion (Li-Ion) battery.

Model-based prognostic approaches use first-principle or regression models to estimate and predict the system’s health state in order to determine the remaining useful life (RUL). Then, in order to handle the prediction results uncertainty, the Bayesian framework is usually used, in which the prior estimates are updated by infield measurements without changing the model parameters. Nevertheless, in the case of system-level prognostic, the mere updating of the prior estimates, based on a predetermined model, is no longer sufficient. This is due to the mutual interactions between components that increase the system modeling uncertainties and may lead to an inaccurate prediction of the system RUL (SRUL). Therefore, this paper proposes a new methodology for online joint uncertainty quantification and model estimation based on particle filtering (PF) and gradient descent (GD). In detail, the inoperability input-output model (IIM) is used to characterize system degradations considering interactions between components and effects of the mission profile; and then the inoperability of system components is estimated in a probabilistic manner using PF. In the case of consecutive discrepancy between the prior and posterior estimates of the system health state, GD is used to correct and to adapt the IIM parameters. To illustrate the effectiveness of the proposed methodology and its suitability for an online implementation, the Tennessee Eastman Process is investigated as a case study.

According to the evolution tendency of the control decision process from a trackside to a train-borne system, various autonomous localization units for railway vehicles were developed. As recommended in railway standards, the design process of each system, here the autonomous localization units (LU), follows the V-model whose first step is to define its availability requirement in order to satisfy the global ETCS system requirements. The classical approach for assigning the subsystem availability is based on the assumption that failure parameters of other units are precisely known. This assumption is too restricted in reality due to the lack of information. In this paper, we propose a new approach that allows taking into account uncertainties in the dependability parameters of the ETCS components for identifying the upper threshold of the LU unavailability to reach ETCS availability requirements. Using fuzzy fault trees, the fuzzy unavailability of the ETCS without the autonomous LU is evaluated. Then, based on its membership function, we assess the satisfaction rate that an advanced ETCS with the autonomous LU can satisfy the ETCS availability target.

International audience; Health monitoring data are increasingly collected and widely used for reliability assessment and lifetime pre- diction. They not only provide information about degradation state but also could trace failure mechanisms of assets. The selection of a deterioration model that optimally fits in with health monitoring data is an important issue. It can enable a more precise asset health prognostic and help reducing operation and maintenance costs. Therefore, this paper aims to address the problem of degradation model selection including goals, procedure and evaluation criteria. Focusing on continuous degradation modeling including some currently used Lévy processes, the performance of classical and prognostic criteria are discussed through numerous numerical examples. We also investigate in what circumstances which methods perform better than others. The efficiency of a new hybrid criterion is highlighted that allows to take into account the information of goodness-of-fit of observation data when evaluating prognostic measure.

International audience; The quality of condition monitoring is an important factor affecting the effectiveness of a condition-based maintenance program. It depends closely on implemented inspection and instrument technologies, and eventually on investment costs, i.e., a more accurate condition monitoring information requires a more sophisticated inspection, hence a higher cost. While numerous works in the literature have considered problems related to condition monitoring quality, (e.g., imperfect inspection models, detection and localization techniques, etc.) few of them focus on adjusting condition monitoring quality for condition-based maintenance optimization. In this paper, we investigate how such an adjustment can help to reduce the total cost of a condition-based maintenance program. The condition monitoring quality is characterized by the observation noises on the system degradation level returned by an inspection. A dynamic condition-based maintenance and inspection policy adapted to such a observation information is proposed and formulated based on Partially Observable Markov Decision Processes. The use and advantages of the proposed joint inspection and maintenance model are numerically discussed and compared to several inspection-maintenance policies through numerical examples.

The non detection of bearing faults in rotating machines can lead to less availability, reliability and safety while increasing the maintenance costs due to unexpected breakdowns and urgent repairing. This paper deals with feature engineering to enhance the performance of an early fault detection and diagnostic of ball bearings of asynchronous electrical motors. The features of different types, ie. time, frequency and time-frequency, are extracted from both current and vibration. Then, they are selected based on a genetic algorithm to continuously capture the health state of the ball bearings. The proposed method is applied on sensor signals acquired from a test bench reproducing a real industrial system. The obtained results show the effectiveness of the method particularly for fault detection and diagnostic using current signals which can be useful in practical applications.

In recent years, the development of autonomous health management systems received increasing attention from worldwide companies to improve their performances and avoid downtime losses. This can be done, in the first step, by constructing powerful health indicators (HI) from intelligent sensors for system monitoring and for making maintenance decisions. In this context, this paper aims to develop a new methodology that allows automatically choosing the pertinent measurements among various sources and also handling raw data from high-frequency sensors to extract the useful low-level features. Then, it combines these features to create the most appropriate HI following the previously defined multiple evaluation criteria. Thanks to the flexibility of the genetic programming, the proposed methodology does not require any expertise knowledge about system degradation trends but allows easily integrating this information if available. Its performance is then verified on two real application case studies. In addition, an insightful overview on HI evaluation criteria is also discussed in this paper.

International audience; Gear reducer motors play an important role in industry due to their robustness and simplicity of construction. However, the appearance of faults in these systems can affect the quality of the product and lead to significant financial losses. Therefore, it is necessary to perform Prognostics and Health Management (PHM) for these systems. This paper aims to develop a practical and effective method allowing an early fault detection and diagnostic for critical components of the gear reducer, in particular gear and bearing defects. This method is based on a new indicator extracted from electrical signals. It allows characterizing different states of the gear reducer, such as healthy state, bearing faults, gear faults, and combined faults. The diagnostic of these states is done by the Adaptive Neuro-Fuzzy Inference System (ANFIS). The efficiency and the robustness of the proposed method are highlighted through numerous experimental tests with different levels of loads and speeds.

International audience; In Prognostic Health and Management (PHM) literature, the predictive maintenance studies can be classified into two groups. The first group focuses on the prognostics step but does not consider the maintenance decisions. The second group addresses the maintenance optimization question based on the assumptions that the prognostics information or the degradation models of the system are already known. However, none of the two groups provides a complete framework (from data-driven prognostics to maintenance decisions) investigating the impact of the imperfect prognostics on maintenance decision. Therefore, this paper aims to fill this gap of literature. It presents a novel dynamic predicive maintenance framework based on sensor measurements. In this framework, the prognostics step, based on the Long Short-Term Memory network, is oriented towards the requirements of operation planners. It provides the probabilities that the system can fail in different time horizons to decide the moment for preparing and performing maintenance activities. The proposed framework is validated on a real application case study. Its performance is highlighted when compared with two benchmark maintenance policies: classical periodic and ideal predicted maintenance. In addition, the impact of the imperfect prognostics information on maintenance decisions is discussed in this paper.

International audience; Smart manufacturing is a part of the fourth industry revolution (Industry 4.0), which offers promising perspectives for high reliability, availability, maintainability, and safety production process. Indeed, smart monitoring methods, that are implemented in this kind of manufacturing process, allow efficient tracking of a system degradation in real time through appropriate sensors. Then, the sensor data are analyzed and processed to extract effective health indicators for fault detection, diagnostic and prognostics. This paper aims to develop a practical methodology for constructing a new health indicator based on heterogeneous sensor measurements to efficiently monitor system states. The proposed methodology is applied to extract the health indicator of a robot cutting tool (i.e. end-flat mill). This indicator is then used to diagnose the different fault types of the tool by an adaptive neuro-fuzzy inference system model.

The pulp and paper industry is a fundamental sector of the economy of many countries. However, this sector requires real collaboration and initiatives from stakeholders to reduce its significant consumption of energy and emission of greenhouse gases. Heat exchangers are examples of equipment in pulp mills that are subjected to undesirable and complex phenomena such as evolution of fouling over time, which leads to inefficiency in terms of energy consumption and unplanned shutdowns, resulting in ineffective maintenance strategies and production costs. Therefore, there is a clear need to develop an accurate predictive maintenance tool that helps mill operators avoid such situations. It is necessary for that tool to effectively track the fouling evolution level and, based on it, deploy a reliable prognostics approach to estimate more accurately the time-to-clean of this equipment. This study presents a new hybrid prognostics approach for fouling prediction in heat exchangers. The proposed approach relies on the fusion of information of different prediction horizons to estimate the time-to-clean. Employing long short-term memory, it allows adaptation of long-term predictions by accurate short-term predictions using multiple non-linear auto-regressive exogenous models. This fusion not only captures the changes in degradation trend over time, but also ensures a good accuracy of prognostics results in both the short- and long-term horizons for planning maintenance actions. The effectiveness of the proposed approach was successfully proven on real industrial data collected from a pulp mill heat exchanger located in Canada.

Smart manufacturing is one of the key parts of the fourth industry revolution (Industry 4.0). It offers promising perspectives for high reliability, availability, maintainability and safety production process, but also makes the systems more complex and challenging for health assessment. To deal with these challenges, one needs to develop a robust approach to monitor and assess the system health state. In this paper, a practical and effective method that can be applied for fault detection and diagnostics of a given system is developed. The proposed method relies on a pattern recognition technique based on the construction of a new health indicator. This health indicator, which can be applied to different types of sensor measurements, is fed to an Adaptive Neuro-Fuzzy Inference System (ANFIS) to detect the health states of the system and diagnose the causes. Furthermore, the performance and the robustness of the proposed method are highlighted by considering various case studies under numerous operating conditions.

International audience; Health monitoring data are vital for failure prognostic and maintenance planning. Continuous monitoring data or frequent inspections can provide a large amount of information on degradation evolution and therefore ensure the quality of deterioration modeling and the lifetime prognostic accuracy. However, they are usually very costly, and sometimes inpractible in real engineering applications. Therefore, it is essential to address the issue of the appropriate amount of monitoring data. This paper proposes a new methodology to help the companies improving their actual inspection/monitoring policy to reduce operation and maintenance costs but also ensure the information quality. We investigate different types of inspection policies including periodic or non-periodic ones by considering multiples functions of the system degradation state that are linear, concave or convex. The best policies are chosen based on a multiobjective optimization problem dealing with the inspection cost and the information level. The advantages and disadvantages of the proposed methodology are discussed through numerous numerical examples for different types of degradation process, particularly Wiener and Gamma processes that have been largely addressed in the framework of degradation modeling.

The first but very significant step in electricity system planning is to make an accurate long-term forecast on electricity consumption. This article aims to forecast the consumption for the Vietnam electricity system (GWH) up to 2030. An econometric model with the Cobb Douglas production function is used. The five variables proposed in the forecasting function are GDP, income, population, proportion of industry and service in GDP, and number of households. The forecasting equation is tested in terms of stationary and co-integration to choose meaningful variables and to eliminate the minor ones which account for none or small impacts on the forecast. The results show that: (1) the qualified forecasting equation only includes 3 major variables: the per capita income, the population, and the number of households, (2) with the medium income scenario, the forecasting consumptions in 2020, 2025, 2030 are 230,195 GWH, 349,949 GWH, 511,268 GWH respectively. (3) The GDP and the proportion of industry and service in GDP do not make major impacts on this forecasting in Vietnam. The method and the result of this article are likely a typical example of forecasting electricity consumption in developing countries which have a transforming economy similar to that in Vietnam.

International audience; Wireless communication links tend to be employed more and more in safety-critical railway applications. Their safe use in an advanced train control system (TCS) is an issue that is addressed in this paper by characterizing the TCS service inter- ruption due to communication errors. More precisely, occurrence probabilities of single errors are first discussed. Then, we obtain probabilistic analytical expressions of several temporal conditions that lead to a TCS service interruption, here a train emergency braking (the critical event). The accuracy of this analytical ap- proach is proved when the results are compared with those given by a simulation approach with a Petri net model. Additionally, as the use case related to the “trains’ separation” is considered in this paper, an analytical evaluation process is proposed to discuss the tolerated time margins that can be fixed to limit the critical-event occurrence probability due to the wireless communication errors.; Wireless communication links tend to be employed more and more in safety-critical railway applications. Their safe use in an advanced train control system (TCS) is an issue that is addressed in this paper by characterizing the TCS service interruption due to communication errors. More precisely, occurrence probabilities of single errors are first discussed. Then, we obtain probabilistic analytical expressions of several temporal conditions that lead to a TCS service interruption, here a train emergency braking (the critical event). The accuracy of this analytical approach is proved when the results are compared with those given by a simulation approach with a Petri net model. Additionally, as the use case related to the 'trains' separation' is considered in this paper, an analytical evaluation process is proposed to discuss the tolerated time margins that can be fixed to limit the critical-event occurrence probability due to the wireless communication errors.