Figure 1: Operating and Maintenance Cost Chart [3]

A new Health Index for RUL estimation of MEMS sensors using dimensionality reduction and Artificial Neural Networks

Article

Full-text available

Nov 2024
IEEE T INSTRUM MEAS

The accurate estimation of the Remaining Useful Life (RUL) of Micro-Electro-Mechanical Systems (MEMS) sensors is crucial for enabling effective condition-based maintenance (CBM) and improving system reliability. This is of vital importance in planning machinery replacements before their actual fault, reducing downtime and thus greatly optimizing costs. However, unlike industrial components (such as bearings, rotating machinery, or power devices), current prognostic strategies for MEMS sensors are limited due to the lack of suitable health indexes capable of describing the degradation path of the sensor’s metrological performances. To address this gap, this paper proposes a novel Health Index for MEMS accelerometers based on static accelerometer outputs by combining time-domain and frequency-domain features using a dimensionality reduction technique. By applying dimensionality reduction, the method captures the essential characteristics of sensor degradation, which are then used to train a Deep Learning-based model for predicting the Remaining Useful Life sufficiently in advance to allow for appropriate maintenance. This paper aims to fill an important gap by providing an innovative solution for MEMS sensor prognostics, with broad implications for industrial sectors reliant on precision sensing technologies. The performance of the proposed health indicator is validated through experimental data, demonstrating its ability to enhance the accuracy of MEMS sensor prognostics. Using data from the proposed accelerated life test and the Arrhenius degradation model, the RUL prediction based on the proposed health indicator has been performed at various operating temperatures, including typical conditions of industrial environments with high thermal stress.

Machine Learning and Internet-of-Things Solutions for Microgrid Resilient Operation

Chapter

Nov 2024

Microgrids inherently have lower inertia in comparison with the power system due to the high penetration of renewable generation sources. On the other hand, the lower spring reserve, high dynamic loads, and inherent volatility of renewable generation sources have demanded real-time control of them. The development of intelligent electrical devices, communication infrastructure, and Internet-of-Things (IoT) technology has enhanced real-time microgrid operation. In contrast, they have exposed the smart microgrid to cyber malicious activities. Cyber attackers try to disrupt the microgrid operation by manipulating the sensors, control center, communication infrastructure, and intelligent electrical devices, leading to decreasing power quality, cascading instability, and, finally, blackout. The effect of a cyber-attack in a standalone microgrid will be more distractive. Given the advantages inherent in these technologies, this chapter conducts an extensive examination of their integration into the management, control, and upkeep of both onshore and offshore microgrids. It provides a detailed account of the communication infrastructure within microgrids and identifies potential vulnerabilities. The chapter then elucidates methods for detecting cyber-attacks, placing a particular emphasis on machine learning, IoT, and Digital Twins. Subsequently, the chapter delves into the convergence of these technologies with maritime microgrids. It expounds upon how these advancements render maritime microgrids susceptible to cyber threats and emphasizes the instrumental role of machine learning in fortifying their cyber resilience. Furthermore, the chapter explores the application of IoT and Digital Twins, combined with machine learning, to enhance the maintenance and operational resilience of maritime microgrids. Finally, the chapter addresses the technical challenges faced by smart seaports and offers resilient operational solutions to mitigate potential disruptions.

Health index estimation through integration of general knowledge with unsupervised learning

Article

Full-text available

Jul 2024
RELIAB ENG SYST SAFE

Design and Development of a Prototype Operating Parameters Monitoring System for High-Voltage Switchgear Using IoT Technologies

Article

Feb 2024
IEEE T POWER DELIVER

Currently, almost all high-voltage switchgear and gas-insulated systems (GIS) are filled with SF ₆ under pressure. Accidental leakage of SF ₆ has dramatic effects on both the environment and the equipment itself. A large leak burdens the environment with hundreds of tons of CO ₂ equivalent and is a major safety risk for network equipment. Therefore, close monitoring of SF ₆ in high-voltage switchgear is virtually required in order to ensure the stability of the power grid. Very few substations currently have the infrastructure to remotely monitor their assets, and retrofit solutions for older substations are either uneconomical or completely impractical. This work focuses on the development of a new open monitoring system that will be inexpensive, minimally intrusive, and infinitely scalable, in terms of both hardware and software. The developed monitoring system should allow transmission system operators to remotely monitor key parameters of their most important assets, which will also provide them with reaction time to avoid critical breakdowns.

A novel machine learning model to design historical-independent health indicators for composite structures

Article

Feb 2024
COMPOS PART B-ENG

Developing comprehensive health indicators (HIs) for composite structures encompassing various damage types is challenging due to the stochastic nature of damage accumulation and uncertain events (like impact) during operation. This complexity is amplified when striving for HIs independent of historical data. This paper introduces an AI-driven approach, the Hilbert transform-convolutional neural network under a semi-supervised learning paradigm, to designing reliable HIs (fulfilling requirements, referred to as 'fitness'). It exclusively utilizes current guided wave data, eliminating the need for historical information. Ensemble learning techniques were also used to enhance HI quality while reducing deep learning randomness. The fitness equation is refined for dependable comparisons and practicality. The methodology is validated through investigations on T-single stiffener CFRP panels under compression-fatigue and dogbone CFRP specimens under tension-fatigue loadings, showing high performance of up to 93% and 81%, respectively, in prognostic criteria.

Auxiliary Particle Filter for Prognostics and Health Management

Article

Dec 2023

Accurately predicting the remaining useful life (RUL) of a system is a crucial factor in prognostics and health management (PHM). This paper introduces an auxiliary particle filter (APF) model, which has the advantages of dynamically updating the model parameters and being optimized in computational speed for prognosis applications in real engineering problems. The development of particle filter (PF) in the recent decade focused on increasing the PF model’s complexity to solve more difficult problems. However, the added complexity negatively impacts the computational speed. The number of particles is commonly reduced to compensate for this increased computational burden, but this significantly reduces the accuracy of PF’s posterior distribution. The developed APF model can estimate unknown states and model parameters at the same time with a large number of particles. This algorithm was demonstrated with a dataset from an electric motor accelerated aging experiment. The results show that this model can quickly and accurately predict the RUL and is robust to measurement noise.

Unsupervised Physics-Informed Health Indicator Estimation for Complex Systems

Article

Full-text available

Oct 2023

Developing Health Indicators (HI) is a crucial aspect of prognostics and health management of complex systems. Previous research has demonstrated the benefits of accurately determining the HI, which can lead to better performance of prognostic models. However, the existing methodologies for determining HI in complex systems are mostly semi-supervised and rely on assumptions that may not hold in real-world scenarios. The existing methods usually involve using a reference set of healthy sensor readings or run-to-failure data to infer HI. But the unsupervised inference of HI from sensor readings, which is challenging in scenarios where diverse operating conditions can mask the effect of degradation on sensor readings, has not been extensively researched. In this paper, we propose a novel physics-informed unsupervised model for determining HI. Unlike previous methods, constrained by assumptions, the proposed method uses prior general knowledge about degradation to infer HI, thereby eliminating the need for a reference set of healthy sensor readings. The proposed unsupervised model is an Autoencoder that incorporates constraints on its latent space to ensure consistency with knowledge about degradation. We assess the efficacy of the proposed model by analyzing a prevalent prognostic case study, specifically the turbofan engine dataset (N-CMAPSS). Our analysis considers the model's sensitivity to data availability and the resulting Health Index's quality, including trendability and monotonicity. Additionally, we investigate the impact of incorporating the Health Index in predicting Remaining Useful Life (RUL). We demonstrate that our proposed method generates a Health Index that exhibits greater monotonicity and trendability than the current state-of-the-art semi-supervised approach. Moreover, our approach for identifying the Health Index leads to enhanced prognostic performance compared to the existing semi-supervised approach.

A comparison study for bearing remaining useful life prediction by using standard stochastic approach and digital twin

Article

Oct 2023
Int J Reliab Saf

Remaining useful life prediction is important for research of maintenance. It is common to use stochastic approaches to predict remaining useful life of components. On the other hand, there is a digital twin model developed by MATLAB for bearing’s real-time remaining useful life prediction. To have a better understanding of the advantages and disadvantages of these models, an experiment was designed and implemented to get real degradation data of bearings for model testing. Two stochastic approaches are selected which are Wiener process and Geometric Brownian Motion. The purpose of the paper is to compare the models for remaining useful life prediction with standard stochastic approaches and digital twin through real degradation data in order to find the comparison among them. Finally, the MATLAB digital twin model outperforms stochastic approaches in the early phases of prediction while remaining comparable in the latter stages. The paper could be used as a reference for further remaining useful life prediction research.

A Health-Aware Framework for Online Replanning of Unmanned Aerial Vehicles in Stochastic Environments

Thesis

Full-text available

Sep 2023

Tim Darrah

This thesis centers around the development of an end-to-end framework for health-aware replanning of UAVs with varying levels of use-based degradation during operations in environments where conditions may change from time to time.

ADVANCED HEALTH MONITORING OF COMPOSITE STRUCTURES THROUGH DEEP LEARNING-BASED ANALYSIS OF LAMB WAVE DATA FOR DEVELOPING HEALTH INDICATORS

Conference Paper

Sep 2023

A health indicator (HI) serves as an intermediary link between structural health monitoring (SHM) data and prognostic models, and an efficient HI should meet prognostic criteria, i.e., monotonicity, trendability, and prognosability. However, designing a proper HI for composite structures is a challenging task due to the complex damage accumulation process during operational conditions. Additionally, designing a HI that is independent of historical SHM data (i.e., from the healthy state until the current state) is even more challenging as HI and remaining useful life prediction are time-dependent phenomena. A reliable SHM technique is required to extract informative time-independent data, and a powerful model is necessary to construct a proper HI from that data. The lamb wave (LW) technique is a useful SHM method that can extract such time-independent data. However, translating the LW data at each time step to the appropriate HI value is a challenge. AI—deep learning in this case—offers significant mathematical potential to meet this difficulty. A semi-supervised learning approach is developed to train the model using the simulated ideal HIs as the targets. The model uses the current LW data, without prior or subsequent data, to output the current HI value. Prognostic criteria are then calculated using the entire HI trajectory until the end-of-life. To validate the proposed approach, aging experiments from NASA’s prognostics data repository are used, which include composite specimens subjected to a tension-tension fatigue loading and monitored using the LW technique. The LW data is first processed using the Hilbert transform, and then, upper and lower signal envelopes in two states – baseline and current time – are used to feed the deep learning model. The results confirm the effectiveness of the proposed approach according to the prognostic criteria. The effect of different triggering frequencies of the LW system on the results is also discussed in terms of the prognostic criteria.

Operating and Maintenance Cost Chart [3]

Similar publications

Citations