EPE Association: EPE 2025 - DS3c: State of Health: Online Monitoring, Failure Diagnosis and Prognosis, Remaining Useful Life Prediction

EPE 2025 - DS3c: State of Health: Online Monitoring, Failure Diagnosis and Prognosis, Remaining Useful Life Prediction
You are here: EPE Documents > 01 - EPE & EPE ECCE Conference Proceedings > EPE 2025 - Conference > EPE 2025 - Topic 05: Sustainable and Affordable Power Electronics > EPE 2025 - DS3c: State of Health: Online Monitoring, Failure Diagnosis and Prognosis, Remaining Useful Life Prediction

   [return to parent folder]

   DC Series Arc Fault Detection with LCL-Type Boost Converter
By Hwa-Pyeong PARK, Mina KIM [View]
[Download]

Abstract: The damaged and loose connector can induce the DC series arc fault condition. Previously, the frequency domain analysis using PV current, such as Fourier transform and wavelet transform, was widely employed for the series arc fault detection. However, performance of arc fault detection is not consistent according to system configurations. This paper investigates the DC series arc fault capability using the frequency spectrum analysis according to the impedance of power converter. From the characteristics of arc fault condition, the boost converter using LCL-type filter can clarify the arc fault condition in the frequency domain analysis. The experimental results using 800-W prototype converter can verify the frequency spectrum change at the arc fault condition with the proposed converter with LCL-type filter.

   Live Monitoring of the Blocking Behaviour of an IGBT Module in the HV-H³TRB Test
By Benedikt KOSTKA, Axel MERTENS [View]
[Download]

Abstract: One of the main electrical characteristics of IGBT modules influenced by humidity-related degradation is the blocking behaviour.This paper presents an approach for online monitoring of the blocking behaviour during a accelerated ageing test based on the high voltage, high humidity, high temperature reverse bias (HV-H$^3$TRB) test by injecting a constant current into the device under test.

   Practical implementation of a New Temperature-Insensitive Aging Indicator of bond-wire contacts for on-line monitoring of IGBT power modules
By Zoubiir KHATIR, Ali IBRAHIM, Richard LALLEMAND [View]
[Download]

Abstract: The online assessment of the health status of power electronic modules is crucial for predictive maintenance, enabling better system reliability and optimized operational efficiency. Traditional aging indicators are often impractical for online monitoring due to their complexity and high implementation costs, particularly because they require junction temperature estimation. In this paper, we focus on the practical implementation of a novel, temperature-insensitive aging indicator tailored for on-line monitoring bond-wire degradation in IGBT power modules. This method, based on the analysis of zero-temperature coefficient (ZTC) of I-V characteristics with degraded top-metal interconnects, allows for an efficient and cost-effective monitoring solution. It is presented an experimental validation through power cycling tests, highlighting the feasibility of real-time application. The proposed indicator shows a strong correlation with conventional aging indicators while offering enhanced sensitivity and independence from temperature variations. Furthermore, we outline a robust methodology for integrating this approach into industrial or embedded applications, ensuring scalability and ease of deployment.

   State-of-Health (SOH) and State-of-Charge (SOC) Estimation for Lithium-ion Battery under Fast-Charging
By Xuelu WANG, Toufik AZIB, Jianwen MENG, Khelil SIDI BRAHIM [View]
[Download]

Abstract: This paper presents two models for lithium-ion battery SOC and SOH estimation using a Toyota Research Institute dataset. A Random Forest Regressor predicts SOH after 10 cycles, while an LSTM neural network estimates real-time SOC. Both models show high accuracy, contributing to advanced health-aware energy management in electric vehicles.

EPE 2025 - DS3c: State of Health: Online Monitoring, Failure Diagnosis and Prognosis, Remaining Useful Life Prediction
You are here: EPE Documents > 01 - EPE & EPE ECCE Conference Proceedings > EPE 2025 - Conference > EPE 2025 - Topic 05: Sustainable and Affordable Power Electronics > EPE 2025 - DS3c: State of Health: Online Monitoring, Failure Diagnosis and Prognosis, Remaining Useful Life Prediction
	[return to parent folder]

	DC Series Arc Fault Detection with LCL-Type Boost Converter By Hwa-Pyeong PARK, Mina KIM	[View] [Download]
Abstract: The damaged and loose connector can induce the DC series arc fault condition. Previously, the frequency domain analysis using PV current, such as Fourier transform and wavelet transform, was widely employed for the series arc fault detection. However, performance of arc fault detection is not consistent according to system configurations. This paper investigates the DC series arc fault capability using the frequency spectrum analysis according to the impedance of power converter. From the characteristics of arc fault condition, the boost converter using LCL-type filter can clarify the arc fault condition in the frequency domain analysis. The experimental results using 800-W prototype converter can verify the frequency spectrum change at the arc fault condition with the proposed converter with LCL-type filter.

	Live Monitoring of the Blocking Behaviour of an IGBT Module in the HV-H³TRB Test By Benedikt KOSTKA, Axel MERTENS	[View] [Download]
Abstract: One of the main electrical characteristics of IGBT modules influenced by humidity-related degradation is the blocking behaviour.This paper presents an approach for online monitoring of the blocking behaviour during a accelerated ageing test based on the high voltage, high humidity, high temperature reverse bias (HV-H$^3$TRB) test by injecting a constant current into the device under test.

	Practical implementation of a New Temperature-Insensitive Aging Indicator of bond-wire contacts for on-line monitoring of IGBT power modules By Zoubiir KHATIR, Ali IBRAHIM, Richard LALLEMAND	[View] [Download]
Abstract: The online assessment of the health status of power electronic modules is crucial for predictive maintenance, enabling better system reliability and optimized operational efficiency. Traditional aging indicators are often impractical for online monitoring due to their complexity and high implementation costs, particularly because they require junction temperature estimation. In this paper, we focus on the practical implementation of a novel, temperature-insensitive aging indicator tailored for on-line monitoring bond-wire degradation in IGBT power modules. This method, based on the analysis of zero-temperature coefficient (ZTC) of I-V characteristics with degraded top-metal interconnects, allows for an efficient and cost-effective monitoring solution. It is presented an experimental validation through power cycling tests, highlighting the feasibility of real-time application. The proposed indicator shows a strong correlation with conventional aging indicators while offering enhanced sensitivity and independence from temperature variations. Furthermore, we outline a robust methodology for integrating this approach into industrial or embedded applications, ensuring scalability and ease of deployment.

	State-of-Health (SOH) and State-of-Charge (SOC) Estimation for Lithium-ion Battery under Fast-Charging By Xuelu WANG, Toufik AZIB, Jianwen MENG, Khelil SIDI BRAHIM	[View] [Download]
Abstract: This paper presents two models for lithium-ion battery SOC and SOH estimation using a Toyota Research Institute dataset. A Random Forest Regressor predicts SOH after 10 cycles, while an LSTM neural network estimates real-time SOC. Both models show high accuracy, contributing to advanced health-aware energy management in electric vehicles.