Abstract:In the past decades, fault diagnosis and prognosis (FDP) approaches were developed in the Riemann sampling (RS) framework, in which samples are taken and algorithms are executed in periodic time intervals. With the increase of system complexity, a bottleneck of real-time implementation of RS-based FDP is limited calculation resources, especially for distributed applications. To overcome this problem, a Lebesgue sampling-based FDP (LS-FDP) is proposed. LS-FDP takes samples on the fault dimension axis and provides a need-based FDP philosophy in which the algorithm is executed only when necessary. In previous LS-FDP, the Lebesgue length is a constant. To accommodate the nonlinear fault dynamics, it is desirable to execute FDP algorithm more frequently when the fault growth is fast while less frequently when fault growth is slow. This requires to change the Lebesgue length adaptively and optimize the selection of Lebesgue length based on fault state and fault growth speed. The goal of this paper is to develop an improved LS-FDP method with adaptive Lebesgue length, which enables the FDP to be executed according to fault dynamics and has low cost in terms of computation and hardware resource needed. The design and implementation of adaptive LS-FDP (ALS-FDP) based on a particle filtering algorithm are illustrated with a case study of Li-ion batteries to verify the performances of the proposed approach. The experimental results show that ALS-FDP keeps close monitoring of fault growth and is accurate and time-efficient on long-term prognosis.Note to Practitioners-Traditional fault diagnosis and prognosis (FDP) approaches are based on the Riemann sampling (RS) method, in which samples are taken and algorithms are executed in periodic time intervals no matter if it is necessary. To reduce computation and make optimal use of computational resources, Lebesgue sampling method is introduced into FDP. In this Lebesgue sampling-based approach, Lebesgue states are defined on the fault dimension axis and algorithm is executed only when the measurement causes a transition from one Lebesgue state to another, or an event happens. Different from RS-based approach, this is a need-based FDP philosophy in which the algorithm is executed only when necessary. This paper studies a parameter adaption method to optimally adjust the Lebesgue state length, which results in the changes of number and location of Lebesgue states, according to fault state and fault growth speed to accommodate the nonlinearity of fault dynamics and achieve a balance between computation and performance. The application to the capacity degradation of a set of Li-ion batteries is presented to demonstrate the effectiveness of the proposed adaptive Lebesgue state length FDP method.

Lebesgue Sampling Based Deep Belief Network for Lithium-Ion Battery Diagnosis and Prognosis

A Physics-Informed Hybrid Data-Driven Approach with Generative Electrode-Level Features for Lithium-Ion Battery Health Prognostics

Lebesgue-Sampling-Based Diagnosis and Prognosis for Lithium-Ion Batteries

Parameters Adaption of Lebesgue Sampling-based Diagnosis and Prognosis for Li-ion Batteries

Uncertainty Management in Lebesgue-Sampling-Based Diagnosis and Prognosis for Lithium-Ion Battery

Low-Cost Adaptive Lebesgue Sampling Particle Filtering Approach for Real-Time Li-Ion Battery Diagnosis and Prognosis.

Health Diagnosis for Lithium-Ion Battery by Combining Partial Incremental Capacity and Deep Belief Network During Insufficient Discharge Profile

Dynamic Bayesian Network-Based Lithium-Ion Battery Health Prognosis for Electric Vehicles

A novel approach for prognosis of lithium-ion battery based on geometrical features and data-driven model

Physics-informed Deep Learning for Lithium-Ion Battery Diagnostics Using Electrochemical Impedance Spectroscopy

A Deep Learning Approach for State-of-Health Estimation of Lithium-Ion Batteries Based on a Multi-Feature and Attention Mechanism Collaboration

A Novel Hybrid Prognostic Approach for Remaining Useful Life Estimation of Lithium-Ion Batteries

Novel informed deep learning-based prognostics framework for on-board health monitoring of lithium-ion batteries

An online fault diagnosis method for lithium-ion batteries based on signal decomposition and dimensionless indicators selection

Lebesgue Sampling-Based Li-Ion Battery Simplified First Principle Model for SOC Estimation and RDT Prediction

Degradation Mode Knowledge Transfer Method for LFP Batteries

Specialized deep neural networks for battery health prognostics: Opportunities and challenges

A Lebesgue-Time-Space-Model and Particle Filter Based Diagnosis and Prognosis Method

Electric Vehicle Lithium-Ion Battery Fault Diagnosis Based on Multi-Method Fusion of Big Data

Realistic fault detection of li-ion battery via dynamical deep learning