Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

BACE-RUL: A Bi-directional Adversarial Network with Covariate Encoding for Machine Remaining Useful Life Prediction

Zekai Zhang, Dan Li, Shunyu Wu, Junya Cai, Bo Zhang, See Kiong Ng, Zibin Zheng

TL;DR

A Bi-directional Adversarial network with Covariate Encoding for machine Remaining Useful Life (BACE-RUL) prediction, which only adopts sensor measurements from the current life cycle to predict RUL rather than relying on previous consecutive cycle recordings.

Abstract

Prognostic and Health Management (PHM) are crucial ways to avoid unnecessary maintenance for Cyber-Physical Systems (CPS) and improve system reliability. Predicting the Remaining Useful Life (RUL) is one of the most challenging tasks for PHM. Existing methods require prior knowledge about the system, contrived assumptions, or temporal mining to model the life cycles of machine equipment/devices, resulting in diminished accuracy and limited applicability in real-world scenarios. This paper proposes a Bi-directional Adversarial network with Covariate Encoding for machine Remaining Useful Life (BACE-RUL) prediction, which only adopts sensor measurements from the current life cycle to predict RUL rather than relying on previous consecutive cycle recordings. The current sensor measurements of mechanical devices are encoded to a conditional space to better understand the implicit inner mechanical status. The predictor is trained as a conditional generative network with the encoded sensor measurements as its conditions. Various experiments on several real-world datasets, including the turbofan aircraft engine dataset and the dataset collected from degradation experiments of Li-Ion battery cells, show that the proposed model is a general framework and outperforms state-of-the-art methods.

BACE-RUL: A Bi-directional Adversarial Network with Covariate Encoding for Machine Remaining Useful Life Prediction

TL;DR

A Bi-directional Adversarial network with Covariate Encoding for machine Remaining Useful Life (BACE-RUL) prediction, which only adopts sensor measurements from the current life cycle to predict RUL rather than relying on previous consecutive cycle recordings.

Abstract

Prognostic and Health Management (PHM) are crucial ways to avoid unnecessary maintenance for Cyber-Physical Systems (CPS) and improve system reliability. Predicting the Remaining Useful Life (RUL) is one of the most challenging tasks for PHM. Existing methods require prior knowledge about the system, contrived assumptions, or temporal mining to model the life cycles of machine equipment/devices, resulting in diminished accuracy and limited applicability in real-world scenarios. This paper proposes a Bi-directional Adversarial network with Covariate Encoding for machine Remaining Useful Life (BACE-RUL) prediction, which only adopts sensor measurements from the current life cycle to predict RUL rather than relying on previous consecutive cycle recordings. The current sensor measurements of mechanical devices are encoded to a conditional space to better understand the implicit inner mechanical status. The predictor is trained as a conditional generative network with the encoded sensor measurements as its conditions. Various experiments on several real-world datasets, including the turbofan aircraft engine dataset and the dataset collected from degradation experiments of Li-Ion battery cells, show that the proposed model is a general framework and outperforms state-of-the-art methods.

Paper Structure

This paper contains 27 sections, 17 equations, 4 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Preliminaries: System Degradation and RUL
  3. Deep Adversarial RUL Prediction
  4. System Architecture
  5. Methodology
  6. Condition Encoding.
  7. RUL Prediction.
  8. Cost Functions.
  9. Optimization.
  10. Experiments
  11. Datasets
  12. Experiment Set-up
  13. Data Pre-processing.
  14. Model Parameters.
  15. Baselines
  16. ...and 12 more sections

Figures (4)

  • Figure 1: The proposed BACE-RUL framework.
  • Figure 2: Predictions for engines and batteries during their lifespans by BACE-RUL. The horizontal axis $c$ represents the elapsed cycles from the beginning, while the vertical axis indicates the RUL at cycle $c$. "E" stands for "Engine".
  • Figure 3: T-SNE figures of four subsets of C-MAPSS. The T-SNE is a method to visualize the distribution of high-dimensional data. The more clumps in the figure indicate the more complicated the distribution is.
  • Figure 4: Predictions for different engines and batteries during their lifespans using baseline methods and the proposed BACE-RUL. The horizontal axis $c$ represents the elapsed cycles from the beginning, while the vertical axis indicates the RUL at cycle $c$. "E" stands for "Engine".