Electromechanical Component RUL Estimation Using Physics-Based Fault Models

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods fail to accurately predict the remaining service life of electromechanical components due to limitations in fault detection and monitoring, leading to unplanned downtimes and maintenance.

Innovation Solution

A method that compares sensor measurement data with fault patterns using a numerical simulation on a physical model, combined with an artificial neural network for high reliability, to determine the remaining service life by simulating stress loads and material properties.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Difficulty of detecting and measuring

If fault detection is performed using sensor data alone, then fault monitoring capability is improved, but prediction accuracy of remaining service life deteriorates

Engineering Contradiction:
Improvefault monitoring capabilityVSAvoidprediction accuracy of remaining service life
Core Design Contradiction:
Difficulty of detecting and measuringVSMeasurement precision

Solution Approach 1:

The patent introduces a physical model as an intermediary between sensor data and RUL prediction. The physical model (comprising mathematical equations representing component behavior) transforms raw sensor measurements into meaningful degradation states, enabling accurate RUL prediction. This mediator bridges the gap between simple fault detection and precise remaining life estimation.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces purely data-driven machine learning approaches with a physics-based modeling approach. Instead of relying on black-box algorithms to predict RUL from sensor data, the system uses explicit physical models (mathematical equations) that represent the actual mechanical and physical degradation processes, providing interpretable and accurate predictions.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If numerical simulation on physical model is used, then prediction reliability is improved, but computational complexity increases

Engineering Contradiction:
Improveprediction reliabilityVSAvoidcomputational complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent changes the parameters of the physical model dynamically based on sensor data. Instead of solving complex simulations continuously, the system updates key parameters (degradation state, stress loads, material properties) based on real-time sensor measurements, maintaining prediction reliability while reducing computational burden through parameter-based adaptation rather than full re-simulation.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent performs preliminary setup of the physical model offline, pre-defining the mathematical structure, material properties, and boundary conditions. This preliminary action allows the online prediction system to only update specific parameters based on sensor data rather than rebuilding the entire simulation model, significantly reducing real-time computational complexity while maintaining reliability.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3736649B1Method for estimating the remaining useful life of at least one electromechanical component
Publication Date: 2024.08.21 SIEMENS AG
  • EP3736649B1 patent drawingFigure 1~2

AI summary

The invention relates to a method for estimating the remaining service life of at least one electromechanical component. To achieve high reliability in estimating the remaining service life, it is proposed that at least one system parameter is determined from sensor data of at least one sensor device (6) to which at least one electromechanical component (4) is assigned, wherein at least one failure case of the electromechanical component (4) is assigned to the system parameter, wherein a stress load on the electromechanical component (4) is simulated on a physical model using a numerical method based on the system parameter, and wherein a remaining service life is determined from the simulated stress load.