Bayesian Aging Model for Device Lifetime Prediction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing device monitoring methods are not sufficiently accurate in predicting failures, leading to potential device downtime and costly repairs, as they rely on physical sensors and statistical predictions that may not account for complex aging dynamics.

Innovation Solution

A method involving an aging model with an End of Life (EOL) boundary condition and equations for aging variables, which calculates a probability density function over time, updates based on measured data, and generates signals for health predictions and Remaining Useful Life (RUL) to prevent failures more reliably.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If physical sensors and statistical predictions are used for device monitoring, then device failures can be detected and prevented, but the prediction accuracy is insufficient leading to potential device downtime and costly repairs

Engineering Contradiction:
Improvefailure prediction accuracyVSAvoiddevice downtime
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent transforms the monitoring approach by changing from direct physical measurement to probabilistic parameter estimation. It uses Bayesian inference to update the probability density function of aging variables based on observed data, enabling more accurate prediction of remaining useful life and reducing false alarms that cause unnecessary downtime

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent introduces an aging model as an intermediary between observed device parameters and failure prediction. The model includes equations that describe the evolution of aging variables over time, allowing indirect but more accurate inference of device health status without requiring direct measurement of degradation

Inventive Principle:
Principle #24Intermediary (Mediator)

2Reliability

If regular maintenance is performed after predetermined time or cycles, then device failures are prevented, but wasteful prophylactic maintenance occurs

Engineering Contradiction:
Improvefailure preventionVSAvoidwasteful maintenance
Core Design Contradiction:
ReliabilityVSLoss of energy

Solution Approach 1:

The patent replaces static scheduled maintenance with dynamic condition-based maintenance. The probability density function of aging variables is continuously updated based on actual device observations, allowing maintenance timing to adapt to the actual degradation rate of each device, thus preventing both failures and wasteful early maintenance

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent performs preliminary prediction of remaining useful life using the aging model before maintenance is actually needed. By calculating the probability distribution of failure time in advance, the system can schedule maintenance optimally - neither too early (wasteful) nor too late (failure risk)

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4307066A1Method for lifetime prediction and monitoring
Publication Date: 2024.01.17 HITACHI ENERGY LTD
  • EP4307066A1 patent drawingFigure 1
  • EP4307066A1 patent drawingFigure 2
  • EP4307066A1 patent drawingFigure 3

AI summary

A method for lifetime prediction and monitoring of a device, and a corresponding system is provided. The method comprises calculating a probability density function over time for an aging variable based on solving at least one equation from an aging model with an End of Life, EOL, boundary condition, measuring an condition related observable of the device; obtaining first data representing measurement of the observable, calculating a likelihood for the aging variable from the first data, updating the calculated probability density function of the aging variable based on the likelihood, and generating a signal indicating a health prediction of the device based on the probability density function, the aging model and the EOL boundary condition.