Power Converter Semiconductor Chip Life Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating the operational state and predicting the remaining useful life of power semiconductor devices in wind turbines, particularly under low frequency sampling data conditions, are inadequate, leading to uncertain decision-making and inefficient maintenance scheduling.
Innovation Solution
A method that receives operational data over consecutive time intervals, determines a parameter indicative of the operational state with an uncertainty indication, and generates alerts based on predefined criteria, using ensemble models or Bayesian inference to estimate peak junction temperature and life usage, allowing for improved prediction of remaining useful life.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If low frequency sampling data is used for estimating operational state and predicting remaining useful life, then data acquisition complexity is reduced, but measurement precision and reliability of the estimation deteriorate
Solution Approach 1:
The patent combines multiple degradation indicators (cumulative life usage, peak junction temperature, remaining useful life) into a unified assessment framework. By integrating these different parameters and using ensemble models that combine multiple prediction approaches, the system achieves reliable estimates even with low-frequency sampling data, resolving the contradiction between reduced data acquisition complexity and maintained estimation precision.
2Ease of operation
If existing estimation methods are used under low frequency sampling conditions, then implementation simplicity is maintained, but prediction reliability and decision-making confidence deteriorate
Solution Approach 1:
The patent implements a feedback mechanism where the system continuously monitors degradation indicators and updates predictions based on observed trends. The ensemble models incorporate feedback from multiple prediction approaches and adjust their outputs accordingly, providing confidence intervals that reflect prediction reliability. This feedback loop maintains implementation simplicity while significantly improving prediction reliability through adaptive learning from actual device behavior.
Solution Approach 2:
The patent changes the parameter representation by introducing confidence intervals and probability distributions alongside point estimates. Instead of providing single deterministic values, the system outputs ranges with associated confidence levels, transforming the reliability metric from a binary reliable/unreliable state to a continuous measure that maintains implementation simplicity while conveying nuanced reliability information for better decision-making.
3Ease of operation
If traditional maintenance scheduling is used without accurate life prediction, then maintenance planning simplicity is maintained, but operational efficiency and resource utilization deteriorate
Solution Approach 1:
The patent enables preliminary action by providing advance predictions of remaining useful life with confidence intervals. This allows maintenance to be planned proactively before actual failures occur, optimizing the timing of maintenance activities. The system performs preliminary assessments that guide maintenance scheduling decisions, improving operational efficiency by preventing unnecessary early replacements while avoiding unexpected failures, all while maintaining relative simplicity in maintenance planning procedures.
Data Source
AI summary
The present disclosure provides examples of methods and systems for estimating an operational state of a semiconductor chip of a power semiconductor device which are based on determining a parameter indicative of an operational state of the semiconductor chip including an indication of uncertainty of the operational state based on the received operational data. The present disclosure further provides examples of methods and systems for predicting remaining useful life of a semiconductor chip.