High-Voltage Equipment Prognosis With Dynamic Failure Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High voltage equipment (HVE) prognosis is challenging due to varying operating conditions, specifications, and failure modes, requiring efficient models for accurate fault prediction and maintenance scheduling across different industries and geographic locations.
Innovation Solution
A monitoring system dynamically selects and tunes models based on performance criteria, using historical data from similar HVEs to predict failure modes and provide proactive maintenance responses, incorporating machine learning, stochastic, and empirical models for improved accuracy and resource efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple models are used to handle varying operating conditions and failure modes, then prognosis accuracy is improved, but device complexity increases
Solution Approach 1:
The system dynamically selects appropriate prognosis models based on real-time operating conditions and equipment state. Instead of using a fixed model, the system adapts model selection to match current operational context, thereby maintaining high accuracy across varying conditions without permanently increasing system complexity.
Solution Approach 2:
The system changes model parameters and selection based on operating conditions such as load, temperature, and equipment age. By adjusting which model is applied based on parameter thresholds and equipment state, the system achieves accurate prognosis across diverse operating scenarios while managing complexity through conditional model selection.
2Measurement precision
If historical data from multiple HVEs is aggregated for model tuning, then model accuracy is improved, but data processing complexity increases
Solution Approach 1:
The system segments historical data from multiple HVEs into relevant subsets based on operating conditions, equipment type, and failure modes. By processing and tuning models on segmented data rather than raw aggregated data, the system achieves accurate models while reducing the computational complexity of data processing.
Solution Approach 2:
The system introduces intermediate data processing layers that aggregate and preprocess historical data before model tuning. These intermediary processing steps organize multi-source data into structured formats, improving model accuracy while managing the complexity of handling data from multiple HVEs.
3Reliability
If dynamic model selection is implemented, then prognosis reliability is improved, but computational resources consumed increase
Solution Approach 1:
The system applies partial model selection by choosing only the most appropriate model(s) for current operating conditions rather than evaluating all available models. This partial action approach maintains high prognosis reliability by selecting sufficient models without the excessive computational cost of comprehensive model evaluation.
Solution Approach 2:
The system performs preliminary model selection based on operating condition thresholds and equipment state before detailed prognosis analysis. By pre-selecting appropriate models based on coarse-grained condition assessment, the system reduces computational resource consumption during actual prognosis generation while maintaining reliability.
Data Source
AI summary
Various aspects of prognosis of an installed high voltage equipment (HVE) by a monitoring system are described. One or more models are dynamically selected from a plurality of models tuned from data obtained from a plurality of HVEs communicatively connected with the monitoring system. A failure mode of the installed HVE is predicted, based on input parameters associated with the installed HVE, using the one or more models. At least one prognostic response is determined for the installed HVE, based on the predicted failure mode, using the one or more models. The at least one prognostic response is provided for the installed HVE.