Hybrid Turbomachinery Risk Model for Predictive Maintenance Scheduling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing maintenance optimization methods for turbomachinery assets lack sufficient predictive accuracy, leading to increased maintenance costs and reduced asset reliability and availability.
Innovation Solution
A computer-based predictive maintenance service that combines empirical, physics modeling, and data-driven approaches to identify anomalies, classify their severity, and estimate the time to maintenance, thereby optimizing maintenance scheduling and scope.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single approach (empirical, physics-based, or data-driven) is used for maintenance estimation, then the method is simple to implement, but the predictive accuracy is insufficient
Solution Approach 1:
The patent merges three distinct maintenance estimation approaches (empirical, physics-based, and data-driven) into a unified hybrid model. Each approach processes different aspects of maintenance prediction, and their results are integrated to produce a comprehensive maintenance estimation that leverages the strengths of all three methods while compensating for their individual weaknesses.
Solution Approach 2:
The hybrid maintenance model functions as a composite system where multiple methodological 'materials' (empirical formulas, physics-based simulations, data-driven analytics) are combined to create a more robust and accurate prediction framework, analogous to how composite materials combine different substances to achieve superior properties.
2Reliability
If fixed time scheduling is used for maintenance planning, then the scheduling is simple to manage, but the maintenance costs increase and asset reliability decreases
Solution Approach 1:
The patent transitions from static fixed-time scheduling to dynamic condition-based scheduling. The maintenance timing is continuously adjusted based on real-time asset condition monitoring, predictive analytics, and evolving operational data, allowing the maintenance plan to adapt dynamically to actual asset needs rather than following a rigid predetermined schedule.
Solution Approach 2:
The system implements continuous feedback loops where maintenance performance, asset condition data, and operational outcomes are monitored and fed back into the predictive models. This feedback mechanism enables the system to learn from past maintenance events and improve future maintenance timing predictions, creating a self-optimizing maintenance planning process.
Data Source
AI summary
A computer implemented method for the maintenance optimization of a fleet or group of turbomachinery assets is disclosed. The method comprises the step of model training and setup, aiming at setting configurations parameters, that can be executed offline, and the step of online calculation on new input data, which is based on detected data and extracted statistical features. An anomaly identification and classification follow, thus calculating a risk assessment, for estimating the risk that an anomaly might cause any event that requires a maintenance task to be executed on one or more assets of the fleet.