Hybrid Model for Electro-Mechanical System Life Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for monitoring and controlling electro-mechanical systems are limited in accuracy due to physics-based models that are not validated, unable to predict maintenance and downtime effectively, and are not scalable for multiple systems, making it difficult to determine reliability and life prediction.
Innovation Solution
A hybrid model is generated by combining real-time sensor data with simulated responses to diagnose failures and predict life trends, using a combination of data-driven and physics-based approaches, allowing for accurate diagnosis and life prediction of electro-mechanical systems individually and in fleets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If physics-based models are used for monitoring electro-mechanical systems, then the monitoring can be performed without extensive sensor data, but the accuracy of diagnosis and life prediction is not validated
Solution Approach 1:
The patent combines physics-based models with data-driven models to create a hybrid model. The physics-based model provides the theoretical framework while the data-driven component validates and refines it using actual sensor data, thereby improving both diagnosis accuracy and model reliability simultaneously
Solution Approach 2:
The system uses sensor data to continuously validate and update the physics-based models. The comparison between predicted values from the physics model and actual sensor measurements provides feedback that refines the model parameters, ensuring both accuracy and reliability
2Adaptability or versatility
If physics-based models are used for monitoring, then the approach is theoretically sound, but the models are not scalable to multiple electro-mechanical systems
Solution Approach 1:
The hybrid model framework is designed to be universal and applicable to multiple electro-mechanical systems. Once the physics-based model is developed for a system type, it can be scaled to monitor fleets of similar systems by adjusting parameters, providing multi-functionality without proportionally increasing complexity
3Ease of manufacture
If incomplete physical system models are used, then the modeling process is simplified, but the accuracy of remaining useful lifetime estimation is reduced
Solution Approach 1:
The patent merges simplified physics-based models with data-driven models to compensate for the incompleteness of the physical model. The data-driven component learns from historical sensor data to predict lifetime, making up for the simplifications in the physics model while keeping the overall approach manageable
Data Source
Figure 1A
Figure 1B
Figure 1C
AI summary
System, method and control unit for diagnosis and life prediction of one or more electro-mechanical system (480) are provided. The method includes receiving sensor data from a plurality of sensors (482, 484) associated with operation of the electro-mechanical system (480). The method includes determining at least one system response (485) associated with at least one failure mode of the electro-mechanical system (480) from the sensor data, wherein the sensor data is indicative of the at least one failure mode of the electromechanical system (480). The method further includes receiving at least one simulated response (495) associated with the at least one failure mode of the electro-mechanical system (480), wherein the at least one failure mode is simulated on a system model of the electro-mechanical system (480). The method includes generating a hybrid model of the electromechanical system (480) in real-time based on the at least one system response (485) and the at least one simulated response (495), wherein the hybrid model combines the at least one system response (485) and the at least one simulated (495). The method also includes generating a diagnosis of the electro-mechanical system (480) based on the hybrid model, wherein the diagnosis includes identification of one or more failures in the electro-mechanical system (480) and wherein the one or more failure indicates initiation of degradation of the one or more electro-mechanical system (480). The method includes predicting a life trend (460) of the electromechanical system (480) based on the diagnosis.