Industrial Regime Change Detection Using Weighted Predictive Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to accurately detect changes in the regime of connected industrial assets or processes and anomalies in their operation, which is critical for maintaining business continuity and optimizing asset management.
Innovation Solution
A method and system that utilize differential equations, machine learning models, and a specialized meta-model to continuously assign weights to predictive models, allowing for real-time detection of regime changes and anomalies by measuring differences between predicted and observed data, and updating calculations sequentially.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional monitoring methods are used to track industrial asset regimes, then the system structure remains simple, but the detection precision of regime changes and anomalies is insufficient
Solution Approach 1:
The patent segments the monitoring system into multiple specialized predictive models, each responsible for detecting specific regime characteristics. These models process different aspects of asset behavior independently and their results are aggregated, allowing high detection precision without requiring a single overly complex monolithic system.
Solution Approach 2:
The patent introduces a meta-model as an intermediary layer that aggregates predictions from multiple specialized predictive models. This meta-model coordinates the outputs of individual models and produces the final regime detection result, enabling complex detection capabilities while maintaining modular model structures.
2Measurement precision
If multiple predictive models are aggregated to improve detection accuracy, then the predictive precision increases, but the computational complexity and processing time increase
Solution Approach 1:
The patent performs preliminary calibration of differential equations using machine learning models during an offline training phase. This pre-processing step prepares the predictive models in advance, so that during real-time operation, only lightweight inference and aggregation are required, significantly reducing online computational time while maintaining high accuracy.
Solution Approach 2:
The patent implements dynamic weighting of predictive models through the meta-model, which adaptively adjusts the contribution of each model based on current operating conditions. This dynamic approach allows the system to use only the most relevant models for each specific regime, reducing computational overhead while maintaining predictive accuracy.
3Reliability
If differential equations are calibrated using machine learning models to improve regime detection, then the detection reliability improves, but the device complexity and calibration requirements increase
Solution Approach 1:
The patent implements self-calibration mechanisms where the machine learning models automatically adjust differential equation parameters using operational data without requiring manual intervention. The system learns from historical data and continuously refines its predictions, maintaining high reliability while reducing the complexity of manual calibration procedures.
Solution Approach 2:
The patent transforms the calibration problem into a parameter optimization task where machine learning models automatically adjust differential equation parameters based on observed asset behavior. This approach converts complex manual calibration into an automated parameter tuning process, improving reliability while managing calibration complexity through algorithmic optimization.
4Speed
If real-time sequential updating of predictive models is implemented to detect regime changes quickly, then the reactivity improves, but the computational load and energy consumption increase
Solution Approach 1:
The patent implements periodic re-calibration and updating of predictive models at strategically chosen intervals rather than continuous updating. The meta-model determines when regime transitions are likely and triggers model updates only at these critical moments, maintaining high reactivity to actual regime changes while minimizing unnecessary computational energy consumption during stable operation.
Data Source
Figure 1
Figure 2
AI summary
A method for detecting a change in the regime of a connected industrial asset or process and represented in operating data measurements from the industrial asset or process. The method comprises providing a database containing data generated from the operating data measurements from the industrial asset or process; providing differential equations describing different regimes of the industrial asset or process; applying machine learning models to the data in the database generated by the industrial asset or process, to calibrate the differential equations; basing at least one predictive model on the differential equations, to obtain a predictive model or group of predictive models per regime of the industrial asset or process;an aggregation of the predictive model or group of predictive models to predict a prediction of the data generated by the asset or industrial process; an implementation of a specialized metamodel to continuously assign, at determined time steps, a weight to each predictive model so as to maximize a predictive power of the whole, and sequentially verify the predictive power, by measuring at each time step, for the data generated by the asset or industrial process, a difference between data predicted by the aggregation of models and data observed in the aggregation step; a measurement of the weight of each predictive model in the aggregation, in order to determine a dominant predictive model having the greatest weight among all the weights measured within the aggregation; an observation of a process of change of regime of the asset or process when another predictive model becomes dominant within the aggregation;a sequential updating of all the calculations necessary in the preceding steps; and taking into account the observation of a change process resulting from the observation of a change process.;