Hydrocarbon Excursion Forecasting With Ensemble Sensor Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional predictive models for hydrocarbon processing systems struggle with balancing bias and variance, sensitivity to outliers, and instability, leading to erratic predictions and inadequate forecasting of equipment excursions.
Innovation Solution
Employing an ensemble predictive model that combines multiple predictive models trained on different data sets to aggregate predictions, using a voting classifier ensemble model with adjustable voting thresholds, to enhance robustness and accuracy in forecasting future excursions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional predictive models are used for forecasting equipment excursions, then the system can provide predictions, but the predictions are erratic and unreliable due to bias-variance imbalance and sensitivity to outliers
Solution Approach 1:
The patent combines multiple individual predictive models into an ensemble model that aggregates their predictions. This merging approach reduces the bias-variance imbalance and sensitivity to outliers that plague single models, producing more reliable and accurate excursion forecasts through collective decision-making of multiple models.
Solution Approach 2:
The ensemble model functions as a composite predictive system, combining different predictive model types (e.g., decision trees, neural networks, statistical models) into a unified forecasting mechanism. This composite structure leverages the strengths of each component model while mitigating their individual weaknesses, particularly regarding bias and variance.
2Reliability
If multiple predictive models are combined into an ensemble model, then prediction reliability improves, but system complexity increases
Solution Approach 1:
The ensemble model is segmented into multiple independent predictive models, each handling specific aspects of the prediction task. This segmentation allows the system to distribute the computational burden and model complexity across separate components, making the overall system more manageable while improving reliability through diversity.
Solution Approach 2:
The ensemble model framework serves multiple functions simultaneously: it performs individual predictions through each component model, aggregates results through voting or averaging mechanisms, and provides robustness against outliers. This multi-functionality justifies the increased complexity by delivering comprehensive prediction capabilities.
3Measurement precision
If an ensemble model with voting classifier is used, then forecasting accuracy improves, but computational requirements and processing time increase
Solution Approach 1:
The individual predictive models within the ensemble are trained in advance on historical data, performing preliminary learning before actual forecasting is needed. This preliminary training allows the ensemble to quickly aggregate pre-trained model predictions during operation, reducing real-time computational burden while maintaining high accuracy.
Solution Approach 2:
The ensemble model uses copying of prediction results from multiple component models, aggregating their outputs through voting or averaging. This copying approach is computationally more efficient than having each model re-analyze data independently, as it leverages the pre-computed predictions of individual models to achieve accurate forecasting.
Data Source
AI summary
A method for predicting a future excursion in a hydrocarbon processing system includes obtaining a plurality of sensor datasets from a corresponding plurality of different sensor units of the hydrocarbon processing system, wherein a sensor dataset N of the plurality of sensor datasets corresponds to a sensor unit N of the plurality of different sensor units; applying each of the plurality of sensor datasets to a corresponding plurality of predictive models contained by an ensemble model, wherein the sensor dataset N corresponds to a predictive model N of the plurality of predictive models; providing by the plurality of predictive models a plurality of separate prediction outputs based on the plurality of sensor datasets; and providing by the ensemble model a final prediction output regarding an occurrence of the future excursion that is based on each of the plurality of separate prediction outputs of the predictive models.