Chemical Process Soft Sensor Recalibration for Stable Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current soft sensors for chemical processes, particularly in multi-step preparations like methionine production, require manual recalibration and lack objective quality assessment, leading to subjective control decisions and inefficiencies.
Innovation Solution
An automatically re-calibrating soft sensor system that adjusts its calibration based on deviations between predicted and actual process values, using a training set augmented with new data to refine its predictions and integrate them into the distributed control system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual recalibration of soft sensors is performed, then control accuracy can be maintained, but operator workload increases and subjective judgment reduces consistency
Solution Approach 1:
The soft sensor system performs automatic self-calibration by utilizing process data and laboratory measurement data to recalibrate itself without operator intervention. The system automatically detects when recalibration is needed, performs the calibration using available data, and updates its model parameters autonomously, thereby eliminating manual recalibration tasks while maintaining control accuracy.
Solution Approach 2:
The system implements a feedback mechanism where laboratory measurements are continuously compared with soft sensor predictions. When deviations exceed predefined thresholds, the system automatically triggers recalibration using the laboratory data as reference, creating a closed-loop system that maintains accuracy without requiring continuous operator monitoring or manual intervention.
2Ease of operation
If automated recalibration is implemented, then operator workload is reduced, but system complexity increases
Solution Approach 1:
The system leverages existing multi-functional infrastructure including the distributed control system (DCS), laboratory information management system (LIMS), and data historians. By reusing these existing systems for calibration purposes rather than building dedicated calibration hardware, the automated functionality is added without proportionally increasing system complexity.
Solution Approach 2:
The system introduces a calibration manager as an intermediary software component that coordinates between the soft sensor model, process data sources, laboratory measurement systems, and the DCS. This intermediary handles the complexity of data integration, model updating, and system coordination, isolating the automated calibration logic from the core control system and making the overall architecture more manageable.
3Measurement precision
If frequent recalibration is performed, then measurement accuracy is maintained, but loss of time occurs due to data collection and processing
Solution Approach 1:
The system continuously collects and stores process data and laboratory measurements in data historians and LIMS in advance, preparing calibration datasets before they are needed. When recalibration is triggered, the system can immediately use pre-collected data without requiring additional measurement time, thereby maintaining accuracy while minimizing recalibration downtime.
Solution Approach 2:
The system performs partial recalibration by updating only the specific model parameters that have drifted, rather than recalibrating the entire soft sensor model from scratch. This selective approach maintains measurement accuracy for affected parameters while significantly reducing the time and computational resources required compared to full model recalibration.
Data Source
AI summary
A method for controlling a chemical process, by preparing methanol, hydrogen sulfide, methyl mercaptan, hydrocyanic acid, acrolein, 3-methylthiopropionaldehyde, 5-(2-methylmercaptoethyl)-hydantoin, methionine, a salt of methionine, and a derivative of methionine. The method includes providing a training set TS1, wherein TS1 is process values PV1 and process values PV2 being correlated to one another, and/or laboratory values LV1 and process values PV2 being correlated to one another. The method includes training a processing unit on the training set TS1 to identify a pattern of correlation between one or more measured process variables and at least one process variable. The method includes developing a calibration function CF1 for a calibrated soft sensor from the identified pattern of correlation and predicting at least one operating parameter for the chemical process as an approximation to LV1 and/or PV1. A system for controlling a chemical process.