Online Derived Measurements for Sensorless Process Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Industrial production processes often operate blindly due to lack of online physical measurements for process variables and product properties, leading to prolonged out-of-specification products and inefficient control.
Innovation Solution
A system and method that combines physical principles and statistical regressions with artificial intelligence/machine learning to generate online, frequently updated measurements for process monitoring, control, and optimization, using existing online and offline data to model process variables and product properties.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If physical sensors are used for online measurements, then measurement reliability is improved, but device complexity and cost increase due to additional hardware installation
Solution Approach 1:
The patent creates virtual copies of physical sensor measurements through data modeling and reconstruction algorithms. Instead of installing additional physical sensors, the system generates synthetic measurement signals based on correlations between available process data and the desired measurements, thereby achieving reliable online measurements without increasing hardware complexity
Solution Approach 2:
The patent replaces the mechanical/physical sensor system with an information-processing system. By using statistical models, machine learning algorithms, and data fusion techniques, the system substitutes physical measurement hardware with computational methods that reconstruct missing measurements from available process data
2Loss of information
If physical sensors are installed for all process variables, then measurement completeness is improved, but loss of time increases due to sensor installation, maintenance, and calibration
Solution Approach 1:
The system enables self-service measurement reconstruction by automatically generating missing measurements from available process data without requiring external sensor installation or manual calibration. The data modeling system continuously reconstructs missing measurements using real-time process data, eliminating the need for physical sensor maintenance and calibration activities
3Productivity
If measurements are taken frequently with physical sensors, then productivity is improved through better control, but loss of energy increases due to continuous sensor operation and data processing
Solution Approach 1:
The patent applies partial action by reconstructing measurements only when and where needed based on data availability and process conditions. Instead of continuously operating physical sensors for all measurements, the system selectively reconstructs missing measurements using available process data, reducing energy consumption while maintaining sufficient measurement frequency for effective process control
Data Source
AI summary
A system, method, and/or apparatus is provided for production processes in which online frequently derived measurements are determined use existing online and/or offline reference measurements and real-time and/or historical data to model process variables and/or product properties to achieve enhanced production goals.