Process Optimization via Pre-Trained Regression Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Complex processes in manufacturing and process industries are difficult to model due to their inherent complexity, limiting conventional optimization methods to sub-processes rather than process-wide optimization.
Innovation Solution
A scalable prediction-optimization framework that uses pre-trained regression models to generate a system-wide optimization model, including decision variables, constraints, and an objective function, to determine operating mode trajectories across multiple sub-processes, enabling process-wide optimization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional optimization methods are used for complex processes, then sub-process optimization is achievable, but process-wide optimization is not attainable due to inherent complexity
Solution Approach 1:
The patent segments the complex process into multiple sub-processes, each modeled by independent regression models. This allows the overall process to be optimized by coordinating these segmented models, transforming an intractable process-wide optimization problem into manageable sub-problems that can be solved individually and then integrated.
Solution Approach 2:
The patent changes the parameter representation by using regression models with adjustable parameters that capture the essential behavior of each sub-process. By parameterizing the complex process relationships through these models, the system transforms qualitative complexity into quantitative parameters that can be optimized mathematically.
2Productivity
If process-wide optimization is attempted, then system-wide performance improvement is achieved, but computational scalability deteriorates
Solution Approach 1:
The patent performs preliminary action by pre-training regression models for each sub-process before the actual optimization. These pre-trained models capture the steady-state and transient behavior of sub-processes, allowing the subsequent process-wide optimization to proceed efficiently without re-computing fundamental process relationships during the optimization itself.
Solution Approach 2:
The patent introduces dynamics by using regression models that can adapt to changing operating conditions. The models capture transient behavior and can be updated as process conditions change, allowing the optimization system to remain computationally efficient while adapting to dynamic process environments.
3Measurement precision
If multiple regression models are integrated for system-wide optimization, then modeling accuracy improves, but model complexity increases
Solution Approach 1:
The patent merges multiple individual regression models into a unified process-wide optimization framework. By combining the sub-process models with the optimization model, the system achieves accurate process behavior prediction while maintaining a structured approach that prevents the complexity from becoming unmanageable.
Solution Approach 2:
The patent creates a universal optimization framework that can handle multiple sub-processes with different characteristics. The regression models serve multiple functions: they predict steady-state behavior, capture transient responses, and provide the mathematical structure needed for optimization, reducing the need for separate models for each function.
Data Source
AI summary
An apparatus and method for optimizing a process, comprising: receiving live operational data associated with a plurality of sub-processes of a process; selecting a pre-trained regression model from a plurality of pre-trained regression models for each sub-process of the plurality of sub-processes; generating a system-wide optimization model comprising a multi-period mathematical program model, including: one or more decision variables; a plurality of constraints, wherein: a first constraint of the plurality of constraints comprises one of the pre-trained regression models, and a second constraint of the plurality of constraints comprises an operational constraint; and an objective function; generating, via the optimization model, an operating mode trajectory comprising a plurality of intermediate operating modes at a plurality of intermediate times during a planning interval; and displaying a set-point trajectory recommendation in a graphical user interface based on the operating mode trajectory.