Layered Economic Optimization for Paper Machine Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing model-based industrial process controllers face challenges in optimizing economic processes without compromising regulatory control, as they often require deep integration with control problems and can result in significant deviations from desired process output targets.
Innovation Solution
A layered approach is introduced, where economic optimization is performed separately from the standard control algorithm, using an extended cost function with adaptive weight parameters to balance economic optimization with regulatory control, ensuring minimal impact on process output target tracking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If economic optimization is deeply integrated with control problems, then economic optimization capability is improved, but process output target tracking accuracy deteriorates
Solution Approach 1:
The system divides the control architecture into two independent layers: a regulatory control layer that ensures accurate tracking of process output targets, and an economic optimization layer that optimizes economic objectives. This segmentation allows each layer to perform its specific function without interfering with the other, resolving the contradiction between economic optimization and target tracking accuracy.
Solution Approach 2:
The system introduces an intermediary adaptive weight parameter that mediates between economic optimization and regulatory control. This parameter dynamically adjusts the influence of economic objectives on the control algorithm, ensuring that economic optimization occurs without compromising the accuracy of process output target tracking.
2Productivity
If deep integration of optimization with control is performed, then economic optimization is achieved, but system complexity increases
Solution Approach 1:
By segmenting the system into independent regulatory control and economic optimization layers, the patent avoids the complexity of deep integration while still achieving economic optimization. Each layer operates with its own algorithms and parameters, simplifying the overall system architecture.
Solution Approach 2:
The control algorithm is designed to serve multiple functions: it performs both regulatory control for accurate target tracking and economic optimization through the extended objective function. This multi-functionality is achieved without increasing complexity by using a unified control framework that accommodates both objectives.
3Reliability
If adaptive weight parameters are adjusted based on tracking errors, then balance between optimization and control is improved, but computational complexity increases
Solution Approach 1:
The system uses feedback from tracking errors to dynamically adjust the adaptive weight parameter. This feedback mechanism ensures that the balance between economic optimization and regulatory control is maintained by increasing the weight on regulatory control when tracking errors are large, and allowing more economic optimization when tracking is accurate.
Solution Approach 2:
The adaptive weight parameter adjustment applies only the necessary degree of correction to maintain balance, rather than completely overriding economic optimization or regulatory control. This partial action approach maintains reliability while avoiding excessive computational complexity.
Data Source
AI summary
A method includes executing a control algorithm for an industrial process using a controller objective function, where the industrial process is associated with at least one controlled variable. The method also includes executing an optimization algorithm for the industrial process using an extended version of the controller objective function. The extended version of the controller objective function includes one or more additional terms added to the controller objective function, and one or more results of the optimization algorithm are provided to the control algorithm. The method further includes, based on tracking errors associated with the at least one controlled variable, adjusting at least one adaptive weight parameter in the extended version of the controller objective function. The at least one adaptive weight parameter is associated with at least one of the one or more additional terms.