Cascaded MPC Proxy Limits for Plantwide Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Industrial process control and automation systems face challenges in providing simultaneous decentralized controls at lower levels and centralized optimization at higher levels, leading to inconsistent solutions and unreachably significant optimization benefits due to the lack of guaranteed solution consistency across multiple layers, resulting in open-loop plantwide optimization rather than closed-loop control.
Innovation Solution
A cascaded model predictive control (MPC) system is implemented, where a master MPC controller receives information from slave MPC controllers on the extent of variable changes within constraints, estimates a feasibility region, and performs plantwide optimization, using proxy limits to merge multiscale models and ensure consistency across layers, enabling decentralized and centralized control simultaneously.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If decentralized MPC solutions are implemented at lower levels, then operability and flexibility in dealing with process upsets, equipment failures, and maintenance are improved, but solution consistency across multiple layers deteriorates
Solution Approach 1:
The control system is divided into multiple hierarchical layers with master MPC controllers at the plant level and slave MPC controllers at the unit level. Each layer operates semi-independently with its own optimization objectives and constraints, allowing decentralized adaptability while maintaining overall consistency through the hierarchical structure and information exchange mechanisms.
2Device complexity
If centralized planning optimization is implemented at higher levels, then a higher-level view that distills out unessential or obscuring details is achieved, but solution consistency across multiple layers deteriorates
Solution Approach 1:
The master MPC controller merges the optimization objectives and constraints from multiple slave MPC controllers into a unified plant-wide optimization problem. This combining approach ensures that the centralized planning optimization maintains consistency with decentralized unit-level controls by integrating their respective feasibility regions and objectives into a coherent global solution.
3Ease of manufacture
If conventional control and automation systems are used, then implementation simplicity is maintained, but optimization benefits remain unreachable due to lack of closed-loop integration
Solution Approach 1:
The cascaded MPC system implements closed-loop feedback by continuously monitoring the actual process state and comparing it with the predicted state from the MPC model. The feedback mechanism allows the master and slave MPC controllers to adjust their control actions in real-time based on deviations from the optimal trajectory, ensuring that optimization benefits are realized while maintaining operational simplicity through automated control.
Data Source
AI summary
A method includes receiving, at a master model predictive control (MPC) controller from a slave MPC controller, information indicating to what extent the slave MPC controller is able to change multiple manipulated variables in each of multiple directions within a variable space without violating process variable constraints of the slave MPC controller. The method also includes estimating a feasibility region associated with the slave MPC controller using the information, where the feasibility region identifies a portion of the variable space in which combinations of manipulated variable values satisfy the process variable constraints. In addition, the method includes performing plantwide optimization at the master MPC controller using the feasibility region, where a solution generated during the plantwide optimization includes one of the combinations of manipulated variable values within the feasibility region.