RTO-APC Coordination via Local Quadratic Approximation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional remote set-point passing strategies between Real-Time Optimization (RTO) and Advanced Process Control (APC) modules in manufacturing processes are inefficient in responding to disturbances, leading to suboptimal economic operation due to the infrequent updating of constraints and lack of awareness by the APC module about unconstrained economic optima, resulting in potential profit losses.

Innovation Solution

A method and system where a Real-Time Optimization module calculates a local quadratic approximation of the economic objective function, which is then transmitted to an Advanced Process Control module to drive the process towards a constrained economic optimum, allowing the APC module to calculate economically optimal steady-state targets at a higher frequency and adapt to changes in constraints.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the RTO module uses a rigorous nonlinear steady-state model to calculate economically optimum operating points, then the economic optimization accuracy is improved, but the calculation frequency is reduced due to computational complexity

Engineering Contradiction:
Improveeconomic optimization accuracyVSAvoidcalculation frequency
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent segments the optimization calculation into two parts: the RTO module performs rigorous nonlinear optimization at lower frequency to generate accurate economic targets, while the APC module performs quadratic programming at higher frequency to track these targets. This segmentation allows each module to operate at its optimal frequency without compromising overall optimization accuracy.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The RTO module performs preliminary calculation of economically optimum operating points using the rigorous nonlinear model at lower frequency. These pre-calculated targets are then used by the APC module for high-frequency tracking, eliminating the need for the APC module to perform complex nonlinear optimization at every cycle.

Inventive Principle:
Principle #10Preliminary action

2Speed

If the APC module operates at high frequency to dynamically drive process variables toward targets, then the response speed to disturbances is improved, but the APC module lacks knowledge of unconstrained economic optima due to receiving only constrained targets from RTO

Engineering Contradiction:
Improveresponse speed to disturbancesVSAvoidloss of economic optimization information
Core Design Contradiction:
SpeedVSLoss of information

Solution Approach 1:

The patent implements a feedback mechanism where the APC module receives not only the constrained economic targets from RTO but also the corresponding Lagrange multipliers. These multipliers provide feedback information about the sensitivity of the economic objective to constraint violations, enabling the APC module to understand the underlying economic optimization landscape without requiring full knowledge of unconstrained optima.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP2507671B1Economics-based coordination of advanced process control and real-time optimization
Publication Date: 2020.12.23 SHELL OIL CO
  • EP2507671B1 patent drawingFigure 1A
  • EP2507671B1 patent drawingFigure 1B
  • EP2507671B1 patent drawingFigure 1C

AI summary

A system and method for coordinating advanced process control and real-time optimization of a manufacturing process are provided. The system and method receive process data and economic data corresponding to the manufacturing process to be controlled and optimized. Based on the process data, the economic data and a nonlinear steady-state model of the process, an economic objective function is calculated by a real-time optimization module. A reduced-order non-linear approximation of the economic objective function is thereafter calculated by the real-time optimization module and transmitted to an advanced process control module. The advanced process control module utilizes the reduced-order non-linear approximation of the economic objective function to control the manufacturing process towards the constrained economic optimum.