Process Systems Enterprise Limited
email this page print this page
pdf overview

gPROMS® in control

gPROMS is widely known as an Advanced Process Modelling environment. Less well-known is the fact that — as a mathematical modelling and solution engine — it can be used in many areas of control design and validation, either stand-alone or in conjunction with industry standard software such as The Mathworks' MATLAB® and Simulink®.

Control system design and optimisation capabilities

The standard modelling capabilities of gPROMS provide numerous facilities for control system design and optimisation. Here are a few:

Simultaneous process and control system design. gPROMS can be used to design and optimise equipment and control system simultaneously, for example, to determine the best reactor diameter and controller tunings in a single run.

Optimal control scheme selection. gPROMS' mixed-integer optimisation capabilities will determine the best-performing control scheme from a number of alternatives, given a range of process disturbances, while simultaneously calculating optimal tuning constant values.

Development and validation of operating procedures. gPROMS' dynamic optimisation capabilities can be used to determine optimal operating procedures for start-up, shutdown and general grade transition, as well as optimal trajectories for batch operations.

Dynamic optimisation for controller tuning. Controllers built within gPROMS can be optimally tuned for a range of disturbances with a single optimisation run.

gPROMS software products and capabilities for Control Engineering

In addition, the following gPROMS products and capabilities are specifically aimed at Control Engineers.

gO:Simulink (the gPROMS Object for Simulink)is a unique and powerful tool that enables control engineers to use complex non-linear gPROMS Advanced Process Models directly within The Mathworks' Simulink® environment.

Simulink flowsheet gPROMS distillation column model embedded within a Simulink application

This enables control engineers to work in their environment of choice using the detailed gPROMS non-linear Differential-Algebraic Equation (DAE) models built by process engineers during the equipment design

gPROMS takes care of all the mathematics required to reduce the problems to the form expected by Simulink, however complex the set of partial and ordinary differential and algebraic equations within the gPROMS model.

The converse option allows the execution of Simulink models within gPROMS, for additional speed.

gO:MATLAB (the gPROMS Object for MATLAB) is a similar tool that enables control engineers to execute gPROMS Advanced Process Models directly within The Mathworks' MATLAB® environment.

This enables control engineers to take advantage of all the MATLAB toolbox facilities without compromising on process model fidelity.

As with gO:Simulink, the solution of the gPROMS models is hidden from the user.

The model linearisation capability

Linearisation

gPROMS task for generating the linear dependencies of distillate flow and purity on reactor inlet flows, concentration and temperature:

gPROMS linearisation section

It is possible to use the gPROMS LINEARISE command from within ModelBuilder to generate a linearised model at any point during the execution of a gPROMS simulation.

The linearised model can then be used — for example — in linear control system design techniques, or in Model Predictive Control application.

Using gPROMS to generate linearised models means that you can:

The user can specify any subset of the model variables {x, y, u} as inputs of interest U and outputs Y

gPROMS automatically determines the minimal subset X of states x required to express the effects of U on Y, computes the A, B, C, D matrices and transmits information to external software via the gPROMS Foreign Process Socket

Simulink Connection Partner logoPSE is a Simulink Connections partner
MATLAB Connection Partner logoPSE is a MATLAB Connections partner