Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd Logo
Upload

Welcome to Scribd!

  • 43 views

    Simulationoptimizationbyautomationvia Python

    Uploaded by

    Sandra Gilbert
    System design and optimization are common interests for researchers. Commercial simulation software like Aspen Plus can simulate systems using predefined components. However, optimization is more difficult. Researchers can either use the software's optimizer, which may not achieve goals, or link external programs to automate simulations and use the software for calculations. The author discusses their approach of simulating cases in Aspen HYSYS then linking it to Matlab/Python for optimization. They share code for linking HYSYS to Python and discuss challenges in automation and variable discovery. For optimization algorithms, they prefer stochastic methods like genetic algorithms and differential evolution, which are easy to apply to chemical cases. However, connecting programs and choosing algorithms does not guarantee smooth optimization, as simulations must

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd

    Simulationoptimizationbyautomationvia Python

    Uploaded by

    Sandra Gilbert
    43 views1 page
    System design and optimization are common interests for researchers. Commercial simulation software like Aspen Plus can simulate systems using predefined components. However, optimization is more difficult. Researchers can either use the software's optimizer, which may not achieve goals, or link external programs to automate simulations and use the software for calculations. The author discusses their approach of simulating cases in Aspen HYSYS then linking it to Matlab/Python for optimization. They share code for linking HYSYS to Python and discuss challenges in automation and variable discovery. For optimization algorithms, they prefer stochastic methods like genetic algorithms and differential evolution, which are easy to apply to chemical cases. However, connecting programs and choosing algorithms does not guarantee smooth optimization, as simulations must

    Original Description:

    Simulation optimization by automation via Python

    Original Title

    SimulationoptimizationbyautomationviaPython

    Copyright

    © © All Rights Reserved

    Available Formats

    DOCX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    System design and optimization are common interests for researchers. Commercial simulation software like Aspen Plus can simulate systems using predefined components. However, optimization is more difficult. Researchers can either use the software's optimizer, which may not achieve goals, or link external programs to automate simulations and use the software for calculations. The author discusses their approach of simulating cases in Aspen HYSYS then linking it to Matlab/Python for optimization. They share code for linking HYSYS to Python and discuss challenges in automation and variable discovery. For optimization algorithms, they prefer stochastic methods like genetic algorithms and differential evolution, which are easy to apply to chemical cases. However, connecting programs and choosing algorithms does not guarantee smooth optimization, as simulations must

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    43 views1 page

    Simulationoptimizationbyautomationvia Python

    Uploaded by

    Sandra Gilbert
    System design and optimization are common interests for researchers. Commercial simulation software like Aspen Plus can simulate systems using predefined components. However, optimization is more difficult. Researchers can either use the software's optimizer, which may not achieve goals, or link external programs to automate simulations and use the software for calculations. The author discusses their approach of simulating cases in Aspen HYSYS then linking it to Matlab/Python for optimization. They share code for linking HYSYS to Python and discuss challenges in automation and variable discovery. For optimization algorithms, they prefer stochastic methods like genetic algorithms and differential evolution, which are easy to apply to chemical cases. However, connecting programs and choosing algorithms does not guarantee smooth optimization, as simulations must

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    of 1

    System design and optimization are one of the common interests of many researchers.

    One of
    the easiest ways to simulate systems is to apply commercialized simulation software, such as
    Aspen Plus and Aspen HYSYS, in which a lot of commonly used components are already
    embedded. And the users also have the possibility to design their own components for some of
    the simulation software.

    However, the optimization problem is not always as easy as the design/simulation problem. In
    general, researchers have two options, if they prefer simulating systems in simulation software:
    1. using optimizer that is offered by the simulation software. However, sometimes it is not that
    easy to achieve the goal the user wants to optimize; 2. linking external programs to automate
    the simulation software and use simulation software as a calculation machine.

    Here I use my research approach as an example to give some hints to those who want to
    optimize their systems and bring some discussion, hopefully, about different approaches of
    optimization for engineering problems, especially, chemical/process/power engineering. For
    my research, first, I simulated some cases in Aspen HYSYS and then my interests go to the
    optimization. I tried several times by using the optimizer offered by Aspen HYSYS, but I was
    not satisfied with the results I got. Therefore, I was forced to connect HYSYS with
    Matlab/Python to deal with the optimization problem.

    I posted the code for connecting HYSYS with Matlab in my profile. And in this part, I will
    more focus on the link between HYSYS with Python. The codes are very simple, please see
    the attachment. However, the automation after linking two programs can be a little bit tricky.
    In Matlab, you can check which variables are available for automation, but for Python, I have
    not found any function that can do that yet. Nevertheless, the best way to discover all the
    variables that the programmers left intentionally to be connected should be via COM browser.
    There all the variable and the type of variables are listed.

    For the optimization algorithm, I prefer, and I am using the stochastic algorithms, such as
    genetic algorithm (GA) and differential evolution (DE). They do have their limitations, but
    easy to be written and applied for almost all the chemical cases. As for as I know, only DE has
    been included as a package in Scipy for minimalization. And some researchers state that DE is
    faster than GA to converge.

    However, connecting both of programs, knowing the variables for automating and choosing
    the suitable optimization algorithm do not mean that the optimization will be carried out
    smoothly. Because the optimization stops or goes in the wrong direction, if the simulation cases
    are not well designed, which leads to error-prone simulations. Therefore, be aware of your
    simulation inputs and calculation procedures when you design your systems, which will
    definitely save a lot of time later if you are not only interested in the simulation but also the
    optimization as well.

    Thanks in advance for any comments and discussions about optimization problems.

    You might also like