I want to optimize a process. The process has 12 important process parameters ( 12 factors and each has three levels). I want to use RSM tool. The RSM is best approach with respect to optimization as compared to other techniques. But it is not capable to manage data with 12 factors. it is possible that some one split or divide the whole process into small sections (3 factors and three levels) and then optimize the process. Is it possible that someone divide the large process into small portions/sctions and then do optmization. If yes, have you any research paper for a reference. Can you share with me.
Need expert opinion and guidance
In other words you have 12 input parameters and how many outputs? The response surface or metamodel can be constructed easily with 12 or more input parameters using MATLAB or Isight or Python. Now, if you want to reduce the 12 parameters you have to perform a sensitivity analysis and observe the importance of each parameter. By doing that you can eliminate the ones that do not affect the output.
The answer is given by Plackett-Burman designs. See the following book:
https://www.amazon.com/Analysis-Experiments-Chapman-Statistical-Science/dp/1420060600/ref=sr_1_1?s=books&ie=UTF8&qid=1544671680&sr=1-1&keywords=John+lawson+Design+of+experiments
and the attached papers. Best, David Booth
Optimizing an Oil Refinery Production is one of my projects. Such a problem has many thousands of equations, many PDEs, and hundreds of constraints. Do you have any Operations Research background (OR)? An OR background would be helpful to understand the complexity. Here are two articles that may enlarge your view of the problems. Research Proposal Think Outside-the-Box, verify Business Strategies, & Operati...
andResearch Proposal Oil Refinery Production, What is the company's goal? Maximiz...
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Materials