Innovative time efficient method to optimize buildings' performance using Design of Experiment, Polynomial Regression and Genetic Algorithms

Dynamic thermal simulation use is essential for studying complex building systems. One of the major challenges to widely deploy optimization in this field' engineering and research is the high computational cost to perform the required simulation-based iterations. To participate in overcoming t...

Full description

Saved in:
Bibliographic Details
Published in:Journal of physics. Conference series 2021-01, Vol.1714 (1), p.12014
Main Authors: Serbouti, A, Rattal, M, Oualim, E M, Mouhsen, Az
Format: Article
Language:English
Subjects:
Artificial neural networks
Computer simulation
Design of experiments
Design optimization
Genetic algorithms
Hypercubes
Multiple objective analysis
Performance indices
Physics
Polynomials
Regression
Sensitivity analysis
Thermal simulation
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Dynamic thermal simulation use is essential for studying complex building systems. One of the major challenges to widely deploy optimization in this field' engineering and research is the high computational cost to perform the required simulation-based iterations. To participate in overcoming this hindrance, we propose in this paper an innovative time efficient Multiobjective Optimization methodology, based on coupling Latin Hypercube Design of Experiments to Artificial Neural Network polynomial regression and Genetic Algorithms (GA). This method allows to benefit from the robustness of Genetic Algorithms and from the rapidity of prediction of polynomial regression. RBD-FAST sensitivity Analysis indexes are also generated without any extra-time. We successfully applied the method to analyse and optimize a constrained 8-objectives problem with 13 input parameters in six climatic zones in Morocco.
ISSN:1742-6588
1742-6596
DOI:10.1088/1742-6596/1714/1/012014