PRESTRESSED I-GIRDER OPTIMIZATION USING GENETIC ALGORITHM

• Main Author: ADIBASKORO (NIM : 15010086), TITO
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/24549
• Institution: Institut Teknologi Bandung

Genetic Algorithm (GA) is a method of optimization mimicking the very process of evolution and natural selection. The results can be seen in nature, as all <br /> <br /> <br /> creatures are optimized to fit their environments. Proves to work rather well, it suggests that this method can be adopted to optimize engineering problems. With <br /> <br /> <br /> the right set up and modeling of individuals and their environment, Genetic Algorithm can optimize pretty much any kind of problems, including - in this case - prestressed I-girder. With it usually being costly, optimization will save considerable amount of budget as well as resource. With numerical approach on many cases, and using AASHTO LRFD 2007 as code and constraints, thorough analysis is done, including ultimate strength, service stresses and deflection, detailing, geometrical feasibility, etc. One of the results of the present study is a GA based optimization software with the suitable approach in order to carry out the process of optimizing prestressed I-girders, including the proving of the effectiveness of the newly applied improvements. By using this software, the other result of this study is to show the best result as well as the other optimum <br /> <br /> <br /> solutions, since running it multiple times will give multiple yet closely optimized selections of results. Lastly, sensitivity analysis is carried out to check the <br /> <br /> <br /> uncertainty level of the software performance regarding the input variation.