Optimasi Multi-objektif pada 3-PUU Paralel Manipulator menggunakan Metode Genetic Algorithm

Setya, Wega Tama Adi (2024) Optimasi Multi-objektif pada 3-PUU Paralel Manipulator menggunakan Metode Genetic Algorithm. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6007212012-Master-Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2026. Download (4MB) | Request a copy

Abstract

Paralel manipulator banyak diaplikasikan dalam proses pemesinan, seperti proses pengelasan Friction Stir Welding (FSW). Dalam penelitian ini 3-PUU paralel manipulator dipilih untuk proses FSW, sehingga penelitian ini bertujuan untuk mensintesa dimensi dan memaksimalkan indeks performa. Dengan memanfaatkan metode genetic algorithm (GA) pada penelitian ini dilakukan empat permodelan optimasi. Pertama bertujuan untuk memaksimalkan indeks performa kinematika yang diwakili dengan local conditioning index (LCI), kedua memaksimalkan indeks performa dinamika yang diwakili dengan mass matrix (d1), ketiga dan keempat memaksimalkan kedua performa indeks menggunakan multi-objektif dengan nilai weigthing coefficient yang berbeda. Terdapat tiga parameter dimensi yang dicari pada proses optimasi indeks performa, yaitu jarak titik koordinat base ke titik Ai (a), radius moving platform (b), dan panjang lengan (l). Dari keempat permodelan optimasi menggunakan metode GA, didapatkan bahwa parameter dimensi hasil optimasi keempat dapat dipilih sebagai implemetasi robot FSW karena menghasilkan nilai indeks performa 0.76 dengan dimensi parameter a = 530 mm, b = 195 mm, dan l = 770 mm serta volume kerja las yang paling besar dengan nilai 〖2910 mm〗^3.
=================================================================================================================================
Parallel manipulators are widely applied in machining processes, such as Friction Stir Welding (FSW) welding process. In this study 3-PUU parallel manipulators were selected for the FSW process, so this study aimed to synthesize dimensions and maximize performance indices. Using the genetic algorithm (GA) method, four optimization models were performed in this study. The first aims to maximize the kinematic performance index represented by the local conditioning index (LCI), the second is to maximize the dynamic performance index represented with the mass matrix (d1), the third and fourth maximize both index performance using multi-objective with different values of the weigthing coefficient. There are three dimensional parameters that are sought in the performance index optimization process, namely the distance of the coordinate point of the base to the point Ai (a), the radius of the moving platform (b), and the length of the arm (l). From the four optimization modeling using the GA method, it was obtained that the dimensional parameter of the fourth optimization result can be chosen as an implemetation of the FSW robot because it produces a performance index value of 0.76 with parameter dimensions a = 530 mm, b = 195 mm, and l = 770 mm and the largest weld working volume with a value of 〖2910 mm〗^3.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Paralel manipulator, Optimasi kinematika, Optimasi dinamika, Optimasi Multi-objektif, Genetic Algorithm, kinematic optimization, Dynamic optimization, Multi-objectiv optimization
Subjects:	Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. Q Science > QA Mathematics > QA336 Artificial Intelligence Q Science > QA Mathematics > QA402.5 Genetic algorithms.
Divisions:	Faculty of Industrial Technology and Systems Engineering (INDSYS) > Mechanical Engineering > 21101-(S2) Master Thesis
Depositing User:	Wega Tama Adi Setya
Date Deposited:	10 Feb 2024 17:47
Last Modified:	10 Feb 2024 17:47
URI:	http://repository.its.ac.id/id/eprint/106617

Actions (login required)

View Item