Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Multi-objective Optimization of Pulsed Gas Metal Arc Welding Process Using Neuro NSGA-II
Abstract Weld quality is a critical issue in fabrication industries where products are custom-designed. Multi-objective optimization results number of solutions in the pareto-optimal front. Mathematical regression model based optimization methods are often found to be inadequate for highly non-linear arc welding processes. Thus, various global evolutionary approaches like artificial neural network, genetic algorithm (GA) have been developed. The present work attempts with elitist non-dominated sorting GA (NSGA-II) for optimization of pulsed gas metal arc welding process using back propagation neural network (BPNN) based weld quality feature models. The primary objective to maintain butt joint weld quality is the maximization of tensile strength with minimum plate distortion. BPNN has been used to compute the fitness of each solution after adequate training, whereas NSGA-II algorithm generates the optimum solutions for two conflicting objectives. Welding experiments have been conducted on low carbon steel using response surface methodology. The pareto-optimal front with three ranked solutions after 20th generations was considered as the best without further improvement. The joint strength as well as transverse shrinkage was found to be drastically improved over the design of experimental results as per validated pareto-optimal solutions obtained.
Multi-objective Optimization of Pulsed Gas Metal Arc Welding Process Using Neuro NSGA-II
Abstract Weld quality is a critical issue in fabrication industries where products are custom-designed. Multi-objective optimization results number of solutions in the pareto-optimal front. Mathematical regression model based optimization methods are often found to be inadequate for highly non-linear arc welding processes. Thus, various global evolutionary approaches like artificial neural network, genetic algorithm (GA) have been developed. The present work attempts with elitist non-dominated sorting GA (NSGA-II) for optimization of pulsed gas metal arc welding process using back propagation neural network (BPNN) based weld quality feature models. The primary objective to maintain butt joint weld quality is the maximization of tensile strength with minimum plate distortion. BPNN has been used to compute the fitness of each solution after adequate training, whereas NSGA-II algorithm generates the optimum solutions for two conflicting objectives. Welding experiments have been conducted on low carbon steel using response surface methodology. The pareto-optimal front with three ranked solutions after 20th generations was considered as the best without further improvement. The joint strength as well as transverse shrinkage was found to be drastically improved over the design of experimental results as per validated pareto-optimal solutions obtained.
Multi-objective Optimization of Pulsed Gas Metal Arc Welding Process Using Neuro NSGA-II
Pal, Kamal (Autor:in) / Pal, Surjya K. (Autor:in)
Journal of The Institution of Engineers (India): Series C ; 100 ; 501-510
10.05.2018
10 pages
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Multi-objective Optimization of Pulsed Gas Metal Arc Welding Process Using Neuro NSGA-II
| Online Contents | 2019
Multi-objective optimization of laminated composite beam structures using NSGA-II algorithm
| British Library Online Contents | 2017
Multi-objective optimization of laminated composite beam structures using NSGA-II algorithm
| British Library Online Contents | 2017
| DOAJ | 2020
NSGA-II based multi-objective optimization in design of Pall friction dampers
| Online Contents | 2013