Hybrid particle swarm optimization and design of experiment /
Mai Salah Eldin Abdelaziz Mohamed
Hybrid particle swarm optimization and design of experiment / طريقة أمثلية حشد الجزيئات المهجنة بإستخدام تصميم التجارب Mai Salah Eldin Abdelaziz Mohamed ; Supervised Mohamed H. Gadallah , Sayed M. Metwalli - Cairo : Mai Salah Eldin Abdelaziz Mohamed , 2017 - 114 P. : charts ; 30cm
Thesis (M.Sc.) - Cairo University - Faculty of Engineering - Department of Mechanical Design and Production
A hybrid Particle Swarm Optimization algorithm and design of experiment approach is developed and tested. The hybridization between the two methods has two different ways: The first one is a trial to make use of design of experiment to reach the optimum selection and combinations of particle swarm optimization's most significant factors (maximum inertia weight }max, minimum inertia weight }min, acceleration coefficients, C1 and C2) using 3 levels orthogonal arrays OAs. An L27OA is employed to study the four factors at three levels. The particle swarm optimization is then applied on a number of benchmark problems to find the optimum solution. The second method is using design of experiments on the problem variables prior to particle swarm optimization. According to the number of parameters of the problem, a suitable orthogonal array is used and number of levels for each parameter is assigned. The obtained feasible solutions are employed as an initial swarm for particle swarm optimization algorithm instead of using large randomly selected swarms
Design of experiment Hybrid PSO Particle swarm optimization
Hybrid particle swarm optimization and design of experiment / طريقة أمثلية حشد الجزيئات المهجنة بإستخدام تصميم التجارب Mai Salah Eldin Abdelaziz Mohamed ; Supervised Mohamed H. Gadallah , Sayed M. Metwalli - Cairo : Mai Salah Eldin Abdelaziz Mohamed , 2017 - 114 P. : charts ; 30cm
Thesis (M.Sc.) - Cairo University - Faculty of Engineering - Department of Mechanical Design and Production
A hybrid Particle Swarm Optimization algorithm and design of experiment approach is developed and tested. The hybridization between the two methods has two different ways: The first one is a trial to make use of design of experiment to reach the optimum selection and combinations of particle swarm optimization's most significant factors (maximum inertia weight }max, minimum inertia weight }min, acceleration coefficients, C1 and C2) using 3 levels orthogonal arrays OAs. An L27OA is employed to study the four factors at three levels. The particle swarm optimization is then applied on a number of benchmark problems to find the optimum solution. The second method is using design of experiments on the problem variables prior to particle swarm optimization. According to the number of parameters of the problem, a suitable orthogonal array is used and number of levels for each parameter is assigned. The obtained feasible solutions are employed as an initial swarm for particle swarm optimization algorithm instead of using large randomly selected swarms
Design of experiment Hybrid PSO Particle swarm optimization