다중 구속조건에 대한 위상최적화 기법의 성능 비교

Abstract

This paper presents the comparative study about optimization algorithms in optimal topology design with multiple constraints. As the optimization problem has more constraints, it becomes more difficult to find the optimum solution in the feasible region and the value of the objective function tends to converge to the local optimum. This makes the optimization methods show the different performances. To compare the performances of the optimization algorithms for the multiple constraints problems, MMA (Method of Moving Asymptotes), IPOPT (Interior Point OPTimizer), and MATLAB’s fmincon (SQP and interior-point method) were concerned for the structural topology optimization with manufacturing constraints such as symmetry and casting direction. The criteria considered to evaluate the solvers are the number of iteration and the value of objective function. These are presented by using the performance profiles which show the results effectively in a graphical way.