A new method of sequential approximate optimization for structural optimization problems

In this paper a new method of sequential approximate optimization is introduced to solve some structural optimization problems. This new method employs a new intermediate variable which has two variable parameters to linearize the objective and constraint functions in terms of the intermediate variable. The approximated optimization problem obtained by using this method is separable, and it can always be convex if the parameters are properly chosen. The approximated optimization problem is then simply solved by using the dual method. The new method can be considered as a generalization of the current sequential approximate optimization methods such as CONLIN, MMA and DSQP, and it can also be reduced to a modified optimality criteria method. Several guidelines and calculation methods for determining the parameters in the intermediate variable are discussed in this paper. It is shown that the new method can greatly accelerate the convergence of the optimization process compared with previous methods, and is more general and flexible to deal with various optimization problems.