NEWSUMT-A: A General Purpose Program for Constrained Optimization Using Constraint Approximations

Abstract

The solution of complex constrained optimization problems is often prohibitively expensive because of the computational expense of evaluating the constraints. Under these circumstances the only viable solution is to employ inexpensive constraint approximations accompanied by move-limits that guard against large errors in the approximated constraints. The NEWSUMT-A program is a general purpose optimization program which automates the process of selecting constraint approximations and move-limits to ensure smooth convergence. The present paper describes the theoretical basis and the implementation of the optimization procedure. Examples from the field of structural optimization are presented. These involve stress and displacement constraints which require a costly finite element analysis. Additionally, the performance of two first-order constraint approximations is compared.

Optimal design of imperfect, anisotropic, ring-stiffened cylinders under combined loads

“…The problem defined in (1) is solved using a sequential unconstrained minimization technique based on the quadratic extended interior penalty function (Haftka and Starnes 1976) implemented in the NEWSUMT-A computer program (Grandhi et al 1985). In NEWSUMT-A the unconstrained pseudo-objective function that is minimized tends to be numerically ill-conditioned as the design vector, X, approaches a constraint boundary.…”

Reliability-based optimization - A proposed analytical-experimental study

“…In both cases, the algorithm that was used for the optimization was an extended interior penalty function technique incorporated in the code NEWSUMT-A. 15 This solves linear and nonlinear optimization problems with both equality and inequality constraints. The user provides subroutines for objective function and constraint evaluation.…”

A Survey of Recent Developments in Rotorcraft Design Optimization

“…The effect of coupling between blade response and the trim solution was not considered. The optimization problem was solved using new sequential unconstrained minimization technique (NEWSUMT), 15 a constrained optimization program based on the extended penalty function and Newton's method with approximate second derivatives. Numerical results showed a reduction in the 4 per revolution vertical hub shear of 15-40%.…”