New algorithms for solving large-scale optimization problems with application to engineering design
Abstract
In this research, four new algorithms for solving large-scale problems in engineering design are presented. The four algorithms are the dynamic basis algorithm and the inscribed sphere algorithm for solving linear programming problems, the active constraint method with maximum step rule for solving linear goal programming problems and the parallel dual projection method for solving positive definite quadratic programming problems. The objective of each algorithm is to remove barriers in solving large scale engineering problems which are present in existing methods. The algorithms are tested on a variety of problems. The optimal shape design of a flywheel is generated by local geometric control and as a test of the SQP algorithm.
Degree
Ph.D.
Advisors
Sandgren, Purdue University.
Subject Area
Mechanical engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.