Article ID: | iaor2004712 |
Country: | United Kingdom |
Volume: | 44 |
Issue: | 12 |
Start Page Number: | 1589 |
End Page Number: | 1598 |
Publication Date: | Dec 2002 |
Journal: | Computers & Mathematics with Applications |
Authors: | Snyman J.A., Hay A.M. |
Keywords: | programming: quadratic |
In this paper, a constrained optimization method, called the Dynamic-Q method, is presented. Simply stated, the method consists of applying an existing dynamic trajectory optimization algorithm to successive spherical quadratic approximations of the actual optimization problem. The Dynamic-Q algorithm has the advantage of having minimal storage requirements, thus making it suitable for problems with large numbers of variables. The Dynamic-Q method is tested and results obtained are compared to results for a sequential quadratic programming (SQP) method. Indications are that the new method is robust and efficient, and a particularly well suited to practical engineering optimization problems.