Efficient Computation for Large Scale Optimization.
Abstract
Several classes of algorithms for solution of the general nonlinear programming (constrained optimization) problem, and four specific implementations of these were chosen and evaluated with respect to expected speed of computation. A test problem based on the path generation problem of terrain following/terrain aviodance flight was developed, and the performance of the chosen optimization procedures was compared. It was found that the generalized reduced gradient method was faster and moe reliable that either of two augmented Lagrangian methods and a quadratic approximation method. However, the solution time for the TF/TA type problems was found to be far in excess of what would be required. Several simplifications of the problem statement were attempted in order to decrease computation time without compromising the integrity of the solution.
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 01, 1982
- Accession Number
- ADA129293
Entities
People
- John A. Fleming
Organizations
- Texas A&M University