Differential Dynamic Programming - A Unified Approach to the Optimization of Dynamic Systems.
Abstract
Several exact expressions for the change in cost, delta, due to a change in control are derived. These expressions are shown to play a central role in control theory and computation, in that they facilitate the derivation, not only of optimization algorithms, employing strong variations in control (the differential dynamic programming algorithms), but also of conditions of optimality. The expressions are particularly useful for deducing, under suitable conditions, sufficiency of the minimum principle for (local) optimality. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1971
- Accession Number
- AD0732008
Entities
People
- D. Q. Mayne
Organizations
- Harvard University