COMPUTER OPTIMIZATION OF NONLINEAR CONTROL SYSTEMS BY MEANS OF DIGITIZED MAXIMUM PRINCIPLE
Abstract
A digitized version of the maximum principle is derived by a simple and elementary method. In the limit of the discrete time interval approaching zero, the digitized maximum principle is reduced to Pontryagin's maximum principle. The digitized version can be readily programmed on a computer, and all four types of optimization problems: (1) minimum time, (2) maximum range, (3) minimum cost between terminal points, and (4) maximum gain (or minimum cost) in a given interval, are solvable by this method. The computation of the optimum process requires relatively little computer memory, and is exact for systems describable by a set of difference equations. The use of a computer to implement the computed optimum process in various ways is discussed. A more general condition for bang-bang control is also derived. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 1961
- Accession Number
- AD0269563
Entities
People
- S.s.l. Chang
Organizations
- New York University