| Article ID: | iaor2000999 |
| Country: | United Kingdom |
| Volume: | 19 |
| Issue: | 6 |
| Start Page Number: | 377 |
| End Page Number: | 392 |
| Publication Date: | Nov 1998 |
| Journal: | Optimal Control Applications & Methods |
| Authors: | Seierstad Atle |
A solution of the maximum principle is optimal if it is ‘surrounded’ by solutions of the maximum principle, or ‘embedded in a field of extremals’. An extension of this well-known principle to infinite horizon problems is stated, and a proof of it is outlined. It is especially useful in non-concave problems, where sufficient conditions based on concavity fail. The usefulness of such an extension is illustrated in two examples.