| Article ID: | iaor19921379 |
| Country: | United Kingdom |
| Volume: | 1991, Ellis Horwood Ltd., 0-13-382615-5, Levy |
| Start Page Number: | 429 |
| End Page Number: | 85 |
| Publication Date: | Oct 1991 |
| Journal: | Heuristic Programming in Artificial Intelligence: The Second Computer Olympiad |
| Authors: | Grigoriev Andrei V. |
| Keywords: | combinatorial analysis, optimization: simulated annealing, game theory |
Instead of traditional exhaustive search procedures in logical games programming, a new method of analysis of variations is proposed under the simulated annealing algorithm scheme. The statistical mechanics concept of a global energy minimum is set to be a counterpart of a winning position in a game. The described approach has minimal memory requirements and may be used as an additional tool for the integral estimation of a position. Applicability of the method to various strategy games is discussed.