Knowledgeable Opponent Models for Enemy Submarine Tactics in Training Simulators
Abstract
This report describes four models which show promise for simulating a knowledgeable opponent for enemy submarine tactics in training simulators. While the models are primarily designed to simulate an opponent by selecting his decision alternatives, the models can also be used to simulate friendly forces as well. The four approaches are: (a) the Elicited Probability approach, (b) the Adaptive Decision Modeling approach, (c) the Heuristic Search approach, and (d) the Production Rules approach. A set of attributes for rating each approach are defined and described. The attributes are in three general categories. Attributes related to the modeling capability of the approach, those related to the development required to use the approach in a sub simulation, and those that relate to the expected performance of a simulation system based on a given approach. These attributes are then used to rate each approach. Finally, several representative decisions are discussed and the method of application for each approach described.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 01, 1979
- Accession Number
- ADA076236
Entities
People
- Antonio Leal
- Don May
- Efraim Shaket