Stochastic modeling-based adaptive control for maritime defense in simulation computer games

The object of the study is the modeling process of virtual adversary behavior and automated control systems for mine weapons in game-based naval combat scenarios, taking into account uncertainty and incomplete information, particularly in conditions of partial or erroneous functioning of the sensor system. One of the most problematic aspects is ensuring effective decision-making in situations where the sensor system exhibits Type I and Type II errors or its feedback is completely absent due to malfunctions or damage. The study employs stochastic modeling methods, mathematical expectation estimation for all possible combat scenarios, and adaptive control algorithms that consider the accuracy of the sensor system and the a priori probability of enemy presence. An adaptive control method for anti-ship defense and a corresponding implementation system have been developed, which includes an adaptive controller capable of performing the core computations in real time to determine optimal control actions for mine weapon deployment. The results of numerical experiments were obtained for various scenarios: with fixed parameters, variable minefield density, sensor system accuracy changes, and different a priori probabilities of ship appearances. These experiments enabled a comprehensive evaluation of the method's effectiveness. The conducted experiments confirm that the proposed method enables effective control of mine weapons in the presence of Type I and Type II errors with probabilities ranging from 0 to 0.9 during the detection of enemy and neutral ships. As a result, the proposed solution provides the capability for adaptive control of combat operations even under high uncertainty, enhances the realism of virtual adversary behavior in simulation games, and lays the groundwork for the development of intelligent automatic control systems in naval combat scenarios.