Task allocation in multi-AUV dynamic game based on interval ranking under uncertain information

Abstract

This paper presents a novel game allocation method for solving the task allocation problem in underwater multi-AUV dynamic confrontations. The proposed method addresses incomplete information using a multi-objective evaluation model and interval ranking. An improved particle swarm optimization (PSO) algorithm, incorporating a hybrid strategy, is introduced to determine the Nash equilibrium of the game model. Firstly, the paper introduces the problem of underwater multi-AUV confrontations and establishes a game model based on interval numbers. Secondly, it develops a multi-target interval evaluation model considering survival value, strike income, and ammunition consumption. Combined with the analytic hierarchy process (AHP), each objective in the evaluation model is weighted to adapt to the battlefield situation with changing confrontation tasks and decision preferences. TOPSIS, relative entropy, and probability degree sort the interval income for optimization. Finally, an improved PSO algorithm called GSPSO, combining the good point set theory (G) and speed control factor (S), solves the optimal allocation problem. Simulation results demonstrate that the probability degree-based ranking method achieves higher resolution incomes, aiding precise optimization. The algorithm improves convergence speed, accuracy, and global optimization ability by increasing population diversity and iterative effectiveness, ensuring real-time decision-making in complex game scenarios.