Multi-UUV Maneuvering Counter-Game for Dynamic Target Scenario Based on Fractional-Order Recurrent Neural Network

• Published in: IEEE transactions on cybernetics Vol. 53; no. 6; pp. 4015 - 4028
• Main Authors: Liu, Lu, Zhang, Shuo, Zhang, Lichuan, Pan, Guang, Yu, Junzhi
• Format: Journal Article
• Language: English
• Published: United States IEEE 01.06.2023; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Decision making; fractional calculus; Fuzzy sets; Game theory; Games; Heuristic algorithms; interval fuzzy game theory; Maneuvers; multi-underwater unmanned vehicle (UUV); Neural networks; Optimization; recurrent neural network (RNN); Recurrent neural networks; Strategy; Uncertainty; Underwater; Unmanned vehicles; Vehicle dynamics
• ISSN: 2168-2267; 2168-2275; 2168-2275
• DOI: 10.1109/TCYB.2022.3225106

In this article, a multi-underwater unmanned vehicle (UUV) maneuvering decision-making algorithm is proposed for a counter-game with a dynamic target scenario. The game is modeled with interval-valued intuitionistic fuzzy rules, and an optimal maneuvering strategy is realized using a fractional-order recurrent neural network (RNN). First, underwater environments with weak connectivity, underwater noise, and dynamic uncertainties are analyzed and incorporated into the interval-valued intuitionistic fuzzy set. Then, the maneuvering decision-making model and the expected return of the multi-UUV countermeasure are designed based on the interval-valued intuitionistic fuzzy rules. Subsequently, to optimize the counter-game maneuvering strategy, a fractional-order RNN is formulated based on the Karush-Kuhn-Tucker optimality conditions. In addition, the existence and uniqueness of the optimal maneuvering solutions as well as the stability of the equilibrium point are discussed. Finally, simulation and experimental results are compared to determine the effectiveness of the proposed algorithm. The influence of the fractional order on the convergence rate and optimization error of the proposed algorithm is also minutely examined.