Research on Neural Network Adaptive Dynamic Sliding Mode Formation Algorithm for Multiple Unmanned Surface Vehicles

This paper investigates an adaptive dynamic sliding mode control algorithm based on Chebyshev neural networks, applied to the formation control problem of multiple unmanned surface vehicles (USVs). The algorithm combines the self-learning capability of neural networks with the robustness of sliding mode control, enabling stable formation control under uncertain environments, external disturbances, and actuator hysteresis. The Chebyshev neural network is utilized to estimate nonlinear functions in the system, reducing computational complexity and improving real-time performance. Additionally, the dynamic sliding mode control method effectively suppresses chattering. Simulation results demonstrate that the proposed method excels in control accuracy, real-time performance, and robustness, providing an effective solution for the formation control of multiple USVs.