Reconfiguration of Unmanned Ship Microgrid Based on Intelligent Fusion Algorithm

Aiming at the problem that unmanned ship microgrids need to be quickly reconstructed to restore power supply after a fault occurs, this study proposes an unmanned ship microgrid fault reconstruction method based on the integration of ant colony algorithm and reinforcement learning. By integrating dynamic reward and punishment mechanism and reinforcement learning Q-learning, the power grid reconstruction system can respond to dynamic load changes and select the optimal path. This method overcomes the defects of traditional optimization algorithms that are prone to fall into local optimality and low computational efficiency. Experimental results show that this method is superior to existing algorithms in terms of convergence speed, load recovery efficiency and generator set balance, which verifies its superiority in complex power grid environments.