In order to analyze and solve the problem of the fault reconfiguration of shipboard power systems (SPSs), a fault reconstruction model with the optimization objectives of load loss and switching operation times is established, and an improved grey wolf optimization (GWO) algorithm is proposed.

Aiming at the deficiencies of the traditional GWO algorithm, a chaotic tent map is added; a cosine function is used to improve the convergence factor so that it maintains a large value in the initial stage, then decreases slowly and increases the attenuation rate in the later stage; non-dominated ranking and congestion calculation are added to improve the decision-making grey wolf selection strategy; and the grey wolf individual is discretized so that it can be used for reconfiguration.

The example of the network fault reconfiguration of a SPS shows that in the case of load and generator faults, the Pareto solution obtained by the proposed method is smaller than the improved differential evolution (DE) algorithm, improved particle swarm optimization (PSO) algorithm and improved genetic algorithm (GA), and has certain advantages in the optimal number of iterations; that is, its optimization speed.

The proposed method can overcome the problems of the traditional GWO algorithm such as its slow convergence speed, poor diversity of initialization population and susceptibility to falling into local optimization. This study proves that a superior power system reconfiguration scheme can be obtained to better ensure the safe and stable operation of ships.