Double-Layer Reactive Power Optimal Configuration Method for Large-Scale Offshore Wind Farms Based on an Adaptively Improved Gravitational Search Algorithm

To address the issue of frequent power frequency overvoltage disconnection accidents in offshore WF caused by the capacitive effect of submarine cables, this paper proposes a double-layer RP optimal configuration method for large-scale offshore WF based on an adaptively improved GSA. Firstly, this paper considers both the RP capabilities of offshore WT themselves and RP compensation equipment, designing a two-layer “configuration-control” optimization framework for RP. The upper layer establishes an optimization configuration model with the objective of minimizing the total investment cost and operational expenses of the equipment. The lower layer establishes a RP optimization operation model with the objective of minimizing a weighted index that comprehensively considers system network losses, voltage deviations, and RP capacity margins. Then, to address the issue of traditional GSA being prone to local optima, this paper introduces a random factor into the mass calculation, combines elite concepts to selectively synthesize gravitational forces based on fitness values, and assigns larger random numbers to forces corresponding to superior particles. By introducing control parameters to adaptively update particle positions, an adaptively improved GSA is proposed, which is employed to solve the established double-layer RP optimization configuration model for large-scale offshore WF. Finally, simulation analysis is conducted on a large-scale offshore WF constructed using MATLAB R2020a. Compared with the basic GSA algorithm, the proposed method reduces the system loss by 50.59% and the voltage deviation by 64.75%. The research demonstrates that the proposed method can effectively enhance the stability of grid voltage and proves the effectiveness of the improved GSA and the proposed two-layer “configuration-control” optimization model.