摘要

To guide the construction of large-scale offshore wind farms, optimization for substation siting and connection topology are both necessary, which is a multiobjective optimization problem. Non-iterative methods are based on greedy strategies and they are only suitable to optimize the connection topology. Iterative methods can update the solutions iteratively to approach the optimum using common optimizers such as particle swarm and firefly algorithm (FA), which are more adaptive in multiobjective optimization. Thus, it is feasible to explore iterative methods to synchronously optimize substation siting and connection topology. This paper proposes a modified FA for the optimization of substation siting and connection topology in a large-scale offshore wind farm. The objective function comprehensively considers critical factors including substation siting, partition of wind turbines, connection topology, cable types, and power loss. The optimization ability of the proposed FA is enhanced by adopting reproduction and resetting mechanisms with dynamic hyperparameters. An implementation that bridges the topological space and Euclidean space is detailed to help with improving the convexity and continuity of search spaces. To validate the efficacy, the proposed FA is first tested in an offshore wind farm with a single substation and then it is applied in a large-scale offshore wind farm with multiple substations to demonstrate the synchronous optimization of substation siting and connection topology.

  • 单位
    南方电网技术研究中心