The geographical positioning and spatial arrangement of AC/DC hybrid substation projects represent a sophisticated multi-factor optimization challenge, often resulting in suboptimal final location choices and structural configurations due to limitations in the algorithm's global exploration efficiency and solution refinement capabilities. Poor regional site selection and layout of DC joint construction stations can lead to an increase in construction, operation, and maintenance costs. By enhancing its optimization mechanism, the whale swarm algorithm can attain superior global search capabilities and better problem-solving abilities for optimal solutions, enabling it to effectively search for high-quality layout positions for site selection and arrangement challenges. In light of this, we explore a regional site selection and layout approach for AC/DC hybrid construction stations utilizing the improved whale swarm algorithm. We first examine the network architecture of AC/DC hybrid construction stations. Based on this analysis, we choose the minimum network operational loss as the objective function. Additionally, we establish various constraints, including load, distance, capacity, and geographical restrictions, to formulate a regional siting and layout model for AC/DC hybrid construction stations. To bolster the global search capacity of the whale swarm algorithm, we enhance it by integrating a comprehensive mutation operator and stochastic sinusoidal inertial weights. Subsequently, we employ the improved whale swarm algorithm to solve the formulated siting layout model, achieving efficient and precise determination of the optimal regional siting layout for the hybrid construction station. The results demonstrate that our method can efficiently identify a near-optimal siting layout plan with rapid convergence. The siting layout for the experimental AC/DC hybrid construction station can circumvent geographical constraints and achieve a network operational loss of just 1.87%, exhibiting excellent global search performance and optimization stability.
Chen et al. (Mon,) studied this question.