Dam-Axis Siting with an Improved Adaptive Variable Neighborhood Search Algorithm

This study addresses the problem of upper-reservoir dam-axis siting in pumped-storage hydropower projects, where achieving a balanced cut–fill ratio (CFR) and minimizing construction cost are critical under complex terrain conditions. Existing methods often rely on manual interpretation or static GIS-based analysis, which struggle to optimize geometric alignment and earthwork equilibrium. To overcome these limitations, we propose an Improved Adaptive ariable Neighborhood Search (IAVNS) algorithm that integrates high-resolution Digital Elevation Model (DEM) data into a two-layer adaptive framework combining local geometric refinement with global exploration. The inner layer performs hierarchical position and elevation adjustments through adaptive neighborhood operators, while the outer layer conducts fitness-guided subregion migration to enhance convergence diversity. Experiments on the Qiannan pumped-storage project demonstrate that the proposed IAVNS algorithm achieves a near-balanced cut–fill ratio (CFR) of 1.31. In terms of solution quality, IAVNS shows an improvement of 22.8% over the Genetic Algorithm (GA) and 16.5% over the classical Variable Neighborhood Search (VNS). Regarding convergence speed, IAVNS performs 49.5% faster thanGA and 29.1% faster than VNS.The algorithm also attains a success rate of 76.9%, exceeding GA (27.8%) and VNS (43.2%). These results verify that IAVNS effectively enhances computational efficiency and geometric precision in dam-axis design. The method provides a promising and potentially extensible framework that could be further explored in broader hydraulic or terrain-sensitive engineering applications.