Pumped Hydro Storage (PHS) complements renewable energy sources (e.g., wind, photovoltaic) by mitigating their intermittency and volatility, thereby enhancing power supply reliability and efficiency. However, PHS site selection is constrained by topography, water resources, environment, and economics, limiting exploitable sites. The scientific aim herein is to address this constraint by establishing a framework for large-scale autonomous identification of potential PHS sites suitable for two topological structures in Northwest China (T1: linking existing reservoirs; T2: combining existing and new reservoirs). This research develops a four-stage framework: exclusion/evaluation indices (via threshold theory and Delphi method), ArcPy-based suitability models for rapid wide-area site identification and capacity calculation, and multi-criteria decision-making for candidate ranking. This approach extends existing research by focusing on dual topological structures in complex regions, overcoming limitations of single-structure or small-scale studies. Results show 1294 potential sites (T1:33; T2:1261) with total storage capacity of 23,391 GWh. High-suitability T1 sites concentrate in northern Xinjiang, while T2 sites dominate Xinjiang's Altai Mountains, Gansu's Qilian Mountains, Qinghai's Zhiduo County, and Ningxia's Xiji/Guyuan Counties.