A Hybrid Multi-Stage Importance–Performance Evaluation Framework for Green Technologies in Mountain Railway Engineering

Mountain railway engineering is constrained by fragile ecology, complex terrain, and intensive construction interfaces, making the project-level selection and implementation of green technologies a critical management issue. Existing studies mainly focus on single-technology assessment or static weighting and provide limited support for linking planning-stage prioritization with implementation-stage demonstration. This study proposes a multi-stage importance–performance evaluation framework for green technologies in mountain railway engineering. The framework integrates rough set theory for factor screening, the CRITIC–grey relational method for planning-stage importance assessment, Extremely Randomized Trees for implementation-stage importance identification, and the Logical Framework Approach for deviation diagnosis and management feedback. In the example application, four of seven candidate evaluation factors were retained after rough set reduction, and the corresponding weights for environmental impact, lifecycle, technical efficiency, and compatibility were 0.247, 0.263, 0.264, and 0.238, respectively. Under the simulated construction scenario, stage-to-stage deviations ranged from −0.24 to 0.11. These results show that the proposed framework can identify key factors, compare prescribed and implementation-stage importance patterns, and support adjustment-oriented management decisions. Because the implementation-stage demonstration is based on simulated rather than real project data, this study is positioned as a methodological and conceptual demonstration rather than an empirically validated case study.