A comprehensive energy efficiency evaluation index system was established based on a framework that integrated both subjective and objective factors for determining indicator weights, to address the energy efficiency evaluation challenges in highway self-consistent energy systems. First, guided by indicator selection principles and methodologies, an energy efficiency evaluation index tree for highway self-consistent energy systems was constructed, encompassing four key dimensions: energy performance, environmental sustainability, reliability, and economic feasibility. Corresponding quantitative calculation methods were defined for each indicator. Next, in determining indicator weights, expert knowledge was incorporated to refine an improved Delphi method based on CRITIC, forming a subjective evaluation approach. Simultaneously, an enhanced entropy weight method adjusted by similarity error was adopted as the objective evaluation method. The subjective and objective weights were then optimally integrated using a variance minimization approach. Subsequently, the set pair analysis (SPA) method was applied to identify the optimal solution among multiple system design alternatives. Finally, the model was implemented to evaluate six highway self-consistent energy system configurations, with the optimal solution determined as: a grid-connected system combining 800 photovoltaic panels and 10 wind turbines, which showed strong consistency with the ideal solution. The proposed evaluation method effectively leveraged indicator data, comprehensively considered subjective and objective factors, and enhanced the scientific rigor and practical applicability of the results. It provides a robust scientific foundation for the assessment and optimization of highway self-consistent energy systems.