In China, there are many old buildings whose renovation could enhance building environment quality, reduce energy consumption, lower carbon emissions, and create economic benefits. However, there is currently a lack of efficient architectural renovation optimization methods to provide decision-makers and designers with highly feasible renovation plans. This paper proposes an innovative optimization method for renovating old buildings to address this issue. The research employs a hierarchical optimization framework integrating Artificial Neural Networks (ANN), the NSGA-II algorithm, and the entropy-weighted TOPSIS method. The first layer's three-objective optimization using the trained neural network model significantly improves the algorithm's efficiency in searching for optimal solutions. The second layer's entropy-weighted TOPSIS optimization includes eight objective optimizations, such as energy, economy, and lighting, thus broadening the assessment dimensions. Taking an old multi-story university dormitory in the Shenyang area as an example, the study optimizes nine passive design variables across eight objectives. The hierarchical optimization results show that the high-scoring renovation plans can save 1.52 to 3.17 million yuan in energy costs over the building's remaining lifecycle and reduce carbon dioxide emissions by 644–962 kg per square meter. SRRC results indicate that external windows significantly impact economic efficiency and energy consumption, emphasizing the importance of selecting the right type of windows in actual engineering projects. The new method proposed in this study can offers a comprehensive and highly feasible renovation plan for old buildings.