Optimization of university dormitory renovation in severe cold regions under the impact of climate change

China has many aging buildings characterized by poor thermal insulation performance and low indoor comfort levels. Climate change can impact the performance of these buildings, making climate adaptation renovations essential. This study proposes an innovative optimization framework for renovations under four climate scenarios, integrating BP neural networks, VIKOR, NSGA-II, and NSGA-III. Taking an old university dormitory in Shenyang as a case study, the research establishes a baseline model using environmental measurements and building energy consumption inventories. Two optimization algorithms perform low-dimensional (three-objective) and high-dimensional (eight-objective) optimizations. The hierarchical optimization strategy allows NSGA-II to optimize building performance with relatively low computational resources when dealing with high-dimensional objectives. Sensitivity analysis results indicate that in severe cold regions, the importance of cooling energy consumption during summer increases as future temperatures rise. The solutions selected through hierarchical optimization can achieve energy savings of 47.9 %-54.1 % compared to the original buildings and reduce 671-762 kg of CO2 emissions per square meter during the remaining lifecycle. This framework enables the evaluation of old building performance under the influence of climate change, and its results can contribute to the research on renovation strategies for old buildings in severe cold regions.