Nowadays, as the role of energy retrofit on the existing building stock is recognized towards energy savings and emissions’ reductions, the actions to be undertaken towards this aim require complex decisions, in terms of the choice among active and passive strategies and among often conflicting objectives of the retrofit. Depending on the actor of the retrofit (e.g., private, public), the main objective could be minimizing the investment, minimizing the energy demand or cost, or minimizing emissions. To facilitate the selection of the optimal retrofit actions, here the application of active archive non-dominated sorting genetic algorithm (aNSGA-II) towards multi-objective optimization is illustrated. The results of the algorithm implementation are analyzed with respect to a residential building located in Rome, Italy. The genes (i.e., the implemented strategies) are described and the optimal solution in the R⁴ space is discussed, alongside with considerations about the solutions pertaining to the Pareto frontier. The applied method allowed to considerably lower the computational time and identifying the multi-objective optimal solution, which was able to reduce by 49.2% annual energy demand, by 48.8% annual energy costs, by 45.2% CO₂ emissions while still maintaining almost 60% lower investment cost with respect to other criterion-optimal solutions.

如今，随着人们认识到能源改造对现有建筑存量在节能和减排方面的作用，在主动和被动策略以及经常相互冲突的目标之间进行选择时，为实现这一目标而需要采取的行动需要复杂的决策的改造。根据改造的参与者（例如，私人，公共），主要目标可能是最小化投资，最小化能源需求或成本，或最小化排放。为了便于选择最佳的改进措施，在此说明了主动归档非支配排序遗传算法（aNSGA-II）在多目标优化中的应用。针对位于意大利罗马的一栋住宅楼，分析了算法实现的结果。基因（即讨论了⁴空间，并考虑了与帕累托前沿有关的解决方案。应用的方法可以显着减少计算时间并确定多目标最优解决方案，该解决方案能够将年能源需求减少49.2％，将年能源成本减少48.8％，将CO ₂排放量减少45.2％，同时仍保持近60％的排放量与其他标准最佳解决方案相比，投资成本更低。