An intelligent retrofit decision-making model for building program planning considering tacit knowledge and multiple objectives

Due to the enormous building stock and high energy consumption of the construction sector, green retrofitting of existing buildings has recently become a critical issue. China has implemented building retrofitting on a large scale based on various norms and standards. In practice, however, the effectiveness of the whole building retrofit program is often jeopardized because some decision-makers are limited by their experience and fail to evaluate the program in its entirety. To overcome this problem, this study offers an intelligent decision support model considering tacit knowledge for program decision-making with conflicting objectives. By comparing several data mining approaches and using 152 retrofitted existing buildings as examples, a tacit knowledge mining model based on the XGBoost algorithm with an accuracy of 73.91% is constructed. The predicted results of the knowledge mining model can be used as the input of the multi-objective decision-making model. In addition, the retrofit cost, thermal insulation requirement, and total retrofit area are chosen as the objectives of the multi-objective decision-making model. Then, the model's applicability to building retrofit programs is tested using five buildings as examples. Finally, the results demonstrate that the proposed model can partially replace experts in supporting policymakers and owners throughout the planning stage. •An intelligent decision-making model for retrofit program planning was proposed.•A multi-objective method embedded with machine learning algorithms were developed.•The accuracy for mining tacit knowledge using XGBoost was 73.91%.•Case study validation demonstrates the workability of the proposed model.