Using algorithms to designate pre-fabricated wall materials: A case study with two implementation methods

Designating wall materials manually is neither efficient nor competent to find the ideal solution. Therefore, the current study aims to develop a new designation approach based on multi-objective optimization which can automatically explore a larger space of solutions and obtain the optimal material combination. The core concept is to use wall properties as objectives to search for satisfying combinations, rather than the thresholds to validate manually proposed solutions. A case study is carried out to find material combinations with favourable U value, specific heat capacity, and the total material price. It is realised by two different research methods: using the k-nearest neighbours (k-NN) algorithm in the Python environment, and using the Strength Pareto Evolutionary Algorithm 2 (SPEA2) in the Grasshopper. Finally, the usability test was conducted among 15 architects. Test results confirmed that this new approach could save much time and find solutions with better performance. The present study has significance in reducing architects’ repetitive work while speeding up the decision-making process.