Building energy optimisation using machine learning and metaheuristic algorithms

The focus of this research is on development of new methods for Building Optimisation Problems (BOPs) and deploying them on realistic case studies to evaluate their performance and utility. First, a new optimisation algorithm based on Ant Colony Optimisation was developed for solving simulation-based optimisation approaches. Secondly, a new surrogate-model optimisation method was developed using active learning approaches to accelerate the optimisation process. Both proposed methods demonstrated better performance than benchmark methods. Finally, a multi-objective scenario-based optimisation was introduced to address uncertainty in BOPs. Results demonstrated the capability of the proposed uncertainty methodology to find a robust design