Machine learning (ML) has been recognised as a powerful method for modelling building energy consumption. The capability of ML to provide a fast and accurate prediction of energy loads makes it an ideal tool for decisionmaking tasks related to sustainable design and retrofit planning. However, the accuracy of these ML models is dependent on the selection of the right hyper-parameters for a specific building dataset. This paper proposes a method for optimising ML models for forecasting both heating and cooling loads. The technique employs multi-objective optimisation with evolutionary algorithms to search the space of possible parameters. The proposed approach not only tunes single model to precisely predict building energy loads but also accelerates the process of model optimisation. The study utilises simulated building energy data generated in EnergyPlus to validate the proposed method, and compares the outcomes with the regular ML tuning procedure (i.e. grid search). The optimised model provides a reliable tool for building designers and engineers to explore a large space of the available building materials and technologies.
The use of Building Information Modelling (BIM) has gained considerable interest in new build projects. However, its use in existing assets has been limited to geometric models utilising Point Cloud Data (PCD) as the primary source of data. The inclusion of non-geometrical data from distributed sources in the geometric model to make it semantically rich has been fraught with considerable challenges. In this paper, an approach is proposed to provide a framework for generating semantically-rich parametric models for existing assets. While the geometric information like length, width, area, and volume can be extracted from a PCD, non-geometric data may need to be appended to this for generating genuinely semantically rich models. The Comma Separated Values (CSV) format is utilised to represent the data that can be extracted from PCDs. In addition, the non-geometric information derived from other sources is appended to the CSV file. Subsequently, the Resource Description Framework (RDF) data is generated from the data presented in the CSV files. RDF is a commonly used Semantic Web technology for storing, sharing, and reusing information on the Web. The RDF data is then used to create the IFC data model by translating RDF into IFC. The IFC file is used to generate 3D BIM by importing it into any IFC-compliant application. The proposed approach was validated on one part of the Edinburgh castle, a relatively complex historical building. The choice of building for validating the approach was driven by technical as well as pragmatic reasons. Technically, the approach will have proven its robustness if it could be shown to work for a complex rather than a relatively simple building. Pragmatically, the authors had access to data on Edinburgh Castle due to an ongoing partnership with the Historic Environment Scotland (HES). However, as a result of the validation process, it is suggested that the proposed approach should be applicable to any existing building.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.