main-content

This book outlines the data-driven modelling of building energy performance to support retrofit decision-making. It explains how to determine the appropriate machine learning (ML) model, explores the selection and expansion of a reasonable dataset and discusses the extraction of relevant features and maximisation of model accuracy.

This book develops a framework for the quick selection of a ML model based on the data and application. It also proposes a method for optimising ML models for forecasting buildings energy loads by employing multi-objective optimisation with evolutionary algorithms. The book then develops an energy performance prediction model for non-domestic buildings using ML techniques, as well as utilising a case study to lay out the process of model development. Finally, the book outlines a framework to choose suitable artificial intelligence methods for modelling building energy performances.

This book is of use to both academics and practising energy engineers, as it provides theoretical and practical advice relating to data-driven modelling for energy retrofitting of non-domestic buildings.

### Chapter 1. Introduction

Abstract
In the UK, buildings are responsible for 46% of all carbon dioxide ($$CO_2$$) emissions [20]. This figure is 40% in the USA and 27% in Australia [15]. Accordingly, the enhancement of energy efficiency of buildings has become an essential matter in order to reduce the amount of gas emission as well as fossil fuel consumption. An annual saving of 60 billion Euro is estimated as a result of the improvement of EU buildings energy performance by 20% [21].

### Chapter 2. Building Energy Performance Assessment Methods

Abstract
Buildings are responsible for a vast amount of GHG emission. Therefore, most countries have set regulations to decrease the gas emission and energy consumption of buildings. These regulations are diverse targeting different areas, new and existing buildings and usage types. This paper reviews the methods employed for building energy performance assessment and summarise the schemes introduced by governments. The challenges with current participates are discussed and solutions will be recommended.

### Chapter 3. Multi-objective Optimisation and Building Retrofit Planning

Abstract
This chapter, first, reviews evaluation indices for the efficient retrofit plan to enhance building energy performance, second, provides the concept and mathematical demonstration of multi-objective optimisation (MOO) and finally presents the potential of using MOO for supporting the development of retrofitting strategies.

### Chapter 4. Machine Learning for Building Energy Forecasting

Abstract
In recent years, Artificial Intelligence (AI) in general and Machine Learning (ML) techniques in specific terms have been proposed for forecasting of building energy consumption and performance. This chapter provides a substantial review on the four main ML approaches including artificial neural network, support vector machine, Gaussian-based regressions and clustering, which have commonly been applied in forecasting and improving building energy performance.

### Chapter 5. Machine Learning Models for Prediction of Building Energy Performance

Abstract
This chapter investigates the accuracy of most popular ML models in the prediction of building heating and cooling loads carrying out specific tuning for each ML model and using two simulated building energy data. The use of grid search coupled with cross-validation method in examination of the model parameters is demonstrated. Furthermore, sensitivity analysis techniques are used to evaluate the importance of input variables on the performance of ML models. The accuracy and time complexity of models in predicting heating and cooling loads are demonstrated.

### Chapter 6. Building Energy Data-Driven Model Improved by Multi-objective Optimisation

Abstract
This chapter 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 chapter utilises simulated building energy data 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.

### Chapter 7. Modelling Energy Performance of Non-domestic Buildings

Abstract
This chapter presents an energy performance prediction model for the UK non-domestic buildings supported by machine learning. The aim of the model is to provide a rapid energy performance estimation engine for assisting multi-objective optimisation of non-domestic building energy retrofit planning. The study lays out the process of model development from the investigation of requirements and feature extraction to the application on a case study. It employs sensitivity analysis methods to evaluate the effectiveness of the feature set in covering retrofit technologies. The machine learning model which is optimised using advanced evolutionary algorithms provides a robust and reliable tool for building analysts enabling them to meaningfully explore the expanding solution space.