Search

The advent of advanced metering infrastructure has led to vast quantities of energy data becoming available. Despite this, the typical methods employed for the performance verification of energy efficiency improvements have not progressed, as they continue to rely on expert knowledge and simplistic statistical modelling techniques. This leads to uncertainty in the quantity of savings arrived at, with this uncertainty acting as a barrier to investment in energy efficiency. In response to this, the industry is evolving towards more advanced and automated methods known as M&V 2.0. This however presents the challenge of keeping the resources required to perform M&V at a minimum level, while also improving the accuracy, reliability and trust in the process. The research presented in this thesis can be largely classified into two prominent tasks. These are the development of a machine learning-based methodology for the construction of accurate baseline energy models and the establishment of a framework and final solution for M&V 2.0. It will be shown through theoretical and practical work that: - Machine learning techniques reduce the uncertainty introduced into the performance verification process by the baseline energy regression model. Additionally, the utilisation of a broader scope of analysis with respect to traditional methods is advantageous in further improving model accuracy. - Novel, computationally efficient data processing methods, including cleaning and feature selection, can be tailored for the industrial buildings sector to minimise the resources required to carry out performance verification. - The void in knowledge resulting from the established M&V protocols can be populated by a prescriptive methodology that utilises machine learning techniques to accurately and reliably quantify energy savings; thus, empowering performance verification practitioners in the use of advanced analytics on granular data sets and removing the need for expert knowledge. - The industrial sector requires a specific framework for the application of M&V 2.0 practices. An M&V 2.0 framework is developed to offer a solution to the challenge of persisting energy savings. A performance deviation detection system enables integration with ongoing monitoring and targeting practices. - M&V 2.0 does not have to increase the resources required to carry out performance verification. A novel, cloud computing-based solution, IntelliMaV, is capable of quantifying energy savings in near real-time with minimal uncertainty at high confidence levels. This thesis addresses some of the key challenges facing the performance verification industry including utilising the large quantities of energy data available in industrial facilities and evolving practices to a level of maturity that will enable it to support M&V 2.0. The implications of such challenges are shown to be significant beyond the individual project level, with the effectiveness of European energy policy dependent on accurate and reliable M&V. The methodology, framework and IntelliMaV application developed all address these challenges, while aiding the transition to M&V 2.0 practices. Despite these advancements, this is not the final solution for the industry. A collective effort must be made to continue to modernise performance verification practices to ensure M&V remains a valued and beneficial practice in energy management into the future.

Automatic text summarization, a process of distilling the most important information from a text document, is to create an abridged summary with software. Basically, in this task, we can regard the "summarization" as a function which takes a single document or multiple documents as an input and has the summary as an output. There are two ways that we can manage to create a summary: extractive and abstractive. The extractive summarization means that we select the most relevant sentences from the input and concatenate them to form a summary. Graph-based algorithm like TextRank, Feature-based models like TextTeaser, Topic-based models like Latent Semantic Analysis (LSA), and Grammar-based models could be viewed as approaches to extractive summarization. Abstractive summarization aims to create a summary similar to humans. It keeps the original intent, but uses new phrases and words not found in the original text. One of the most commonly used models is the encoder-decoder model, a neural network model that is mainly used in machine translation tasks. Recently, there is another combination approach that combines both extractive and abstractive summarization, like Pointer-Generator Network, and the Extract then Abstract model. In this course, we're given both a small dataset (about 500 documents) and a big dataset (about 11,300 documents) that mainly consist of web archives about a specific event. Our group is focusing on reports about a terrorist event -- Attack Westminster. It occurred outside the Palace of Westminster in London on March 22, 2017. The attacker, 52 year-old Briton Khalid Masood, drove a car into pedestrians on the pavement, injuring more than 50 people, 5 of them fatally. The attack was treated as "Islamist-related terrorism". We first created a Solr index for both the small dataset and the big dataset, which helped us to perform various queries to know more about the data. Additionally, the index aided another team to create a gold standard summary of our dataset for us. Then we gradually delved into different concepts and topics about text summarization, as well as natural language processing. Specifically, we managed to utilize the NLTK library and the spaCy package to create a set of most frequent important words, WordNet synsets that cover the word, words constrained by part of speech (POS), and frequent and important named entities. We also applied the LSA model to retrieve the most important topics. By clustering the dataset with k-means clustering, and selecting important sentences from the clusters using an implementation of the TextRank algorithm, we were able to generate a multi-paragraph summary. With the help of named entity recognition and pattern-based matching, we confidently extracted information like the name of the attacker, date, location, nearby landmarks, the number killed, the number injured, and the type of the attack. We then drafted a template of a readable summary to fill in the slots and values. Each of these results individually formed a summary that captures the most important information of the Westminster Attack. The most successful results were obtained using the extractive summarization method (k-means clustering and TextRank), the slot-value method (named entity recognition and pattern-based matching), and the abstractive summarization method (deep learning). We evaluated each of the summaries obtained using a combination of ROUGE metrics as well as named-entity coverage compared to the gold standard summary created by team 3. Overall, the best summary was obtained using the extractive summarization method, with both ROUGE metrics and named-entity coverage outperforming other methods. NSF: IIS-1619028 AttackWestMinster_Team4_Project_Report.pdf: Final project report. AttackWestMinster_Team4_Project_Report_Overleaf.zip: Final project report archive from LaTeX project in Overleaf. AttackWestMinster_Team4_Final_Presentation.pdf: Final presentation in PDF format. AttackWestMinster_Team4_Final_Presentation.pptx: Final presentation in PowerPoint format. AttackWestminster_Team4_Code.zip: All source code related to our project.