Methodological framework for efficient energy management by intelligent data analytics

Link to CroRIS site

Although Croatia and other EU countries have strategies and national energy efficiency plans that set and monitor targets for reducing energy consumption, the data is not sufficiently scientifically analyzed to enable more efficient energy management and, in particular, machine learning methods are lacking. The aim of the project is to carry out an intelligent analysis of data on the energy efficiency of buildings in the public sector in order to offer methods and models that enable better planning of national energy policy and cost management. The project objectives are: (1) to identify characteristic groups (clusters) of public buildings for which energy characteristics, measures to improve these characteristics and energy consumption will be monitored, (2) to develop a methodological framework for predicting and analyzing the impact of individual variables on energy consumption savings, CO2 emissions and costs through data analysis, (3) to apply advanced analytical methods to assess the financial viability and long-term financial impact of energy efficiency measures, and (4) to propose adjustments to the model of the natural gas supply chain as one of the main energy sources in the public sector through simulation modeling. Machine learning methods such as cluster analysis, neural networks, decision trees, support vector methods and others as well as financial methods, supply chain mapping and simulation modeling are tested to build the model. Through statistical testing, methods that produce more accurate models will be selected and their integration into a new methodological framework will be proposed as a scientific contribution to the project. In addition, the impact of individual input variables on energy consumption and costs will be revealed and a unique energy efficiency database will be created, which will also serve for future analysis. In addition to researchers, the beneficiaries of this project may include government institutions providing subsidies, energy policy makers and implementers, building users and energy service providers.