BUILDING LIFE CYCLE MANAGEMENT AT THE OPERATION STAGE USING ARTIFICIAL NEURAL NETWORK MODELS AND MACHINE LEARNING

Abstract

The use of artificial neural networks and machine learning methods for the analysis of heat loss in buildings is of significant relevance in modern construction. These technologies are highly accurate and efficient in data processing. Artificial neural networks have the ability to analyze vast amounts of information and identify complex patterns, which significantly increases the accuracy of determining heat loss in buildings. In turn, machine learning methods make it possible to take into account various influencing factors, such as geographic location and meteorological conditions, thereby making a significant contribution to improving the quality of analytical results. Such approaches provide more reliable and accurate conclusions, which is critical for effective energy management and reducing heat loss in buildings. In this article, the authors conducted a study of heat losses of buildings and their prediction at the operational stage using artificial neural networks and machine learning methods. The technique is based on the analysis of data on heat loss and their relationship with various building parameters. Forecasting was carried out using artificial neural networks in the Statistica software package and the machine learning method based on the scikit-learn library. The proposed approach allows you to effectively manage the energy consumption of a building, optimizing its energy efficiency and improving the life cycle management of a capital construction project. The results demonstrate the high accuracy and convergence of the model with actual values, as well as its ability to predict performance.