Vizualizace monitorování provozně technických funkcí v Inteligentní budově v rámci IoT

Abstract

The aim of this dissertation is to ensure the calculation of the prediction of carbon dioxide concentration from PI System data. The first step is to explore the possibilities that IBM Cloud and Microsoft Azure offer in data analysis and machine learning. The knowledge gained when working with cloud tools is written up in a separate chapter. In the next part of the paper, experiments are developed. Their results lead to the selection of a suitable prediction method and its setting for the needs of this work. Connectivity for reading and writing data from the PI Server is created via the PI Web API Code that is used both for the long-term calculation of CO2 prediction and for communication with PI Web API is written in Python. Furthermore, the work with the PI Integrator for the Business Analytics tool and the connection to Apache Kafka are described within real-time data streaming. Finally, a visualization of the operational and technical functions of the two bytes in the PI Vision tool is created. The visualization also contains data of predicted CO2 values.