Machine learning and predictive control-based energy management system for smart buildings

The concepts of smart cities are introduced to sustainably accommodate the large and growing population. In this regard, several constituents are identified, viz., smart buildings, smart industries, smart waste management, smart transportation, etc. A city comprises of several buildings like residential apartments, commercial complexes, corporate offices, etc., where people reside and spend most of their time. Now, a building can have various areas where it can be improved by incorporating different technologies. The energy production via renewables like solar photovoltaics, mini wind turbines, biomass, energy storages, diesel generator, integration to the smart grid, managing various loads. This book chapter provides a brief state of the art literature on smart buildings as important “building blocks” for realizing the vision of smart cities. It explains various components of the building instrumentation and automation, challenges, and future research trends. A discussion on smart homes technology having interconnected appliances with internet of things, information and communication technologies, and decision making as well as control algorithm design using the field programmable gate array is also included. Lastly, an application of machine learning in forecasting various time-varying parameters from both load and generation sides, which is to be used in predictive control is presented.