Author : Nambivel Raj and Pramod Patil

In this paper we propose an energy Management framework based on Machine Learning algorithms and IOT to provide energy savings for buildings. We specifically consider buildings with energy storage systems attached which then allow multiple use cases for energy savings. We start with one particular application of peak shaving which helps by providing savings in energy cost by reducing peak demand charges. The algorithms help forecast and calculate key variables which are then leveraged by the IOT platform to translate the insight into actions for savings. The objective is to have an end to end framework to help save by reducing monthly consumption cost. We propose end to end framework and mechanism and also evaluate available machine learning algorithms that can be used. This framework is a combination of smart edge devices and a cloud data platform used for remote monitoring by customers which provides real time data and information about their consumption patterns and trends. Also the framework created is such that it can be extended for further applications like Frequency Regulation Service and Virtual Power Plant without modifications to architecture.

Keywords: Battery Energy Storage Systems, peak shaving, smart edge device, IOT Gateway, data platform, Machine Learning, IOT, cloud platform, energy savings, Virtual Power Plant