Intelligent Forecasting And Control Of Energy Storage For Smart Buildings

Abstract

A reliable and cost-effective electricity supply is pivotal for the advancement of society, sustainable economy, and wellbeing of its citizens. INTERACT project aims to deliver an intelligent and cost-optimal solution for energy management for the building equipped with PV and energy storage. This project shall consist of a forecasting module and a controller. The forecasting module shall predict the building load intelligently and dynamically. Similarly, the PV power forecasting module shall also have a self-correcting mechanism to deal with the modeling issues. The third variable that adds novelty to the project is the estimation of the power available from the grid. These forecast variables shall drive the optimal dispatch of the energy storage to meet the building load in the most cost economical way and also according to the user preferences. The first part of the controller development shall be to perform mathematical modeling of the problem. Subsequently, the problem shall be convexified to find a deterministic solution to the optimization problem. The forecast module along with the controller model shall be deployed on a prototype in the laboratory environment. The industrial partner shall be responsible for communicating the field requirements, deploying the controller in the field, and debugging practical issues. The innovative approach shall provide flexibility to the building operator in reducing its energy costs and also to ensure the availability of power during the defined intervals. INTERACT shall lead to a comprehensive and adaptable solution to the energy problem in Pakistan.

Primary Goals

The goal of this project is to develop and test a prototype energy management system for buildings that minimize the cost of energy and ensure the availability of the energy supply. It is assumed that a PV system and storage have been installed at the building.

Infrastructure Needed: Access to GPU cluster for developing the forecasting model and workstation

Principal Investigator

Co-Principal Investigators

Mr. Imad Asif

imadasif@ikspvt.com

Industrial Collaborator(s)

IKS Services Pvt Limited

Scientific Field(s)

AI, Forecasting, Energy Informatics, Renewable Energy, Power Systems

Infrastructure Needed

Access to GPU cluster for developing the forecasting model and workstation

Duration

12 months

Funding

5.979 Million PKR