Distributed Discrete Level Energy Scheduling for Residential Load Control in Smart Grids
The digitisation of the world over the last decade sees the increase in the mass adoption of connected digital technologies and applications. Electronics and appliances have shifted from a luxury to a daily necessity. The shift to this Energy-based living standards resulted in an evolving energy landscape where utilities must adapt on how they interact with the electricity consumption habit of residential users or customers for an optimised energy or electricity scheduling and pricing
Dr. Chai Chin Choy of A*STAR Institute for Infocomm Research (I2R)’s Smart Energy & Environment Cluster (SEEC), explored a general mathematical framework that models and solves the problem of optimizing energy or electricity scheduling for such scenario from the utilities’ perspective.
In this project, Dr. Chai equipped each residence with an energy management system (EMS) with sufficient computational power and a smart meter that communicates with its various electrical devices. Two-way communication is then established with the EMS central controller at the energy provider through the appropriate communication network protocols for example, a local area network.
There were three considerations and approaches taken while on this project:
Firstly, considerations were given to the discrete level energy scheduling with variable starting time and termination time, taking into account variable starting and ending times of elastic loads (dependant on the users’ need).
Secondly, to resolve the mixed integer programming problem from earlier, closed form solutions that facilitate distributed implementation were derived. These solutions maximize the total utility from the users’ perspective, as well as maximize the total revenue from the energy provider’s perspective, a desirable objective which has been neglected in previous works.
Lastly, the team proposed a distributed load selection and discrete level energy scheduling scheme that achieves distribution of computation complexity to users, while each user can still converge to its local sub-optimal solutions. The proposed scheme requires only aggregate load and pricing information exchanges between the energy provider and each user.
Different from many previous works, this is a novel approach to reduce computational load at the utilities by exploiting smart metering with two-way communications between the utilities and each residential users. While maximizing the welfare of residential users in terms of their utilities, the proposed solution also maximizes the total revenue or sale at the utilities through pricing control, a highly desirable objective which has not been emphasized in many previous works.
* This paper clinched the Best Paper Award in IEEE Region Ten Conference (TENCON) 2017.