Browsing by Author "Sunny, M.R.,"

Now showing 1 - 4 of 4

An approach to imbalance power management using demand response in electricity balancing market
(Institute of Electrical and Electronics Engineers Inc., 2019-09) Kabir, M.A.,; Sunny, M.R.,; Zhang, C.
Imbalance power management is a key operational task for a system operator. Imbalance power is produced from the difference in electrical energy supply and demand in the real operational time which deviates power system stability. In this paper, a balancing market model using demand response is developed to mitigate imbalance power. In context, a case study of balancing market model is investigated using flexible residential load along with nord pool spot market and imbalance data Therefore, the market model is simulated in MATLAB for five weeks. The simulation result is evaluated in order to determine imbalance reduction value, flexibility benefits and demand response value. Finally, the study is concluded with the propositions of future balancing market model contribution in imbalance reduction using demand response.
Assessment of Grid-connected Residential PV-Battery Systems in Sweden-A Techno-economic Perspective
(Institute of Electrical and Electronics Engineers Inc., 2021) Kabir, A.,; Sunny, M.R.,; Siddique, N.I.
Ensuring a secure supply of electricity without harming climate is a key challenge for future power system and many renewables-based cutting-edge technologies are introduced to overcome this challenge. This paper aims to study the grid-connected residential PV-battery system at behind-The-meter scenarios in Sweden from a technical and economic perspective. The system is designed with PV arrays, inverters, Lithium-ion or lead-Acid batteries. The optimal PV, lithium-ion and lead-Acid battery size are determined at two locations (Arlanda and Karlstad) in Sweden based on the highest value of Net Present Value (NPV), Profitability Index (PI), the lowest value of Levelized cost of energy (LCOE) and payback period considering system losses, electricity market price, cost and PV incentives and a comparison is performed between these two locations against some techno-economic performance metrics for two combinations of PV-battery systems. Then, the best-case energy profile, bill savings, battery performance are also investigated. Finally, the system is simulated using System Advisory Model (SAM), a renewable analysis software from NREL, USA, and is found that the grid-connected PV with lithium-ion battery system is feasible and more economical considering available PV incentives in Sweden.
Residential Energy Management: A Machine Learning Perspective
(IEEE Computer Society, 2020) Sunny, M.R.,; Kabir, M.A.,; Naheen, I.T.,; Ahad, M.T.
In smart grids, residential energy management is a vital part of demand-side management. It plays a pivotal role in improving the efficiency and sustainability of the power system. However, challenges such as variability of consumption profiles require machine learning to understand and forecast residential demands. Moreover, machine learning based intelligent load management is required for effective implementation of demand response programs. In this article, applications of machine learning algorithms in residential demand forecasting, load profiling, consumer characterization, and load management are comprehensively discussed. The article also examines the characteristics and availability of relevant databases, and explores research challenges and possibilities.
Smart Meter Data Compression and Load Profile Classification using UMAP and Random Forest
(Institute of Electrical and Electronics Engineers Inc., 2021) Sunny, M.R.,; Kabir, M.A.,; Islam, R.,; Nazifa, S.
In this paper, Uniform Manifold Approximation and Projection (UMAP) is used to compress electricity consumption data. The Random Forest (RF) classification algorithm is then used on the compressed data to learn the consumption patterns of two distinctive user base - household consumers and SMEs (small and medium businesses). Compression ratio achieved by UMAP and classification accuracy of our classifier model are compared with conventional methods and various machine learning pipelines proposed in recent studies. The results demonstrate that our proposed technique achieves better compression ratio and classification accuracy compared to the conventional methods.