Machine Learning Assisted Energy Optimization in Smart Grid for Smart City Applications

Abstract

Peer-to-peer electricity transaction is predicted to play a substantial role in research into future power infrastructures as energy consumption in intelligent microgrids increases. However, the on-demand usage of Energy is a major issue for families to obtain the best cost. This article provides a machine learning predictive power trading framework for supporting distributed power resources in real-time, day-to-day monitoring, and generating schedules. Furthermore, the energy optimization algorithm used in machine learning (EOA-ML) is proposed in this article. The machine learning-based platform suggested two modules: fuel trading and intelligent contracts based on machine learning implemented predictive analytical components. The Blockchain module enables peers to track energy use in real-time, manage electricity trading, model rewards, and irreversible transaction records of electricity trading. A predictive analysis component based on previous power usage data is designed to anticipate short-term energy usage in the Intelligent Contracts. This study utilizes data from the provincial Jeju, Korea’s electricity department on true energy utilization. This study seeks to establish optimal electricity flow and crowdsourced, promoting electricity between consumers and prosumers. Power trading relies on day-to-day, practical environmental control and the planning of decentralized power capitals to satisfy the demands of smart grids. Furthermore, it employs data mining technologies to obtain and study time-series research from the past electricity utilization data. Thus, the time series analytics promotes power controllingto better future efficient planning and managingelectricity supplies. It utilized numerous statistical methods to assess the effectiveness of the suggested prediction model, mean square error in different models of machine learning, recurring neural networks. The efficacy of the proposed system regarding the delay, throughput, and resource using hyperleader caliper is shown. Finally, the suggested approach is successfully applied for power crowdsourcing between prosumer and customer to reach service reliability based on trial findings. The actual and predicted cost analysis has been increased (95%). It minimizes the delay rate to (40.3%) by improving the efficiency rate.

Efficient Usage of Energy Infrastructure in Smart City Using Machine Learning

“…As a result of the increased efficiency, the delay rate is reduced to 40.3%. [3] Electric vehicles (EVs) are rapidly gaining popularity as a fundamental component of smart mobility in smart city applications due to their ability to help lower greenhouse gas emissions. However, the strain on power grid infrastructure from widespread EV deployment is one of the biggest obstacles.…”

Classification analysis of burnout people's brain images using ontology‐based speculative sense model

“…The time complexity for the second level in the decision tree is less than the one needed to build the decision tree. The child nodes are created based on satisfying the condition, and the burnout brain images are classified based on ranges 21 …”

“…The child nodes are created based on satisfying the condition, and the burnout brain images are classified based on ranges. 21…”

Data-Driven Methods of Machine Learning in modeling the Smart Grids

“…The most common ML algorithms used in the SG analysis are artificial neural networks (ANN), Gaussian Regression (GR), Support Vector Machine (SVM), Random Forest (RF), Linear Regression (LR), and K-nearest neighbor (KNN) [1]. An ML-based energy optimization algorithm enables to tracking of realtime energy use, irreversible transaction records of electricity trading, managing electricity trading, and model reward [2]. The Gray-Wolf algorithm helps to manage the optimum programming of agents, loads, storage, and switches in the SG [3].…”