Home /Research /Electric vehicle integration system for power management on a blockchain-based smart grid platform with AI capabilities
LEARNING

Electric vehicle integration system for power management on a blockchain-based smart grid platform with AI capabilities

Mohd. Afzal Khan, Nidhi Gupta, Rahit Kathuria, Sharma Vanshika, Vivek Sharma, Anil Kumar Vaghmare

Year
2023
Citations
3
Access
Open access

Abstract

<title>Abstract</title> A virtual power plant (VPP) is composed of a network of dispersed power generating units, available power users, and storage systems. The AEBIS system is an AI-enabled, block chain-based electric car integration system that is presented in this research for power management in a smart grid platform. The use of AI based chips guarantees a performance that is economical. A safe and transparent service is further achieved with the addition of block chain technology to the system at an affordable memory and latency cost. A VPP is made up of a network of dispersed power generating units, users, and storage systems. During times of peak load, a VPP distributes the electricity generated by numerous connected units to balance the network's load. When used effectively, demand-side energy equipment like mobile robots and electric vehicles (EVs) may also balance the supply and demand of energy. The balancing of supply power is an ambitious goal due to the volatility in power produced by various power units. Furthermore, although it is crucial, the security of communications and the end installations has not received adequate attention. This article suggests an electrical solution that is blocking chain-based and AI-powered. The vehicle integration system for the energy management of the smart grid platform is known as AEBIS. The solution is based on artificial neural networks and federated learning algorithms for EV load prediction and utilizes the fleet of electric vehicles as both a consumer and a supplier of electric energy within a VPP platform. The evaluation findings demonstrate that the suggested approach produced high power consumption forecasts in the standard training scenario with an R2 score of 0.938. Applying a federated learning strategy resulted in a 1.7% accuracy reduction. The suggested system produces a dependable and prompt service to deliver additional electricity from the vehicular network as a result, consequently lowering the amount of power fluctuation and providing an accurate estimation of power consumption. Additionally, adopting an AI chip guarantees economical performance. A safe and transparent service may also be obtained at the level at the cost of an acceptable memory cost and latency thanks to the system's integration of block chain technology.

Keywords

Electric power systemSmart gridComputer scienceEmbedded systemEnergy storageElectric powerEnergy managementElectricityEnergy management systemAutomotive engineering

Related papers

Browse all LEARNING papers