Optimization and Analysis of Electric Vehicle Charging Systems using PMLP-PSVM-MLR Approach

Electrifying transport represents a highly promising alternative to reduce CO2 emissions from the transport industry and prevent possibly catastrophic climate change. Electric vehicles have the potential to transform transportation; yet, their extensive adoption presents issues for the efficient operation of electric power networks, especially at the distribution level. An effectively administered Electric Vehicle Charging System is crucial for maintaining network stability. This paper examines the potential impacts of a comprehensive Electric Vehicle Charging System on a distributed network, proposes control measures to alleviate those impacts, and investigates the benefits of such a system from a performance perspective. The methodology consists of three phases: data preparation, feature extraction, and model training. Data preprocessing encompasses transformation, normalisation, and partitioning, whilst feature extraction is executed via the Hough transform, which converts data from a rectangular grid to a polar coordinate system. Models are subsequently trained utilising PMLP, PSVM, and MLR methodologies. The suggested solution surpasses conventional alternatives, PSVM and MLP, improving network efficiency and stability. The results demonstrate that optimised control systems can markedly mitigate the adverse effects of electric vehicle charging on distribution networks. Subsequent study ought to concentrate on enhancing these tactics and broadening their application to more extensive networks.