User Anxiety-aware Electric Vehicle Charging Scheduling: An Episodic Deep Reinforcement Learning Approach

• DOI: 10.20944/preprints202404.0598.v1
• Publication Date: 2024-04-10T09:22:36Z
• Authors: Ishtiaque Zaman; Miao He

The transportation industry is rapidly transitioning from Internal Combustion Engine (ICE) based vehicles to Electric Vehicles (EVs) promote clean energy. However, large-scale adoption of EVs can compromise the reliability power grids by introducing large uncertainty in demand. Demand response with a controlled charge scheduling strategy for mitigate such issues. In this paper, deep reinforcement learning- developed individual considering user’s dynamic driving behavior and charging preferences. temporal dynamics anxiety about EV battery rigorously addressed. A weight allocation technique applied continuously tune priority cost-saving respect duration. sequential control problem formulated as Markov decision process, an episodic approach deterministic policy gradient (DDPG) algorithm target smoothing delayed update techniques develop optimal strategy. real-world dataset that captures behavior, arrival time, departure duration, utilized study. extensive simulation results reveal effectiveness proposed minimizing energy cost while satisfying requirements.