Novel control strategies for electric vehicle charging stations using stochastic modeling and queueing analysis.

This study presents a comprehensive analytical framework for modeling electric vehicle (EV) charging infrastructures through a stochastic queueing-theoretic approach that explicitly incorporates critical customer behavioral dynamics. The proposed model addresses key phenomena often overlooked in classical frameworks, including customer impatience (reneging), balking behavior, feedback mechanisms, and state-dependent service threshold policies, within a finite-population, multiple-server environment. These behavioral elements reflect realistic operational scenarios in which users may opt not to join extended queues, abandon the system due to excessive delays, or return for service completion based on prior dissatisfaction. The system dynamics are formulated using a continuous-time Markov chain (CTMC), and the corresponding Chapman-Kolmogorov differential equations are derived to characterize state transitions. Employing a matrix-analytic solution technique, the steady-state probability distribution is obtained, enabling the computation of multiple performance metrics such as system occupancy, server utilization, abandonment rates, and throughput. Numerical simulations validate the model's applicability and highlight intricate interdependencies among customer tolerance thresholds, service quality levels, and operational performance indicators. The findings offer valuable insights into capacity planning, congestion control, and service optimization, providing a rigorous decision-support framework for the design and management of EV charging networks under uncertain and dynamic user behavior. The study also outlines practical managerial implications and suggests directions for future research to enhance the adaptability and efficiency of smart charging infrastructures.