Queueing modeling and optimization of hybrid electric vehicle infrastructures using evolutionary algorithms

• Stochastic modeling of HEV charging networks. • Demonstration of mathematical differential equations and its solution methodology. • Optimal and sensitivity investigations based on total anticipated cost. • Implementation of the nature-inspired algorithm: particle swarm optimization. The charging infrastructure for hybrid electric vehicles (HEVs) is essential for sustainable transportation, however, such systems experience challenges due to inconsistent operating patterns. The current study proposes a queueing-based stochastic modeling incorporating working vacations, vacation interruptions, and arrival control strategies to enhance service operations. The set of differential-difference equations is constructed to characterize system dynamics, subsequently leading to the establishment of essential system performance indicators for performance assessment. The steady-state probability distribution is derived by employing the matrix-analytical method. The novelty of the proposed research is to formulate the cost optimization problem and demonstrate optimal converging outcomes through an in-depth comparative investigation of multiple evolutionary algorithms, including the particle swarm optimization (PSO), cuckoo search (CS), grey wolf (GW), and honey badger (HB) optimizers. Further, economic and sensitivity analyses are provided to highlight practical insights for the design and operation of HEV charging infrastructures. The findings of numerical illustrations and optimal investigation are presented in tables and graphs to provide straightforward perspectives. Lastly, the concluding remarks and future perspectives are provided covering the significant contributions of the research findings.