A Multiobjective Approach to Find Optimal Electric-Vehicle Acceleration: Simultaneous Minimization of Acceleration Duration and Energy Consumption

The high energy consumption associated with acceleration requires electric vehicles (EVs) to accelerate to a chosen speed optimally, particularly in urban driving cycles. Existing methods deal with the minimization of acceleration energy without considering the acceleration duration. This paper focuses on solving a multiobjective optimization problem (MOOP) with two conflicting objectives: minimization of acceleration duration and minimization of energy consumption. Two approaches were used to reach a desired speed: using a single acceleration value and using multiple acceleration values. For each approach, demonstrative speed changes were chosen, and the problem was solved using multiobjective genetic algorithms (MOGAs). The results (Pareto-optimal fronts) obtained by these two approaches were compared using suitable performance metrics. To validate the reliability of MOGA results, statistical analysis was carried out. Furthermore, a nonparametric study, i.e., the Wilcoxon signed-rank test, was conducted to compare the effectiveness of both approaches. It was found that multiple accelerations were more effective in minimizing the duration and energy consumption than a single acceleration. For the same duration, multiple accelerations reduced energy consumption by up to 2%. Sensitivity analysis for both approaches with electric motor model parameters was conducted. The simulation results of EV acceleration using the preferred optimal solution based on driving comfort and the Pareto front's knee suggested a strong implication toward driving assistance.