Stochastic optimization of energy storage systems in hybrid electric vehicle under uncertainty

For hybrid electric vehicles (HEV) to operate efficiently and robustly, the energy source sizing design must be optimized. Since, the real-world operating variables such as vehicle load demand, speed profiles, road slopes, vehicle mass, and source characteristics are unpredictable and directly affect power and energy requirements, the design method must take uncertainty into account. This research suggests an optimal hybrid energy storage size methodology under uncertainty for HEV with ultracapacitors (UC) and batteries. For the purpose of designing the battery and UC rating as efficiently as possible, multi-objective optimization based on Monte Carlo simulation (MCS) is employed. Optimization is done in each MCS loop using the non-dominated sorting genetic algorithm II. The hybrid storage power split is accomplished by using the adaptive fast Fourier transform with minimal wave reconstruction error approach. Following the completion of 20 MCS loops to ascertain the optimal HEV sizing, the optimal solution is selected based on variables such voltage stability, energy storage capacity, current handling capacity, transient support, and uncertainty coverage. MCS ensures resilience while NSGA-II efficiently figures out the pareto-optimal trade-offs among the multiple objectives meeting power requirements. As a result, the designed energy storage system is sturdy and appropriately proportioned, balancing component longevity with energy efficiency in real-world scenarios. Considering uncertainty in vehicle mass, road slope, driving and regenerative braking efficiency and battery and UC capacity, the system is robust with six series battery units, eight parallel battery units, eight series UC units, and two parallel UC units. By choosing features that represent actual operating and design conditions for the HEV, accuracy, robustness, and computational economy were guaranteed. The ideal sizing design permits 30 % mass tolerance, 5 % drive train and regenerative brake efficiency variance, sufficient road slope variation, and 5 % manufacturing and aging tolerance of sources used.