With the rapid adoption of electric vehicles globally, integrating large-scale electric vehicle loads into park-level integrated energy systems has become a critical approach to enhancing energy utilization efficiency and reducing grid stress. In China, the electric vehicle market is expanding rapidly, making it essential to develop strategies that accommodate these loads while promoting low-carbon operations. This paper proposes a two-layer optimization framework that combines electric vehicle charging management with advanced power-to-hydrogen technologies to achieve economic and environmental benefits. The model considers real-time pricing, improved power-to-gas processes, and a carbon trading mechanism to minimize costs and emissions.
The park integrated energy system comprises multiple energy sources, including wind turbines, photovoltaic systems, combined heat and power units, gas boilers, and energy storage for electricity, heat, gas, and hydrogen. A key innovation is the enhanced power-to-gas system, which involves electrolyzers, methane reactors, and hydrogen fuel cells to improve hydrogen utilization. The integration of electric vehicles is modeled based on spatiotemporal characteristics to simulate disordered charging, which is then optimized through real-time electricity prices to achieve orderly charging. The system participates in a carbon trading market with a stepped carbon pricing mechanism to further reduce emissions.
Electric vehicle charging behavior is analyzed using data-driven approaches, considering factors like travel purpose, time, and location. The disordered charging load is derived from statistical distributions of departure times, trip durations, and parking times. For instance, the probability density function for the first departure time follows a Gaussian mixture model: $$ f(t_f) = \sum_{i=1}^n \alpha_i N(\mu_i, \sigma_i) $$ where \( t_f \) is the first departure time, \( \alpha_i \) are weights, and \( \mu_i \), \( \sigma_i \) are parameters. Similarly, driving time follows a log-normal distribution: $$ f(t_d) = \frac{1}{t_d \sigma_d \sqrt{2\pi}} \exp\left[ -\frac{(\ln t_d – \mu_d)^2}{2\sigma_d^2} \right] $$ where \( t_d \) is the driving duration. These models help generate realistic electric vehicle charging profiles, which are then optimized in the upper layer of the framework.
The two-layer optimization model consists of an upper layer for electric vehicle charging coordination and a lower layer for park integrated energy system dispatch. In the upper layer, the objective is to minimize grid load variance, maximize load peak-to-average ratio, and reduce charging costs for electric vehicle users. The combined objective function is: $$ F = \lambda_1 f_1 – \lambda_2 f_2 + \lambda_3 f_3 $$ where \( f_1 \) is the grid load variance, \( f_2 \) is the peak-to-average ratio, and \( f_3 \) is the charging cost, with weights \( \lambda_1, \lambda_2, \lambda_3 \) summing to 1. Constraints include charging power limits, battery state-of-charge boundaries, and charging duration requirements. For example, the state-of-charge after charging must satisfy: $$ S_{\text{end}} E – h d_{D+1} + P t_c \eta \geq 0.2 E $$ where \( S_{\text{end}} \) is the final state-of-charge, \( E \) is battery capacity, \( h \) is energy consumption per km, \( d_{D+1} \) is the next trip distance, \( P \) is charging power, \( t_c \) is charging time, and \( \eta \) is efficiency.
In the lower layer, the park integrated energy system is optimized to minimize total costs, including energy purchase, carbon trading, and wind curtailment penalties. The objective function is: $$ \min F = C_{\text{buy}} + C_{\text{CO}_2} + C_{\text{waste}} $$ where \( C_{\text{buy}} \) is the cost of purchasing electricity and gas, \( C_{\text{CO}_2} \) is the carbon trading cost under a stepped mechanism, and \( C_{\text{waste}} \) is the penalty for wasted wind power. The carbon trading cost is calculated as:
$$ C_{\text{CO}_2} = \begin{cases}
\lambda E_{\text{IES}} & \text{if } E_{\text{IES}} \leq l \\
\lambda (1 + \alpha) (E_{\text{IES}} – l) + \lambda l & \text{if } l < E_{\text{IES}} \leq 2l \\
\vdots & \vdots
\end{cases} $$
where \( \lambda \) is the base carbon price, \( \alpha \) is the growth rate, and \( l \) is the interval length. Energy balance constraints for electricity, heat, gas, and hydrogen are enforced. For example, the electricity balance is: $$ P_{\text{buy,e}}(t) = P_{\text{E,e}}(t) + P_{\text{EL,e}}(t) + P_{\text{ES,e}}(t) – P_{\text{DG}}(t) – P_{\text{CHP,e}}(t) – P_{\text{HFC,e}}(t) $$ where each term represents power from grid purchase, electrical load, electrolyzer, electrical storage, wind generation, CHP, and fuel cell.
The improved whale optimization algorithm is employed to solve this complex optimization problem. Traditional whale optimization algorithm is enhanced by using Tent chaotic mapping for population initialization and a hunting speed control factor for iteration updates. The Tent chaotic sequence is generated as: $$ X(m+1) = \begin{cases} 2X(m) & \text{if } 0 \leq X(m) \leq 0.5 \\ 2(1 – X(m)) & \text{if } 0.5 < X(m) \leq 1 \end{cases} $$ This ensures a uniform distribution of initial solutions. The hunting speed control factor \( V \) is defined as: $$ V = \gamma \left(1 – \left(\frac{t_{\text{gen}}}{t_{\text{maxgen}}}\right)^\theta\right) $$ where \( t_{\text{gen}} \) is the current iteration, \( t_{\text{maxgen}} \) is the maximum iterations, and \( \gamma \), \( \theta \) are parameters. This factor adjusts the search scope, improving global exploration and local refinement.
To validate the proposed strategy, a case study is conducted with a park integrated energy system including 200 electric vehicles. Scenarios are compared: disordered charging versus ordered charging, traditional power-to-gas versus improved two-stage power-to-gas, and different carbon trading mechanisms. Key parameters are summarized in the following tables.
| Parameter | Value |
|---|---|
| Number of Electric Vehicles | 200 |
| Battery Capacity (kWh) | 30 |
| Charging Efficiency | 0.9 |
| Energy Consumption per 100 km (kWh) | 15 |
| Slow Charging Power (kW) | 7 |
| Fast Charging Power (kW) | 20 |
| Average Driving Speed (km/h) | 60 |
| Device | Capacity (kW) | Upper Limit (%) | Lower Limit (%) | Ramp Rate (%) |
|---|---|---|---|---|
| Electrical Storage | 800 | 90 | 10 | 20 |
| Thermal Storage | 600 | 90 | 10 | 20 |
| Gas Storage | 400 | 90 | 10 | 20 |
| Hydrogen Storage | 300 | 90 | 10 | 20 |
| Equipment | Capacity (kW) | Energy Conversion Efficiency (%) | Ramp Rate (%) |
|---|---|---|---|
| Combined Heat and Power | 600 | 92 | 20 |
| Gas Boiler | 700 | 95 | 20 |
| Methane Reactor | 300 | 60 | 20 |
| Hydrogen Fuel Cell | 200 | 95 | 20 |
| Electrolyzer | 500 | 87 | 20 |
Results show that ordered charging of electric vehicles reduces charging costs by 21.85% and decreases load variance significantly. The improved power-to-gas system lowers carbon trading costs by 38.79% and wind curtailment penalties by 63.75%. The stepped carbon trading mechanism further incentivizes emission reductions. Comparative analysis of algorithms indicates that the improved whale optimization algorithm converges faster and achieves better solutions than traditional methods, especially as the number of electric vehicles increases.
In conclusion, the integration of electric vehicles into park integrated energy systems, combined with advanced hydrogen-based technologies and carbon market participation, offers a viable path toward low-carbon operations. The two-layer optimization strategy effectively balances economic and environmental objectives, demonstrating the importance of coordinated dispatch in future energy systems. This approach is particularly relevant for China, where electric vehicle adoption is accelerating, and carbon neutrality goals are paramount. Future work could explore dynamic demand response and multi-energy market interactions to further enhance system flexibility.