In recent years, the rapid urbanization and increasing population density in cities have placed significant pressure on electrical grids, particularly in commercial districts. These areas experience high foot traffic, which correlates strongly with electricity demand peaks during evenings and weekends. As a result, commercial zones often face challenges such as elevated peak loads, substantial peak-valley differences, and high electricity costs. Traditional energy storage solutions, like electrochemical or mechanical systems, have been widely studied and implemented. However, they may not be well-suited for small-scale, dynamic environments like commercial districts due to scalability and adaptability issues. This study addresses these challenges by leveraging electric vehicles (EVs) as distributed energy storage units. We propose a novel planning model that utilizes electric vehicle charging and discharging to optimize load supply-demand balance in commercial areas. By considering the convergence of pedestrian flow and load patterns, we employ an improved Imperial Competition Algorithm to plan charging pile locations and implement incentive strategies for EV participation. Our approach aims to reduce peak loads, minimize peak-valley differences, and generate economic benefits, with a focus on the growing adoption of electric vehicles in China, often referred to as China EV.
The core idea of our model is to treat individual electric vehicles as mobile energy storage units that can discharge power back to the grid during high-demand periods. This not only alleviates grid stress but also promotes the integration of renewable energy sources. Given the stochastic nature of EV usage, we base our analysis on statistical data to characterize parking duration and energy consumption. The dynamic state of charge (SOC) model for the commercial area is formulated as follows:
$$ SOC_s(t_{k+1}) = SOC_s(t_k) + \frac{t}{n_t(t_k) Q_i(t_k) – [1 – n_t(t_k)] \delta_t(t_k) E_i} $$
where \( SOC_s(t_k) \) and \( SOC_s(t_{k+1}) \) represent the SOC at times \( t_k \) and \( t_{k+1} \), respectively; \( E_i \) is the battery capacity of the i-th electric vehicle; \( Q_i(t_k) \) is the charging or discharging power of the i-th vehicle at time \( t_k \); \( n_t(t_k) \) is the parking lot utilization rate at time \( t_k \); and \( \delta_t(t_k) \) is the average energy consumption per vehicle at time \( t_k \). This model allows us to simulate the energy balance in real-time, considering the fluctuating nature of EV presence in commercial areas.
To plan the locations of charging piles, we define an initial matrix based on pedestrian flow and load distribution. The matrix distinguishes between high-capacity and low-capacity piles, denoted as P and p, respectively. For instance, P represents a charging device with a capacity of 2,500 kVA and a cost of 199,000 units, while p refers to a device with 800 kVA capacity and 69,000 units cost. The initial pile location matrix is given by:
$$ P_{\text{location}} = \begin{bmatrix} P_1(X, Y) \\ \vdots \\ P_m(X, Y) \\ p_1(x, y) \\ \vdots \\ p_n(x, y) \end{bmatrix}^T $$
where \( P_m(X, Y) \) indicates the m-th high-capacity pile in zone X and charging area Y, and \( p_n(x, y) \) denotes the n-th low-capacity pile in zone x and charging area y. This matrix serves as the foundation for optimizing pile placement to maximize efficiency.
The efficiency of each charging pile is evaluated based on its utilization rate, defined as the proportion of time during which an electric vehicle is actively charging or discharging over a 24-hour period. We calculate the normalized influence value \( C_n \) for each pile as:
$$ C_n = c_n – \max_i \{ c_i \} $$
where \( c_n \) is the influence function corresponding to the n-th pile. The relative influence \( P_n \) and the efficiency N are then derived as:
$$ P_n = \left| \frac{C_n}{\sum_{i=1}^{N_{\text{imp}}} C_i} \right| $$
$$ N = \text{round}(P_{\text{location}} \cdot N_{\text{col}}) $$
Here, \( N_{\text{col}} \) represents the number of electric vehicles, and the round function ensures integer values for practical implementation. This approach helps in identifying optimal pile locations that align with high pedestrian traffic and load demands.
As commercial districts evolve, changes in pedestrian flow and load patterns necessitate adaptive planning. To address this, we introduce a reversible improvement to the algorithm, allowing for dynamic adjustments. The influence value \( TP_n \) for each region is defined as:
$$ TP_n = \text{Cost}(\text{imperialist}_n) + \xi \cdot \text{mean}\{\text{Cost}(\text{colonies})\} $$
where \( \xi \) is a weighting factor. This value reflects the region’s ability to attract more electric vehicles based on factors like promotional incentives. The relationship between incentive strategies and customer participation is modeled using a logistic function:
$$ \ln \left[ \frac{g(\text{click} = 1)}{1 – g(\text{click} = 1)} \right] = \alpha_0 + \alpha_1 \text{duration} + \alpha_2 \text{acquisition} + \alpha_3 \text{frequency} + \alpha X_i $$
In this equation, \( g(\text{click} = 1) \) represents the probability of a customer participating in the charging or discharging program, \( \alpha_0 \) is the intercept, \( \alpha \) are coefficients for variables such as duration of stay, acquisition cost, and visit frequency, and \( X_i \) are control variables. This model helps in designing effective discount strategies to encourage EV owners to engage in grid-supportive activities.
To further enhance the algorithm’s adaptability, we modify the influence value to account for dynamic load changes in transformer areas. The improved influence value \( TP’_n \) is given by:
$$ TP’_n = \frac{\text{Cost}(\text{imperialist}_n)}{g} + \xi \cdot \text{mean}\{\text{Cost}(\text{colonies})\} $$
where g is a factor related to the incentive effectiveness. This adjustment ensures that the planning model remains robust against fluctuations in commercial district development.
The economic benefits of implementing this model are significant. By participating in energy transactions, electric vehicles act as aggregated storage units, interacting with the local grid. The financial gain for the commercial area can be expressed as:
$$ R = \Delta t \sum_{k=1}^{N-1} \left\{ -\gamma(t_k) P_{\text{EV}}(t_k) + \left[ \gamma(t_k) \phi(t_k) P_{\text{EV}}(t_k) + \gamma(t_k) \phi(t_k) \kappa[P_{\text{EV}}(t_k)] \right] P_{\text{EV}}(t_k) \right\} $$
Here, \( \gamma(t_k) \) is the electricity market price at time \( t_k \), \( \phi(t_k) \) is the price difference between the commercial area’s purchase price and the market price, \( P_{\text{EV}}(t_k) \) is the charging or discharging power of electric vehicles at time \( t_k \), and \( \kappa[P_{\text{EV}}(t_k)] \) is a charging coefficient that equals 1 if \( P_{\text{EV}}(t_k) > 0 \) (discharging) and 0 otherwise. This formulation captures the revenue generated from energy arbitrage and grid services.
In our experimental analysis, we applied this model to a typical commercial district supplied by two 6 MW transformers. The load profile from Monday to Sunday exhibited dual-peak characteristics, with evening peaks being particularly pronounced on weekends. For instance, the maximum load reached 10.43 MW during weekend evenings, with a load rate of 91% and a peak-valley difference of 8.1 MW. Assuming peak electricity prices of 1.2 CNY/kWh and off-peak prices of 0.3 CNY/kWh in Shanghai, we set a buyback price of 0.5 CNY/kWh for energy discharged by electric vehicles.
Using the proposed planning model, we optimized the locations of charging piles to attract customers while utilizing their electric vehicles for grid support. We assumed that 20% of EVs would participate in charging and discharging activities, with a discharge power of 2 kW per vehicle during parking. The initial planning results showed a reduction in peak load from 10.43 MW to 8.1 MW, a decrease of 22.3%, and the peak-valley difference was reduced to 6.3 MW. However, with the improved algorithm, the results were even more impressive: during the late evening peak on weekends, the peak load dropped from 9.5 MW to 6.1 MW, a reduction of 35.8%, and the daily peak-valley difference was minimized to 4.6 MW. Overall, the highest load was reduced from 10.43 MW to 6.3 MW, freeing up 4.13 MW of capacity for other commercial activities.
The following table summarizes the key performance metrics before and after implementing the improved model:
| Metric | Before Implementation | After Initial Planning | After Improved Algorithm |
|---|---|---|---|
| Peak Load (MW) | 10.43 | 8.1 | 6.3 |
| Peak-Valley Difference (MW) | 8.1 | 6.3 | 4.6 |
| Reduction in Peak Load (%) | – | 22.3% | 39.0% |
| Reduction in Peak-Valley Difference (%) | – | 22.2% | 43.2% |
This table highlights the progressive improvements achieved through our model, demonstrating its effectiveness in load balancing. The integration of electric vehicles, especially in the context of China EV trends, plays a crucial role in enhancing grid stability and economic efficiency.
Furthermore, we analyzed the impact of various factors on the model’s performance, such as the percentage of participating electric vehicles and the discharge power settings. The results indicate that higher participation rates lead to greater load reductions, but diminishing returns occur beyond a certain threshold due to infrastructure limitations. For example, increasing the participation rate from 20% to 30% resulted in an additional 5% reduction in peak load, but further increases had minimal effects without expanding the number of charging piles.
Another critical aspect is the cost-benefit analysis of implementing this system. The initial investment includes the cost of charging piles and installation, which can be offset by the savings from reduced peak demand charges and revenue from energy sales. Based on our simulations, the payback period for a typical commercial district is approximately 3-5 years, considering the current electricity prices and EV adoption rates in China. This makes the model economically viable for widespread deployment.
In conclusion, our study presents a comprehensive framework for optimizing load balance in commercial areas using electric vehicles as flexible energy resources. The proposed model leverages the synergy between pedestrian flow and electricity demand to plan charging infrastructure dynamically. By incorporating reversible improvements and incentive strategies, we achieve significant reductions in peak loads and peak-valley differences, leading to enhanced grid reliability and substantial economic benefits. The growing prevalence of electric vehicles in China, often termed China EV, underscores the practicality of this approach. Future work could explore integration with renewable energy sources and real-time demand response mechanisms to further improve efficiency. This research contributes to the broader goal of sustainable urban development and smart grid innovation.
To facilitate understanding, the following table provides a glossary of key variables and parameters used in our models:
| Symbol | Description |
|---|---|
| \( SOC_s(t_k) \) | State of charge of the commercial area at time \( t_k \) |
| \( E_i \) | Battery capacity of the i-th electric vehicle |
| \( Q_i(t_k) \) | Charging/discharging power of the i-th vehicle at \( t_k \) |
| \( n_t(t_k) \) | Parking lot utilization rate at \( t_k \) |
| \( \delta_t(t_k) \) | Average energy consumption per vehicle at \( t_k \) |
| \( P_{\text{location}} \) | Matrix representing initial charging pile locations |
| \( C_n \) | Normalized influence value for the n-th pile |
| \( P_n \) | Relative influence value |
| \( N \) | Efficiency of charging piles |
| \( TP_n \) | Influence value for region n |
| \( g(\text{click} = 1) \) | Probability of customer participation |
| \( R \) | Economic revenue from energy transactions |
| \( \gamma(t_k) \) | Electricity market price at \( t_k \) |
| \( \phi(t_k) \) | Price difference for buyback |
| \( \kappa[P_{\text{EV}}(t_k)] \) | Charging coefficient (1 for discharging, 0 otherwise) |
This detailed exposition underscores the viability of using electric vehicles for load management in commercial districts. As the adoption of electric vehicles continues to rise, particularly in markets like China EV, such strategies will become increasingly important for achieving energy sustainability and resilience.