In the context of the global energy transition and carbon neutrality strategies, the electric car (electric car) industry is experiencing explosive growth. Statistical data indicate that the annual sales of new energy passenger vehicles worldwide reached 18 million units in 2024, with the Chinese market alone accounting for more than 60% of this share. By the end of 2024, the number of new energy electric cars in China had reached 31.4 million, yet the construction of charging infrastructure has failed to fully match this rapid development. The imbalance between the number of new energy electric cars and charging terminals is prominent. The ratio of electric cars to charging piles is approximately 8.05, which is far from the ideal level. The contradiction of insufficient charging facility supply is becoming increasingly acute.
Regional coverage disparities are also severe. In economically developed eastern coastal areas, the density of charging facilities is relatively high, with the number of public charging piles per 10,000 people reaching 150 in some cities. Conversely, in certain central and western regions, the number of public charging piles per 10,000 people is less than 5, showing a significant gap. Furthermore, the mismatch of resources during peak and valley periods is evident. During peak electricity consumption periods, the utilization rate of charging piles often exceeds 75%, resulting in queuing for charging. During valley periods, the utilization rate drops below 40%, indicating severe idle waste of resources. This situation not only impacts the user charging experience but also poses significant challenges to grid dispatching. Hence, I initiated research revolving around the strategic deployment of charging stations. This was driven by the objective to tackle the constraints associated with conventional plug-in infrastructure, understand the potential of nascent technologies like dynamic wireless charging, and suggest comprehensive solutions that enhance the overall support system for electric cars.
1. Research Framework and Fundamental Models for Electric Car Charging Station Planning
I established a comprehensive methodological framework for tackling the electric car charging station planning challenges. The planning challenges under study are related to a complicated ecosystem involving specific participants such as the electric car users, the charging station operators, transmission networks, and the road networks. The methods that I utilized in this work are explicated in the sections that follow. Moreover, in this chapter, I introduce some fundamental tools that have been central to my research, such as Carbon Emission Flow theory, Monte Carlo simulation, and the Gravity P-Median model for location analysis.
1.1 Methodology for Carbon Emission Flow
Carbon emission flow offers a means to trace the transfer of CO2 within a power network by linking energy flows to the physical quantities of emissions. Within my research framework, I define the carbon potential of node i at time period t. It represents the carbon emissions associated with the consumption of one unit of electricity at a specific node. I can derive the initial carbon potential of a node according to the power system state. The calculation integrates the power and carbon intensity of all generators connected to that node, along with the carbon flow density of the incoming branches. Mathematically, I represent this initial node carbon potential as follows:
$$ e^{0}_{i,t} = \frac{\sum_{l\in \Gamma_i} P^{t}_{l} \cdot \phi^{t}_{l} + \sum_{g\in G_i} P^{t}_{g} \cdot \mu^{t}_{g}}{P_{L,i}^{t}} $$
Here, \( G_i \) represents the aggregation of power generation units connected to node i, and the set \( \Gamma_i \) includes the branches transmitting power into node i. Meanwhile, \( P^{t}_{g} \) denotes the real power output of a generator, with its corresponding carbon emission intensity represented by \( \mu^{t}_{g} \). The real power flow through a specific branch is designated as \( P^{t}_{l} \), and its carbon flow density is symbolized by \( \phi^{t}_{l} \). The total load power at node i is denoted as \( P^{t}_{L,i} \).
1.2 Low-Carbon Incentive Electricity Pricing Mechanism
To better guide electric car users towards charging nodes that have lower carbon emissions, I have advocated for a low-carbon incentive electricity price mechanism, which is developed from the principles of carbon emission flow. Through this mechanism, the charge is subsidized based on the carbon potential of the node at which the charging station is situated. Essentially, an incentive is provided to users who charge at nodes with lower carbon potential, thereby reducing grid emissions associated with the electric car charging process. The pricing mechanics can be defined by the subsequent formula:
$$ S_{i,t,sd} = S_{t,base} – a (e_{i} – e_{a}) \cdot \rho $$
In this particular equation, \( S_{t,base} \) signifies the initial Time-of-Use (TOU) electricity price, and \( a \) is the subsidy coefficient that is designed to regulate the level of subsidy. The carbon potential of node i is represented by \( e_{i} \), while \( e_{a} \) refers to the average carbon potential across the grid. Additionally, \( \rho \) is used to denote the carbon price in the electricity market. This system creates an economic signal that incentivizes users to shift their charging behavior towards nodes that are considered more environmentally friendly.
1.3 Prediction of Electric Car Charging Demand
I identified the Monte Carlo simulation method as a suitable approach to predict the charging demands of electric cars because of its benefits in dealing with the stochastic characteristics of electric car use. This simulation method allows me to model the randomness associated with various user behaviors. The overall steps in my simulation process are as follows:
(1) Modeling the probability distribution for daily driving distance: I based my assumption on the premise that drivers of electric cars exhibit similar travel patterns to those of conventional vehicles. Analysis of private vehicle data indicates that the daily driving distance adheres to a log-normal distribution. The corresponding probability density function is expressed as:
$$ f_s(x) = \frac{1}{x \sigma_s \sqrt{2\pi}} \exp\left[-\frac{(\ln x – \mu_s)^2}{2 \sigma_s^2}\right] $$
(2) Modeling the probability distribution for the start time of charging: I found that the return time of private vehicles follows a normal distribution, whose probability density function is segmented to account for different times of day, as given below.
$$ f_s(x)=\begin{cases}
\frac{1}{\sigma_s\sqrt{2\pi}} \exp\left[-\frac{(x-\mu_s)^2}{2\sigma_s^2}\right], & 12 < x \le 24 \\
\frac{1}{\sigma_s\sqrt{2\pi}} \exp\left[-\frac{(x+24-\mu_s)^2}{2\sigma_s^2}\right], & 0 < x \le 12
\end{cases} $$
In my model formulation, I account for the assumption that a typical electric car’s charging process can be approximated using a constant or uniform power. Additionally, all electric cars were configured with identical battery capacities to simplify the calculations. Through the iterative application of these distributions and by integrating the individual charging loads, I was able to derive the aggregate load profile for an electric car fleet.
| Parameter | Mean | Variance |
|---|---|---|
| Daily Mileage | 3.2 | 0.88 |
| Charging Start Time | 17.6 | 3.4 |
By segmenting and simulating different functional zones (residential, commercial, and industrial) within a city, I obtained distinct load curves for each zone. As a result, I observed that the industrial zone produced the largest total charging demand, followed by the residential zone and subsequently the commercial zone, each with its own unique peak time.
Table 1-2 The outcome of Monte Carlo simulation for various zones
| Zone | Total Charging Amount | Charging Peak Time | Peak Load /kW |
|---|---|---|---|
| Residential Area | Medium | 21:00 | ~200 |
| Industrial Area | Highest | 17:00 | ~240 |
| Commercial Area | Lowest | 20:00 | ~150 |
1.4 Gravity P-Median (GPM) Model for Charging Station Siting
The conventional PM model is widely used to minimize the weighted travel distance. I have explored a more comprehensive version known as the Gravity P-Median model. In this model, I have operated under the assumption that the selection of a charging facility by an electric car driver is not exclusively dictated by the nearest distance principle. Instead, users are more inclined to use a facility based on its unique ‘attractiveness’ and the diminishing utility caused by its distance. The demand point i assigning its demand to a facility at j is governed by the subsequent probability function, which integrates the attributes of attractiveness and distance.
$$ f_{ij} = \frac{S_j d_{ij}^{-\beta}}{\sum_{j \in P} S_j d_{ij}^{-\beta}} $$
where \( S_j \) symbolizes the attractiveness of a charging station at candidate point j. The distance decay parameter, \( \beta \), reflects the influence of distance on user choice. In the context of my analysis, I have further refined the attractiveness calculation by considering dynamic elements. These elements, \(k_j\) as a scaling coefficient for station type, \( T_{t,j} \) representing time-dependent coefficients, and the node carbon potential \( e_i \), are expressed in the following equation:
$$ S_j = \frac{k_j \cdot T_{t,j} \cdot a_j}{d_{ij}^{\beta}} $$
The total travel distance, which is the objective function for siting decisions under the GPM framework, is calculated as:
$$ F(P) = \sum_{i \in I} \sum_{j \in P} \frac{S_j d_{ij}^{-\beta}}{\sum_{j \in P} S_j d_{ij}^{-\beta}} D_i d_{ij} $$
In order to demonstrate the effectiveness of my model, I simulated a small example involving 8 demand nodes and 4 candidate facilities. The results were clear. My GPM model was able to select the optimal set of facilities that led to lower expected total costs due to its consideration of user preferences for more attractive facilities.
| Model | Selected Facilities | Total Cost (¥×10³) |
|---|---|---|
| p-median Model | F2, F3 | 92.0 |
| Gravity p-median Model | F2, F4 | 85.3 |
2. A Two-Level Planning Model for Electric Car Charging Stations Considering Demand Response and Nodal Carbon Potential
I formulated a bilevel planning model in light of the pressing need to reconcile grid expansion with carbon peaking and carbon neutrality goals. The core innovation of this model lies in employing two distinct mechanisms, which I structure as a Stackelberg game between the different planning layers.
(1) The Upper-Level Model – Site Selection: At this level, my objective is to minimize the overall travel distance for electric car users. I employed the gravity p-median model as the core analytical method. The model is subject to certain constraints like closeness and service area restrictions.
$$ \min F(P) = \sum_{i \in I} \sum_{j \in P} f_{ij} D_i d_{ij} $$
In this context, the total number of charging stations is predetermined. The selection of the exact location for a station from a pool of candidates using binary variables is reflected in the model. This specific location matrix is later forwarded to the lower level model.
(2) The Lower-Level Model – Capacity Determination: In this stage, I designed the objective function to calculate the highest annual profit for the charging stations. The profit structure is multi-faceted and includes revenues from selling energy and low-carbon incentives. Conversely, it also accounts for various costs, such as the annualized investment costs of the stations, operational expenses, and a penalty cost associated with the waiting time of electric car users. I have integrated the carbon emission flows and changing carbon potential within the objective function.
$$ \max F = S_e + S_c – S_i – S_k – S_w $$
Here, \( S_e \) is the revenue from energy sales which is calculated as:
$$ S_e = \sum_{m=1}^{N_m} \sum_{i=1}^{n_{EV}} \sum_{t=1}^{24} [P_{i,t,EV} (s_{i,t,sd} – s_{t,w})] $$
where \( P_{i,t,EV} \) is the charging load, \( s_{i,t,sd} \) is the charging tariff of the station, and \( s_{t,w} \) is the wholesale electricity price.
The carbon-related revenue term, \( S_c \) is composed of two elements. The first part comes from trading carbon reduction credits in the market, while the second part represents subsidies for reducing carbon potential at the node. I formulated the latter part using the average carbon potential and nodal carbon potential as follows:
$$ S_{c2} = \sum_{m=1}^{N_m} \sum_{i=1}^{n_{EV}} \sum_{t=1}^{24} \{ \gamma (e_a – e_{n}) P_{i,t,EV} \} $$
where \(e_a\) is average carbon potential, \(e_n\) is nodal carbon potential, and \(\gamma\) is the subsidy carbon price. This equation encourages stations to be located at nodes that show lower carbon intensity. Penalty costs (\(S_w\)) are also introduced to offer economic compensation to users who have to wait because a station is at its full capacity, and I utilized this to balance capacity allocation.
Constraints in the lower level include:
1. Total demand constraint: The aggregate capacity of all charging stations within a 24-hour cycle needs to be equal to or exceed the total amount of electric car charging demand. That is formalized as follows:
\( \sum_{i=1}^{n_{EV}} P_{i,EV} \ge \sum_{t=1}^{24} \sum_{i=1}^{n_{EV}} P_{i,t,EV} \)
2. Charging station capacity constraint: I ensure that the capacity of the charging station surpasses 0.7 times its daily peak charging power to prevent over-conservative design.
\( \sum_{i=1}^{n_{EV}} P_{i,EV} \ge 0.7 \cdot \sum_{t=1}^{24} \max(P_{i,t,EV}) \)
3. Power flow constraints: This includes maintaining nodal voltage magnitudes \( V_i \) within defined limits and ensuring branch currents \(I_{ij}\) stay below their thermal limits.
\( V_{i,min} \le V_i \le V_{i,max} \), \( I_{ij} \le I_{ij,max} \)
2.1 Solution Methodology for the Bilevel Model
To handle the high non-linearity in the upper model, I utilized Particle Swarm Optimization (PSO). The upper model takes into account the initial attractiveness and distance matrices, with its algorithm parameters configured to explore combinations of n charging station sites. The resulting location information is passed to the lower model, where I leveraged the CPLEX solver, which effectively handles the economic optimization and capacity allocation challenge for maximizing profit. Once a lower-level solution is found, I calculate the resulting carbon potential at the nodes. This carbon potential data is then channeled back to inform and update the upper-level attractivness parameters, \( S_j \), in an iterative loop until the best cost-benefit solution is achieved.
2.2 Simulation and Scenario Analysis for the Bilevel Model
To validate my proposed model, I created a coupled simulation that integrated a 33-node transportation network with the IEEE 33-node distribution system. I established four distinct scenario conditions to analyze how carbon potential and demand response impact the system. The scenarios I designed varied the source of power generation and whether my low-carbon incentive mechanism was incorporated.
| Scenarios | Generation Mix | Pricing Scheme | Low-Carbon Signal |
|---|---|---|---|
| Scenario 1 | Thermal Dominated | TOU tariff | Not considered |
| Scenario 2 | Thermal Dominated | TOU tariff | Considered |
| Scenario 3 | Renewable Dominated | Modified TOU tariff | Not considered |
| Scenario 4 | Renewable Dominated | Modified TOU tariff | Considered |
(1) Analysis of Siting and Sizing Outcomes:
My simulation generated a set of optimal locations for charging stations, which included nodes such as 3, 13, 16, and 18. The associated capacity results for these stations under each scenario are summarized in the table below.
| Scenarios | Energy Sales Revenue (¥×10⁴) | Carbon Revenue (¥×10⁴) | Construction Cost (¥×10⁴) | Maintenance Cost (¥×10⁴) | Waiting Penalty Cost (¥×10⁴) | Total Profit (¥×10⁴) |
|---|---|---|---|---|---|---|
| Scenario 1 | 264.56 | 19.36 | 125.97 | 45.58 | 14.32 | 98.04 |
| Scenario 2 | 261.32 | 32.92 | 128.31 | 46.43 | 15.68 | 103.81 |
| Scenario 3 | 304.52 | 19.51 | 123.95 | 44.13 | 16.95 | 139.00 |
| Scenario 4 | 296.65 | 33.95 | 124.94 | 42.38 | 17.09 | 146.18 |
From my analysis of the results, I confirmed that the carbon incentive mechanism notably shifts the charging load distribution from high carbon intensity nodes to those with lower values, contributing to significant reductions in overall waiting penalties and increased total profits.
(2) Impact on Node Carbon Potential:
The comparison between the carbon output of the entire system in scenario 1 (where the grid had an average carbon potential near 0.52 kg/(kW·h)) and scenario 4 (which had a much lower average carbon potential of 0.22 kg/(kW·h)) proved the effectiveness of my model. This transition to renewable sources and the use of incentives lead to a significant 57% reduction in overall average carbon potential within the network.
(3) Impact on power grid stability:
I have also analyzed voltage fluctuations in the low-carbon scenario. I found that the low-carbon incentive model effectively schedules charging for valley periods or periods of high renewable output, and when it redistributes loads to greener nodes it prevents heavily loaded nodes from causing severe voltage dips. My calculation of the voltage deviation at node 18 at 20:00 in scenario 4 reveals a decrease in about 16.4% compared to the standard scenario 1. These operational improvements highlight the system-level benefits of including carbon potential into the planning process for electric car charging infrastructure.
3. Integrated Planning and Stackelberg Game Scheduling for Plug-in Stations and Dynamic Wireless Charging Roads
Given the drawbacks that are inherent in plug-in charging, such as its time-consuming nature and contribution to peak load, I investigated the potential of dynamic wireless charging roads. This emerging technology enables electric cars to charge while in motion, effectively mitigating “range anxiety” and lessening the pressure on the grid. Given this, I formulated a joint planning model for these two types of charging infrastructure. This model is constructed to balance the advantages and limitations of each technology.
(1) Joint Planning Model Structure:
The objective of my joint planning model is to maximize the combined operational profits from both the plug-in charging stations and the dynamic wireless charging roads. The model was designed to optimize the locations and capacities of these assets. The financial objective can be described as:
$$ \max \sum_{t=1}^{N_m} ( C_{sa} – C_{pm} – C_{pc} ) – ( C_{inv} + C_{op} ) $$
Here, C_sa represents the charging revenue, C_pm is the cost of purchasing power from the grid, and C_pc refers to penalties for failing to meet energy service level agreements. I also included C_inv as the amortized capital investment for building the charging stations and installing the dynamic wireless charging road equipment. C_op captures the operational and maintenance expenditures for these installations.
(2) Energy Demand Assignment Model:
In this research, I used an energy demand assignment model to more efficiently convert what is normally a complex transportation flow problem into an energy-based calculation, avoiding the need to model individual vehicle movements. This model directly assigns the aggregate energy demand to the nodes and edges of the traffic network based on capacity constraints.
$$ \min \text{CH}(E_f, E_d) $$
The model is subject to constraints that enforce that the energy requirements of electric cars, represented as \( E_f \) and \( E_d \) for plug-in and dynamic charging respectively, must not exceed the maximum energy deliverable by the traffic and grid infrastructure. \( E_u\) signifies the average energy gap for each vehicle and \( E_g\) is the energy gap for an OD pair, which sets the boundary.
An analysis of cost under different budgets was conducted using a Nguyen-Dupuis network-IEEE 14 node system:
| Total Budget | Energy Sales Revenue | Purchase Cost | Construction Cost | Operational Cost | Penalty Cost | Total Profit |
|---|---|---|---|---|---|---|
| 20,000 | 4,953 | 990 | 1,744 | 659 | 193 | 1,367 |
| 50,000 | 8,250 | 1,035 | 4,385 | 859 | 223 | 2,607 |
| 90,000 | 10,830 | 1,159 | 7,848 | 763 | 52 | 1,008 |
From this table, I deduce that the optimal total budget for charging station operators is the mid-range budget. As the budget increases, a larger proportion is allocated to the expensive dynamic wireless charging roads. Despite these high construction costs, the most substantial overall profit is realized at the intermediate budget level because the large investments in dynamic wireless charging present high risk without commensurate returns. Conversely, with the lowest budget, the operator must rely solely on plug-in stations which have lower energy sales revenue and higher penalties due to queuing.
In simulating a larger case study with the Sioux-Falls network and an IEEE 33-node distribution system, the results similarly highlighted the benefits of combining 7 strategically placed plug-in charging stations with 8 dynamic wireless charging roads. These results suggest that dense downtown areas are more fitting for dynamic wireless charging, while sparse suburban locations warrant the use of plug-in stations to maximize overall efficiency and coverage.
3.1 Master-Slave Game Model for Operation Optimization
To further refine the operations, I built a Stackelberg game model that captures the conflicting interests between the charging station operator and the electric car user. This framework was used to determine an optimal pricing strategy and the charging decisions of the users. The operator acts as the leader and establishes a time-of-use pricing scheme that is designed to maximize its own profitability. Acting as followers, the electric car users respond by altering their charging strategies to minimize their own risk and travel costs.
Upper-Level Leader – Charging Station Operator:
In this model, I defined the operator’s revenue function, G1, as the difference between the income from selling electric car charging services and the cost of purchasing electricity.
$$ G_1 = \sum_{t} ( S_{cd1,t} P_{1,t} + S_{cd2,t} P_{2,t} – S_{gd,t} P_{all,t} ) $$
Where \( S_{cd1,t} \) and \( S_{cd2,t} \) are the charging prices for plug-in and dynamic wireless charging services, and \( P_{1,t} \) and \( P_{2,t} \) are the corresponding power consumed at time t. The price of wholesale power at the same time is \( S_{gd,t} \), and total power bought is \( P_{all,t} \).
Lower-Level Followers – Electric Car Users:
I formulated the cost function of electric car users, G2, to incorporate their expenditure on charging and the costs associated with time spent in queues and vehicle depreciation.
$$ G_2 = \sum_{t} ( S_{cd1,t} P_{1,t} + S_{cd2,t} P_{2,t} + S_w T_w + S_{sh} (P_{1,t} + P_{2,t}) ) $$
Here, \( S_w \) signifies the cost per unit time and \( T_w \) is the total waiting time. I included \( S_{sh} \) to express the depreciation cost, which is essential for capturing the complete economic impact on the user. The game is subject to several constraints, including balance constraints, capacity constraints on the grid and storage, maximum and minimum limits for the pricing schemes, and PV generation limits.
The Stackelberg game concludes in a unique equilibrium due to the convex nature of the objective functions of all participants in relation to their strategies. I solved the model using an iterative approach. The leader’s pricing problem is solved with Particle Swarm Optimization, while the user response is optimized using a CPLEX solver, iterating until convergence to stable solutions.
3.2 Analysis of Master-Slave Game Simulation Results
To analyze the effect of my proposed game-theoretic approach, I used the planning results for the Sioux-Falls network-IEEE 33 node distribution system. Based on comparisons between the pre-game and post-game states, I was able to observe a dynamic shift in the time-of-use tariffs. After the game, the price of plug-in charging fell below the static 1.5 ¥/kW·h during all periods, with the lowest price at 0.5 ¥/kW·h. The game also resulted in a dynamic pricing scheme, where the maximum plug-in price of 1.35 ¥/kW·h occurs during the evening peak hours (19:00-23:00). Dynamic wireless charging price, originally at 3 ¥/kW·h, became even higher during morning rush (7:00-9:00) and afternoon rush (16:00-20:00) when both the grid load and vehicle flow peak, thereby discouraging charging at those critical times.
Table 3-2 Grid load-related metrics pre- and post-game implementation
| Metric | W/O Stackelberg game (MW) | With Stackelberg game (MW) |
|---|---|---|
| Peak-Valley Difference | 119.5 | 105.2 |
| Variance | 34.1 | 28.9 |
Following the game, I found that the load profile is more flattened. The peak-valley gap decreased by 11.9% and the variance reduced by 15.2%, as seen in the table. This signifies a more stable and reliable grid operating point. The simulation showed that demand at the evening peak was reduced by as much as 8.1%, while the load during valley times increased by 22.8%. This shifting effect, guided by the pricing incentives, facilitates effective peak shaving and valley filling.
Table 3-3 Costs and revenues accrued by electric car users and operators under game scenarios (in ¥×10⁴)
| Scenario | Depreciation cost | Time cost | Charging cost | Total cost | Operator’s profit |
|---|---|---|---|---|---|
| Pre-Game | 7.32 | 17.24 | 85.29 | 109.85 | 50.94 |
| Post-Game | 7.52 | 21.04 | 71.19 | 97.75 | 47.46 |
I observed that the game effect lowers charging costs for electric cars by up to 16.5% because the price signals urge users to seek cheaper periods for charging. While it is true that the time cost significantly increases by 21.9%, the substantial reduction in charging costs still leads to a beneficial overall reduction in total costs of 11%. The operator experiences a drop in profit of about 6.8%, which is a smaller percentage change than the benefits accrued to the electric car users. This is a more balanced result for a charging market.
4. Conclusion
Through my research, I have developed and validated models to support the coordinated planning and operation of plug-in charging stations and dynamic wireless charging roads for electric cars. In this body of work, I have arrived at several key conclusions from my analysis:
(1) I integrated carbon emission flow principles and a low-carbon incentive pricing mechanism. This approach effectively guides electric car users to charge at nodes with lower carbon potential, channeling load shifts to greener areas of the grid. The framework allowed for the strong mitigation of the environmental impact of electric car charging, contributing to a decrease in overall node carbon potential of up to 57% in scenarios where renewable generation sources are abundant. This result demonstrates that charging infrastructure planning is a powerful tool for achieving system-wide decarbonization goals, without significantly compromising profit.
(2) My simulation exercises in both small (14-node) and larger (33-node) networks provided valuable insights into the allocation of investments between plug-in and dynamic wireless charging infrastructure. I observed an inverse relationship between the number of plug-in charging stations and the total construction budget, confirming that dynamic wireless charging, while beneficial, is suitable for implementation only in scenarios that have enough capital. The model suggests that dynamic wireless charging roads are more appropriate within high-density traffic zones. In contrast, plug-in charging stations are better suited for serving low-density or boundary areas of the transportation network. This strategy provides a cost-effective and space-efficient coverage of charging needs.
(3) The master-slave game model that I introduced creates a mechanism to mediate the conflicting goals of charging station operators and electric car users. In the post-game period, the time-of-use price offered to electric cars at plug-in stations became lower than the static tariff in all hours, and dynamic wireless charging prices escalated during peak traffic periods to restrict demand. The implementation of this equilibrium strategy reduces the users’ total travel costs, mostly due to lower charging fees despite higher travel times. Furthermore, the scheduling flattening effect of the pricing on the overall grid profile decreases the peak-valley gap by 11.9% and an overall load variance reduction of 15.2%. This creates a dual benefit: it cuts costs for electric car users and enhances grid stability by smoothing out demand curves. This dual benefit emphasizes the effectiveness of the strategic game for fostering a synergic relationship between electric car adoption and grid operations.