In recent years, the rapid growth of electric vehicle adoption, particularly in regions like China EV markets, has highlighted the need for advanced energy management systems. Energy routers serve as pivotal devices in the energy internet, enabling the integration of distributed energy resources such as photovoltaic (PV) systems, energy storage, and electric vehicles. The increasing penetration of electric vehicles introduces challenges related to battery management, including state of charge (SOC) imbalances that can lead to overcharging or overdischarging, ultimately affecting battery lifespan and system stability. This paper addresses these issues by proposing a coordinated control strategy for energy routers that incorporates SOC balancing for multiple electric vehicles. We focus on a distribution network energy router with interfaces for the grid, PV, energy storage, DC output, and multiple electric vehicle charging ports. By leveraging DC bus voltage signaling (DBS) for coordination, our approach prioritizes the utilization of PV generation while maintaining system stability. Furthermore, we introduce an enhanced droop control method based on consensus algorithms to achieve SOC balancing among electric vehicles, ensuring equitable power distribution and prolonging battery life. Through extensive analysis and simulation, we demonstrate the effectiveness of our control strategy in various operational scenarios, emphasizing its applicability to modern energy systems with high electric vehicle integration.
The structure of the energy router considered in this study comprises several key ports: a distribution grid port, a PV port, an energy storage port, multiple electric vehicle charging ports, and DC/AC load ports. The grid port utilizes a cascaded H-bridge (CHB) configuration for AC/DC conversion, while a dual-active bridge (DAB) structure provides electrical isolation and energy transfer. The PV port connects to a boost converter for maximum power point tracking (MPPT) or droop control, depending on system conditions. The energy storage and electric vehicle ports employ bidirectional DC/DC converters to manage power flow, allowing for charging and discharging operations. The DC bus voltage serves as a critical signal for coordinating these components, enabling seamless mode transitions based on voltage levels. This topology supports the integration of renewable energy and electric vehicles, facilitating vehicle-to-grid (V2G) capabilities where electric vehicles act as mobile storage units. The flexibility of this design is essential for accommodating the dynamic nature of electric vehicle usage, especially in contexts like China EV deployments, where charging infrastructure must handle varying demands.
To ensure efficient operation, we implement a hierarchical control strategy divided into three layers: power dispatch, DC bus voltage control, and converter control. The power dispatch layer receives scheduling commands from higher-level systems and adjusts the energy router’s operation accordingly. The DC bus voltage control layer monitors the bus voltage and switches between four distinct operating modes to optimize PV utilization and maintain voltage within acceptable limits. These modes are defined based on voltage thresholds, as summarized in Table 1. For instance, when the bus voltage exceeds 1.03 per unit (p.u.), indicating surplus PV power, the system enters Mode 1, where PV operates in droop control to limit output, and excess power is absorbed by the grid, storage, or electric vehicles. In contrast, during low voltage conditions (e.g., below 0.97 p.u. in Mode 3), the grid, storage, and electric vehicles supplement the power deficit. The converter control layer executes local commands, such as MPPT for PV or current-voltage loops for storage and electric vehicle converters, ensuring rapid response to changes in load or generation.
| Mode | Voltage Range (p.u.) | PV Control | Power Balance Equation |
|---|---|---|---|
| 1 | 1.03 – 1.05 | Droop Control | $$ P_{PV} = P_g + P_b + P_L + P_{EV} $$ |
| 2 | 1.00 – 1.03 | MPPT | $$ P_{PV} = P_g + P_L + P_{EV} + P_b $$ |
| 3 | 0.97 – 1.00 | MPPT | $$ P_{PV} + P_g + P_{EV} + P_b = P_L $$ |
| 4 | 0.95 – 0.97 | MPPT | $$ P_{PV} + P_g + P_b + P_{EV} = P_L $$ |
The core of our approach lies in the SOC balancing control for multiple electric vehicles. When electric vehicles are connected to the energy router, they function as distributed storage units, and their SOC levels must be managed to prevent imbalances that could cause premature battery degradation. Traditional droop control, expressed as $$ U_{ref} = U^* – k \cdot I $$, where \( U_{ref} \) is the reference voltage, \( U^* \) is the nominal bus voltage, \( k \) is the droop coefficient, and \( I \) is the output current, often fails to achieve SOC equilibrium due to variations in battery capacity and line impedance. To address this, we employ a consensus-based algorithm that dynamically adjusts the power output of each electric vehicle based on its SOC. The SOC for the \( g \)-th electric vehicle is defined as:
$$ SOC_g = SOC_{g0} – \frac{\int I_g \, dt}{C_{eg}} $$
where \( SOC_{g0} \) is the initial SOC, \( I_g \) is the output current, and \( C_{eg} \) is the battery capacity. Differentiating this equation gives:
$$ \dot{SOC_g} = -\frac{I_g}{C_{eg}} $$
This implies that SOC balancing requires the ratio of output current to capacity to be equal across all electric vehicles. However, in practical scenarios with mismatched impedances, this condition is not met. Therefore, we introduce a dynamic average consensus estimator for SOC:
$$ \overline{SOC_g} = SOC_g + \frac{\beta_i}{|N_g|} \int_0^t \sum_{j \in N_g} (\overline{SOC_j} – \overline{SOC_g}) \, d\tau $$
where \( \overline{SOC_g} \) and \( \overline{SOC_j} \) are the average SOC estimates for electric vehicles \( g \) and \( j \), \( \beta_i \) is the consensus gain, and \( N_g \) is the set of neighboring electric vehicles. Additionally, we define an equilibrium factor \( \eta_i \) to adjust power distribution:
$$ \eta_i = e^{\lambda (\overline{SOC_j} – \overline{SOC_g})} $$
with \( \lambda = \mp 1 \) for charging and discharging modes, and \( \eta_i \) bounded between \( l \) and \( h \) to reflect battery constraints. This factor is used to compute a virtual current \( i_{vEVg} = \eta_i C_{rg} i_{pg} \), where \( C_{rg} \) is a rating factor. A virtual current consensus controller then generates a voltage correction term:
$$ \rho_{vEVg} = k_{pb} (\overline{i_{vEVg}} – i_{vEVg}) + k_{ib} \int (\overline{i_{vEVg}} – i_{vEVg}) \, dt $$
where \( k_{pb} \) and \( k_{ib} \) are proportional and integral gains. This correction is added to the reference voltage, enabling adaptive current sharing that promotes SOC convergence. The overall control strategy for electric vehicle ports is illustrated in Figure 7, which integrates these elements to achieve robust performance.
To validate our proposed control strategy, we conducted simulations using MATLAB/Simulink, with key parameters listed in Table 2. The energy router was tested under various conditions, including islanded and grid-connected modes, with multiple electric vehicles having different initial SOC levels. In Scenario 1, under islanded operation with PV power less than load demand, the energy storage discharged to stabilize the bus voltage, causing its SOC to decrease. When PV power increased, the storage switched to charging, and two electric vehicles with initial SOC values of 46% and 49% were connected in charging mode. The SOC balancing control gradually reduced the disparity, achieving equilibrium around 120 seconds while maintaining bus voltage stability. Similarly, in Scenario 2, during grid-connected operation with low PV power, electric vehicles discharged to support the system, and their SOC levels converged from 81% and 77% to a common value within approximately 600 seconds. These results confirm that our approach effectively manages SOC imbalances without compromising voltage regulation, highlighting its suitability for real-world applications involving electric vehicles.
| Parameter | Value |
|---|---|
| DC Bus Voltage | 700 V |
| Grid Voltage (Phase A) | 220 V |
| DC Bus Capacitance | 1100 μF |
| Boost Output Capacitance | 200 μF |
| DC Bus Filter Inductance | 3 mH |
| Storage Side Inductance | 3 mH |
| VSC Filter Inductance | 4 mH |
| Switching Frequency | 15 kHz |
| Grid Frequency | 50 Hz |
In Scenario 3, we examined cases with high PV generation, where the bus voltage exceeded 1.05 p.u., triggering droop control for the PV converter. The surplus power was initially absorbed by the storage, but when the storage SOC reached 90%, it ceased charging, and the grid began importing power. Throughout these transitions, the bus voltage remained within designated limits, demonstrating the robustness of our hierarchical control. The integration of electric vehicles as flexible resources played a crucial role in mitigating voltage fluctuations, underscoring the importance of SOC management in enhancing system resilience. This is particularly relevant for China EV ecosystems, where the density of electric vehicles can strain local grids. Our control strategy not only addresses technical challenges but also aligns with sustainability goals by maximizing renewable energy use and extending battery life.
In conclusion, this paper presents a comprehensive coordinated control framework for energy routers that incorporates SOC balancing for multiple electric vehicles. By leveraging DC bus voltage signaling and consensus-based algorithms, we achieve stable operation across various modes while ensuring equitable power distribution among electric vehicles. The simulation results validate the effectiveness of our approach in maintaining voltage stability and SOC equilibrium, even under dynamic conditions. Future work could explore real-time optimization and communication latency effects, further refining the strategy for large-scale deployments. As the adoption of electric vehicles continues to rise, especially in markets like China EV, such advanced control mechanisms will be essential for building resilient and efficient energy internet infrastructures.