MATLAB Simulation of Neural Network Energy Management in PV Powered EV Charging Station
Overview of the System
The system in focus comprises a solar photovoltaic (PV) panel, two batteries (one for storage and the other for the EV), and a connection to the power grid through an inverter. These components are interconnected via a common DC link, ensuring the smooth operation of the charging station. The energy management system intelligently regulates the energy flow based on PV power generation, battery status, and grid input, ensuring optimal charging conditions for the EV.
The Role of the PV Panel
The solar PV panel in the system consists of series-connected modules with parallel strings. Under ideal radiation conditions (1,000 W/m²), the panel generates a maximum of 2 kW of power. The PV panel’s operation is optimized using a Maximum Power Point Tracking (MPPT) algorithm, which adjusts its output according to varying radiation and temperature conditions. The system uses an adaptive neuro-physic control algorithm for MPPT, taking radiation and temperature as inputs to determine the optimal output voltage.
A boost converter is employed to connect the PV array to the common DC link, which allows the PV to operate at its maximum power point and ensures a consistent energy supply to the system.
Storage Battery Operation
The storage battery plays a crucial role in energy management by storing excess energy generated by the PV panel. This battery is connected to the common DC link through a bidirectional DC-DC converter, which allows for both charging and discharging operations in buck or boost modes. A voltage control method is implemented for the storage battery, where the reference voltage is set at 500V. The battery discharges to supply power when required, ensuring a reliable energy supply to the EV battery and compensating for periods when PV output is low.
EV Battery Operation
The EV battery is also connected to the DC link through a bidirectional DC-DC converter, using the same voltage control method as the storage battery. This ensures the EV battery receives the appropriate amount of power for charging. The EV battery charges continuously as part of the system’s energy management strategy, ensuring that the electric vehicle is ready for use without interruption.
Integration with the Power Grid
The system integrates a single-phase grid through an inverter, which is connected to the common DC link via an LCL filter. The inverter’s role is to manage the energy transfer between the grid, the batteries, and the PV panel. Energy management is crucial in ensuring the grid receives or supplies energy at the right times based on the system’s needs.
Neural Network-Based Energy Management
The key innovation in this system is the neural network-based energy management approach. This control technique uses inputs from the storage battery and PV power to determine the optimal reference current. The neural network processes this data and generates a reference current that is compared with the actual current produced by the inverter. The error between the reference and actual current is processed by a PID controller, and switching pulses are generated for the inverter. This ensures that the inverter operates efficiently, maintaining a stable energy flow to the EV battery.
Simulation Results and Power Management
The MATLAB simulation of this system was conducted under various irradiation conditions, showing how the power management adjusts dynamically based on changes in radiation. Under high irradiation, the PV panel generates maximum power, and both the EV and storage batteries receive energy. As the irradiation decreases, the grid supplies power to the EV battery to ensure continuous charging, compensating for the reduced PV output.
The simulation demonstrates the ability of the system to maintain a constant DC bus voltage, which remains stable despite fluctuations in radiation. It also highlights how the storage battery discharges to support the EV battery, ensuring that the vehicle is consistently charged even under varying solar conditions.
Conclusion
In summary, the integration of solar PV, energy storage, and a neural network-based energy management system provides a highly efficient solution for managing energy flow in an EV charging station. By dynamically adjusting the charging and discharging of the batteries, and intelligently managing the grid connection, the system ensures optimal performance even under changing environmental conditions. The use of advanced control algorithms such as MPPT and neural network-based management allows for improved reliability and efficiency, making it an effective solution for sustainable EV charging.
Comments