Grid Connected PV System with Neural Network MPPT
Explanation of the Simulation Model
This model is developed for a grid-connected PV system with ANN MPPT. The model consists of:
Main Grid: 154 MW, 34 kW.
Transformer: 34 kW to Volt transformer.
AC Load:
Load 1: 500 kW
Load 2: 30 kW
Both loads have a rating of 400V line-to-line RMS and 50 Hz.
Solar PV Array
The PV array in this model includes:
Panel Power: 414 watts per panel
Open Circuit Voltage: 85.3V
Voltage at Maximum Power Point: 72.9V
Short Circuit Current: 6.09A
Current at Maximum Power Point: 5.69A
Temperature Coefficients:
Voltage: -0.229%/°C
Current: 0.03%/°C
A total of 41 kW PV panels are used, arranged in series and parallel to match the grid requirements.
ANN MPPT Implementation
Data Collection and Training
For the neural network MPPT, we need to measure the irradiation and temperature. Based on these measurements, the reference voltage is generated.
Data Collection:
Short Circuit Current: 6.09A
Current at Maximum Power Point: 5.69A
Open Circuit Voltage: 85.3V
Voltage at Maximum Power Point: 72.9V
Temperature Coefficients:
Voltage: -0.229%/°C
Current: 0.03%/°C
Input and Output Data:
Input: Irradiation and temperature
Output: Reference voltage (Maximum Power Voltage)
Training the Neural Network:
Collected 1000 samples with varying temperature (15-35°C) and irradiation (0-1000 W/m²).
Used the standard solar PV equations to generate input data for the neural network.
Neural Network Training
Input Data: Irradiation and temperature
Target Data: Reference voltage
The neural network is trained to match the input data with the output reference voltage. The model's accuracy is validated with an R-value close to 1, indicating a good fit.
Simulation of the Grid Connected PV System
Constant Irradiation and Temperature
Initial Setup:
Irradiation: 1000 W/m²
Temperature: 25°C
Simulation Results:
Maximum Power from PV: ~41 kW
Load 1 Power: ~30 kW
Load 2 Power: ~10 kW
Grid Power: Remaining power supplied to the grid or received from the grid.
Changing Load Conditions
New Load Setup:
Load 1: 20 kW
Load 2: 10 kW
Total Load: 30 kW
Simulation Results:
PV Power: ~40 kW
Excess power (10 kW) supplied to the grid.
Varying Irradiation
Changing Irradiation:
Initial Irradiation: 1000 W/m²
Reduced to: 500 W/m² at 0.3 seconds
Simulation Results:
PV Power drops from ~40.65 kW to ~20.2 kW.
Grid compensates for the reduced power to meet the load requirements.
Explanation of Inverter Control
Inverter Control Logic:
Measures irradiation and temperature to provide the reference voltage.
DC link voltage is regulated based on this reference voltage.
Real power from the PV is sent to the grid while ensuring reactive power (IQ reference) is zero.
Control Voltage:
Generated in the form of VD and VQ.
Converted to ABC form for modulation.
Sinusoidal three-phase voltages are generated and multiplied with magnitude to control the inverter.
Conclusion
Using ANN MPPT, we effectively extract maximum power from the solar PV system. The system adjusts to varying irradiation conditions and load requirements, ensuring efficient power management between the PV system and the grid.
Commentaires