🔋 Grid Tied PV Battery System in MATLAB – Simulation & Energy Management

🌞 Introduction to Grid Tied PV Battery Systems

A grid-tied PV battery system combines solar photovoltaic panels with a battery energy storage unit and a grid connection to ensure uninterrupted power supply. In this MATLAB/Simulink model, a SEPIC converter connects the PV array to the DC bus, while a neural network-based energy management system optimizes the power flow between PV, battery, and the grid.

📊 PV Panel Specifications

• Panel Power Rating: 115 W per panel

• Open Circuit Voltage (Voc): ~36.3 V

• Short Circuit Current (Isc): ~7.84 A

• Max Power Point (Vmp): ~29 V

• Max Power Point Current (Imp): ~7.35 A

• Array Configuration: 8 panels in series (≈ 1,705 W at 1,000 W/m²)

⚙️ System Architecture

The simulation uses the following components:

• PV Array → SEPIC Converter → DC Bus

• Battery connected directly to the DC Bus

• Inverter with LCL Filter connected to the AC Grid

• Control Methods:

  • PV side → Perturb & Observe MPPT

  • Inverter side → DQ Control Logic + Neural Network EMS

🧠 Neural Network Energy Management

The Energy Management System (EMS) uses battery SOC and PV power as inputs to a trained neural network in MATLAB.

• Generates a reference current (Iref)

• Determines whether to charge the battery, export to the grid, or import from the grid

• Supports customization with user-generated training data

🔄 Inverter Control with DQ Logic

1. Iref converted to sinusoidal form using grid voltage

2. αβ ↔ dq transformations for comparison with measured currents

3. PI controllers regulate inverter output

4. PWM signals generated for IGBT control

🌤 Simulation Results

• High Irradiance (1,000 W/m²):

  • PV generates ~1,700 W

  • Battery charges at ~300–424 W

  • Surplus (~1,000–1,250 W) sent to the grid

  • Inverter and grid voltages/currents remain in-phase ✅

• Low Irradiance (500 W/m²):

  • PV generates ~850 W

  • EMS imports power from the grid to charge the battery

  • Negative Iref indicates grid power import

📌 Key Takeaways

• Flexible EMS: Customizable neural network logic for different power strategies

• Efficient MPPT: P&O ensures maximum PV power extraction

• Bidirectional Operation: Supports both export and import of power

• Realistic Modeling: Includes irradiance variation scenarios