MATLAB Implementation of Microgrid Dynamics with P&O, PSO and ANFIS MPPT

Microgrids are becoming vital in modern energy systems, integrating renewable energy sources and storage technologies to ensure stability and sustainability. In this post, we explore the MATLAB simulation of a microgrid system that incorporates solar photovoltaic (PV) and battery energy storage, highlighting the use of different Maximum Power Point Tracking (MPPT) techniques—Perturb & Observe (P&O), Particle Swarm Optimization (PSO), and PSO with ANFIS (Adaptive Neuro-Fuzzy Inference System).

Overview of the Microgrid System

The simulated microgrid includes:

A 1 MW PV array, which forms the main power generation source.
A 1 MW / 1 MWh battery energy storage system to support load balancing and provide backup.
Grid interconnection through voltage step-up stages and power inverters.

The PV array uses 355 W panels, arranged with 15 in series and 190 parallel strings, capable of generating around 1.01 MW at standard test conditions (STC), i.e., 1000 W/m² irradiation and 25°C temperature.

MPPT Techniques in Use

To maximize power extraction from the PV system, three different MPPT methods are implemented:

1. P&O (Perturb and Observe) MPPT

Calculates power change and voltage change to adjust the duty cycle.
Controls the boost converter to regulate output power.
Though simple and widely used, it can suffer from oscillations near the MPP and slower convergence under rapidly changing environmental conditions.

2. PSO (Particle Swarm Optimization) with ANFIS

Uses PSO to train an ANFIS model to predict the optimal duty cycle.
Inputs: Irradiation and temperature; Output: Optimal voltage reference.
After training with synthetic irradiance and temperature data, the ANFIS model is embedded within the MPPT control block.
Offers a faster and more intelligent tracking method, especially under varying weather conditions.

3. PSO with Sliding Mode Control

Combines PSO for voltage reference generation with sliding mode control for robust performance.
Improves response time and reduces steady-state error.
Particularly effective in noisy environments and sudden load changes.

Inverter and Control Strategy

PV Side Control

Utilizes a three-level neutral point clamped inverter.
Implements dq0 transformation and PI control for voltage regulation.
Real power is prioritized (IQ = 0), ensuring minimal reactive power injection.
Converts Vdq reference back to ABC form for inverter modulation.

Battery System Integration

A two-level inverter is used without a converter.
Supports both charging and discharging modes depending on system demands.
Control logic governs operation as either inverter or rectifier.
Rated at 922 V and 1 MWh capacity, with an initial State of Charge (SOC) of 50%.

Grid and Load Interface

Voltage at the Point of Common Coupling (PCC) is maintained at 600 V, stepped up to 25 kV, and then connected to a 120 kV grid.
Local loads are included to simulate realistic demand-side variations.
The entire system operates at 60 Hz, and various performance metrics are monitored, including:
- PV power
- Battery power
- Reactive power
- PCC voltage and frequency
- Currents from different subsystems

Simulation and Results Comparison

The system is simulated under all three MPPT techniques, and results are compared based on key performance indicators like power tracking accuracy, speed, and response to environmental variations.

Observations:

P&O MPPT showed slower response and slight instability in achieving the maximum power point.
PSO MPPT provided better accuracy and faster convergence.
PSO with ANFIS outperformed both with superior tracking efficiency, minimal oscillations, and better adaptability to rapid environmental changes.

The simulation demonstrated that AI-driven MPPT algorithms, particularly PSO-ANFIS, can greatly enhance the performance and reliability of renewable-integrated microgrid systems.

Conclusion

This MATLAB-based simulation highlights the evolving role of intelligent MPPT techniques in optimizing renewable energy systems. As demonstrated, advanced methods like PSO with ANFIS provide a significant edge over traditional techniques, making them ideal for next-generation microgrids.

Stay tuned for more tutorials and deep-dives into renewable energy simulations—and don't forget to subscribe to our channel for upcoming updates and implementations.