MATLAB Simulation of Grid connected PV with Hybrid Fuzzy Neural Network MPPT
MATLAB Simulation of Grid connected PV with Hybrid Fuzzy Neural Network MPPT
𝐈𝐧𝐭𝐫𝐨𝐝𝐮𝐜𝐭𝐢𝐨𝐧
The MATLAB Simulation of Grid connected PV with Hybrid Fuzzy Neural Network MPPT demonstrates a two-stage grid-connected photovoltaic system designed to extract maximum solar power under both uniform irradiance and partial shading conditions.
The model combines:
Two PV array groups
Hybrid Fuzzy Logic–Neural Network MPPT
High-gain DC–DC converter
Three-phase voltage source inverter
LCL grid filter
DC-link voltage regulation
dq-axis inverter current control
Grid synchronization using PLL
Real-power transfer to a three-phase utility grid
The hybrid MPPT structure combines the duty-cycle decisions obtained from fuzzy logic and neural-network-based variable-step tracking to improve PV operation when solar irradiance changes. Pasted text
𝐒𝐲𝐬𝐭𝐞𝐦 𝐎𝐯𝐞𝐫𝐯𝐢𝐞𝐰
The complete system follows a two-stage power conversion architecture:
PV Array → High-Gain DC–DC Converter → DC Link → Three-Phase Inverter → LCL Filter → Grid
The first conversion stage adjusts the PV operating point and increases the DC voltage. The second stage converts DC power into three-phase AC power suitable for grid connection.
Main System Components
Component | Function |
PV Array Group 1 | Operates under full irradiance |
PV Array Group 2 | Variable irradiance creates partial shading |
Bypass Diodes | Provide alternate current paths during shading |
Hybrid MPPT | Determines the converter duty cycle |
High-Gain Converter | Extracts PV power and raises DC voltage |
DC Link | Maintained around the specified reference |
Three-Phase Inverter | Converts DC power into AC power |
LCL Filter | Filters inverter switching components |
PLL | Synchronizes inverter control with the grid |
dq Controller | Controls active and reactive current components |
Grid | Receives generated PV power |
𝐏𝐕 𝐀𝐫𝐫𝐚𝐲 𝐂𝐨𝐧𝐟𝐢𝐠𝐮𝐫𝐚𝐭𝐢𝐨𝐧
The simulation uses two PV array groups connected in series.
Each PV group contains:
3 PV modules in series
1 parallel string
Identical PV module specifications
Bypass-diode arrangement for partial shading operation
PV Module and Array Parameters
Parameter | Value |
Individual PV module rated power | 349.59 W |
Module voltage | 43 V |
Module current | 8.13 A |
Series modules per group | 3 |
Parallel strings | 1 |
Approximate power per PV group | 1.048 kW |
Number of PV groups | 2 |
Standard irradiance | 1000 W/m² |
Temperature | 25°C |
Under standard operating conditions, the two PV groups together provide approximately 2 kW of photovoltaic power, as demonstrated in the simulation. Pasted text
𝐏𝐚𝐫𝐭𝐢𝐚𝐥 𝐒𝐡𝐚𝐝𝐢𝐧𝐠 𝐂𝐨𝐧𝐝𝐢𝐭𝐢𝐨𝐧
One important feature of this MATLAB/Simulink model is the evaluation of MPPT performance under changing irradiance conditions.
One PV group remains at standard irradiance, while the irradiance applied to the second group is varied to create partial shading.
Irradiance Profile
Operating Stage | Irradiance |
Initial condition | 1000 W/m² |
Shading level 1 | 800 W/m² |
Shading level 2 | 600 W/m² |
Shading level 3 | 400 W/m² |
Temperature | 25°C |
Irradiance change interval | Approximately 0.1 s |
As the irradiance decreases, the available PV power also decreases. This makes the operating condition useful for analyzing how the hybrid MPPT adapts to changing solar conditions. Pasted text
𝐇𝐲𝐛𝐫𝐢𝐝 𝐅𝐮𝐳𝐳𝐲 𝐍𝐞𝐮𝐫𝐚𝐥 𝐍𝐞𝐭𝐰𝐨𝐫𝐤 𝐌𝐏𝐏𝐓
The main control feature is the Hybrid Fuzzy Neural Network MPPT.
Instead of depending on only one tracking method, the model uses two MPPT branches:
Fuzzy Logic based variable-step MPPT
Neural Network based variable-step MPPT
The final switching command is developed by combining the outputs of these two control approaches.
Why Hybrid MPPT?
The hybrid structure is useful because partial shading produces a more complex PV power characteristic than uniform irradiation.
It helps the controller respond to:
Sudden solar irradiance changes
Different PV operating conditions
Partial shading effects
PV voltage variations
PV current variations
Required changes in converter duty cycle
𝐅𝐮𝐳𝐳𝐲 𝐋𝐨𝐠𝐢𝐜 𝐌𝐏𝐏𝐓
The fuzzy-based MPPT monitors the PV electrical quantities and generates an adaptive change in duty cycle.
Its control process includes:
Measuring PV voltage
Measuring PV current
Determining changes in the operating condition
Evaluating the PV power characteristic
Generating a variable duty-cycle step
Applying the updated duty cycle to the MPPT process
Unlike a fixed-step tracker, the variable-step approach allows the duty-cycle adjustment to change according to PV operating conditions.
This can provide:
Faster response during significant irradiance changes
Smaller duty-cycle corrections near the desired operating point
Better adaptation to changing PV conditions
𝐍𝐞𝐮𝐫𝐚𝐥 𝐍𝐞𝐭𝐰𝐨𝐫𝐤 𝐌𝐏𝐏𝐓
The neural-network branch receives the PV operating information and determines the required duty-cycle adjustment.
Neural Network Inputs
PV voltage
PV current
Neural Network Output
Duty-cycle change
This output is subsequently processed by the variable-step MPPT mechanism to generate a duty-cycle command.
The neural network therefore provides an intelligent mapping between the PV operating condition and the required converter control adjustment.
𝐇𝐲𝐛𝐫𝐢𝐝 𝐃𝐮𝐭𝐲 𝐂𝐲𝐜𝐥𝐞 𝐆𝐞𝐧𝐞𝐫𝐚𝐭𝐢𝐨𝐧
The fuzzy controller and neural network operate as separate MPPT branches.
The control sequence can be summarized as:
Measure PV voltage and current.
Send the measurements to the MPPT controllers.
Generate the fuzzy-based duty-cycle command.
Generate the neural-network-based duty-cycle command.
Combine the two duty-cycle outputs.
Obtain the final duty-cycle reference.
Compare the reference with a carrier waveform.
Generate converter switching pulses.
Control the high-gain DC–DC converter.
Extract maximum available PV power.
The model uses the average of the duty-cycle commands obtained from the fuzzy and neural-network branches before generating the converter switching pulses. Pasted text
𝐇𝐢𝐠𝐡-𝐆𝐚𝐢𝐧 𝐃𝐂–𝐃𝐂 𝐂𝐨𝐧𝐯𝐞𝐫𝐭𝐞𝐫
The PV system is connected to a high-gain DC–DC converter before the inverter stage.
The converter performs two important functions:
Controls the PV operating point according to the hybrid MPPT command
Provides the required DC voltage for the grid-side inverter
The described model uses an active-switch LC-network-based high-gain converter.
Converter Control Flow
PV Voltage & Current → Hybrid MPPT → Duty Cycle → PWM → High-Gain Converter
As solar irradiance varies, the MPPT modifies the converter duty cycle to continuously adapt the operating point of the photovoltaic source.
𝐆𝐫𝐢𝐝-𝐒𝐢𝐝𝐞 𝐈𝐧𝐯𝐞𝐫𝐭𝐞𝐫
The output of the high-gain converter is connected to a three-phase inverter.
The inverter is responsible for:
DC-to-AC conversion
DC-link voltage regulation
Grid current control
Active-power delivery
Reactive-current control
Synchronization with grid voltage
Grid-Side Parameters
Parameter | Value |
Grid voltage | 400 V |
Grid frequency | 50 Hz |
DC-link voltage reference | 700 V |
q-axis current reference | 0 |
Inverter type | Three-phase VSI |
Output filter | LCL filter |
𝐂𝐨𝐧𝐭𝐫𝐨𝐥 𝐒𝐭𝐫𝐚𝐭𝐞𝐠𝐲
The inverter uses a synchronous reference frame control strategy.
The major control sections are:
PLL-based grid synchronization
abc-to-dq transformation
DC-link voltage controller
d-axis current controller
q-axis current controller
Feed-forward decoupling
dq-to-abc conversion
PWM generation
𝐏𝐋𝐋-𝐁𝐚𝐬𝐞𝐝 𝐆𝐫𝐢𝐝 𝐒𝐲𝐧𝐜𝐡𝐫𝐨𝐧𝐢𝐳𝐚𝐭𝐢𝐨𝐧
Grid synchronization is necessary before transferring power from the inverter to the utility system.
The model uses grid voltage measurements to determine the grid phase information.
The synchronization section includes:
Three-phase grid-voltage measurement
Stationary-reference transformation
Rotating-reference transformation
q-axis voltage regulation
PI control
Grid angle generation
The resulting electrical angle is used by the dq transformations employed in the current-control loops.
𝐃𝐂-𝐋𝐢𝐧𝐤 𝐕𝐨𝐥𝐭𝐚𝐠𝐞 𝐂𝐨𝐧𝐭𝐫𝐨𝐥
A key requirement of the grid-connected system is maintaining a stable DC-link voltage.
The model uses a reference of:
𝐕DC Reference = 700 V
The measured DC-link voltage is compared with this reference.
The resulting error is processed through a PI controller to generate the d-axis current reference.
The d-axis current therefore controls the real-power exchange required to maintain the DC-link voltage.
𝐝-𝐀𝐱𝐢𝐬 𝐂𝐮𝐫𝐫𝐞𝐧𝐭 𝐂𝐨𝐧𝐭𝐫𝐨𝐥
The d-axis channel primarily regulates the active power flow.
The control sequence is:
Measure DC-link voltage
Compare with 700 V reference
Generate d-axis current reference
Measure inverter current
Convert inverter current to dq components
Compare measured d-axis current with its reference
Process the error through the PI controller
Apply feed-forward decoupling
Generate the d-axis inverter control command
As PV power changes, the grid current adjusts while the DC-link controller attempts to maintain the required voltage.
𝐪-𝐀𝐱𝐢𝐬 𝐂𝐮𝐫𝐫𝐞𝐧𝐭 𝐂𝐨𝐧𝐭𝐫𝐨𝐥
In the demonstrated configuration:
𝐈q Reference = 0
This operating mode is used so that the generated photovoltaic power is primarily delivered to the grid as real power.
The measured q-axis current is compared with zero, and the error is processed through the current controller.
This provides controlled grid-current operation without commanding additional reactive-current injection in the demonstrated case. Pasted text
𝐏𝐖𝐌 𝐆𝐞𝐧𝐞𝐫𝐚𝐭𝐢𝐨𝐧
After obtaining the d-axis and q-axis control quantities:
The control quantities are transformed back into three-phase signals.
Three sinusoidal control references are obtained.
These references are compared with a triangular carrier.
Six switching pulses are produced.
The pulses operate the six inverter switches.
This allows the three-phase inverter to generate controlled AC current for grid connection.
𝐖𝐨𝐫𝐤𝐢𝐧𝐠 𝐏𝐫𝐨𝐜𝐞𝐬𝐬
The overall simulation can be understood in a simple sequence:
Step 1 – Solar Power Generation
Two photovoltaic groups generate DC power.
Step 2 – Partial Shading Creation
The irradiance of one PV group is progressively reduced.
Step 3 – PV Measurement
PV voltage and current are measured continuously.
Step 4 – Hybrid MPPT
Fuzzy logic and neural-network MPPT controllers determine suitable duty-cycle adjustments.
Step 5 – High-Gain Conversion
The converter operates according to the hybrid MPPT switching command.
Step 6 – DC-Link Regulation
The inverter controller maintains the DC-link voltage around 700 V.
Step 7 – Grid Synchronization
The PLL tracks the phase of the three-phase grid.
Step 8 – dq Current Control
Active and reactive current components are controlled independently.
Step 9 – Inverter PWM
Six PWM pulses operate the three-phase inverter.
Step 10 – Grid Power Transfer
Available photovoltaic power is transferred to the 400 V, 50 Hz grid.
𝐒𝐢𝐦𝐮𝐥𝐚𝐭𝐢𝐨𝐧 𝐑𝐞𝐬𝐮𝐥𝐭𝐬
The simulation monitors several important variables:
PV power
Grid voltage
Grid current
DC-link voltage
Grid-side active power
Observed Behaviour
Operating Condition | Simulation Behaviour |
Standard irradiance | PV power is close to 2 kW |
Partial shading begins | PV power decreases |
Irradiance decreases | Grid-current magnitude decreases |
Changing PV generation | Power delivered to the grid varies |
DC-link control active | DC-link remains close to 700 V |
q-axis reference = 0 | Real-power-oriented operation is maintained |
The source demonstration shows the PV power dropping to approximately 1.75 kW during one of the partial-shading operating stages. Pasted text
𝐃𝐂-𝐋𝐢𝐧𝐤 𝐑𝐞𝐬𝐩𝐨𝐧𝐬𝐞
One of the most noticeable results is the DC-link voltage behaviour.
During startup:
A transient voltage rise occurs.
The controller quickly acts on the inverter current.
The DC-link voltage settles near its 700 V reference.
After settling, the DC-link remains approximately constant even when:
PV irradiance changes
PV output power decreases
Partial shading occurs
Grid-current magnitude varies
This demonstrates the role of the outer DC-link voltage-control loop in maintaining stable inverter operation.
𝐆𝐫𝐢𝐝 𝐕𝐨𝐥𝐭𝐚𝐠𝐞 𝐚𝐧𝐝 𝐂𝐮𝐫𝐫𝐞𝐧𝐭
The simulation also displays the three-phase grid voltage and current waveforms.
Grid Voltage
The three-phase voltage remains balanced and sinusoidal at the grid side.
Grid Current
The grid current changes according to available photovoltaic power.
For example:
Higher PV power → higher grid-current magnitude
Lower irradiance → reduced PV power
Reduced PV power → lower grid-current magnitude
Therefore, changes in solar energy are reflected mainly in the amount of current and real power supplied to the grid.
𝐊𝐞𝐲 𝐅𝐞𝐚𝐭𝐮𝐫𝐞𝐬
Hybrid Fuzzy–Neural Network MPPT
Variable-step maximum power point tracking
Partial shading simulation
Two PV array groups
Bypass-diode implementation
High-gain DC–DC converter
Two-stage grid-connected architecture
Three-phase inverter
LCL grid filter
700 V DC-link regulation
400 V, 50 Hz grid
dq-axis current control
PLL-based synchronization
Feed-forward decoupling
Active-power-oriented operation
PWM-based inverter switching
PV and grid performance monitoring
𝐖𝐡𝐚𝐭 𝐂𝐚𝐧 𝐁𝐞 𝐀𝐧𝐚𝐥𝐲𝐳𝐞𝐝?
This MATLAB/Simulink model is useful for studying:
PV characteristics under changing irradiance
Partial shading effects
Hybrid intelligent MPPT
Fuzzy logic control
Neural-network control
High-gain converter behaviour
DC-link voltage regulation
Grid synchronization
dq-reference-frame control
Active and reactive current control
Three-phase inverter switching
Grid-connected renewable-energy conversion
𝐀𝐩𝐩𝐥𝐢𝐜𝐚𝐭𝐢𝐨𝐧𝐬
The simulation is suitable for students, researchers, engineers, and educators studying areas such as:
Solar photovoltaic energy conversion
Grid-connected PV systems
Intelligent MPPT algorithms
Fuzzy logic control
Artificial neural networks
Power electronic converters
Renewable-energy integration
Three-phase inverter control
Smart-grid applications
Partial shading analysis
Advanced MATLAB/Simulink control development
𝐖𝐡𝐲 𝐓𝐡𝐢𝐬 𝐌𝐨𝐝𝐞𝐥 𝐢𝐬 𝐔𝐬𝐞𝐟𝐮𝐥
A conventional PV simulation under constant irradiance cannot fully demonstrate the challenges experienced by an actual solar system.
This model introduces varying solar irradiation and partial shading while combining:
PV Generation + Hybrid Intelligent MPPT + High-Gain Converter + DC-Link Control + Grid Inverter Control
This allows users to study the complete power-conversion chain rather than analyzing only the MPPT block independently.
𝐂𝐨𝐧𝐜𝐥𝐮𝐬𝐢𝐨𝐧
The MATLAB Simulation of Grid connected PV with Hybrid Fuzzy Neural Network MPPT provides a complete platform for studying photovoltaic power extraction and grid integration under changing environmental conditions.
The hybrid MPPT combines fuzzy logic and neural-network-based variable-step tracking, while the high-gain converter performs the first power-conversion stage. A three-phase inverter with PLL synchronization, dq current control, feed-forward decoupling, PWM control, and 700 V DC-link regulation transfers the available PV power to the 400 V, 50 Hz grid.
The simulation clearly demonstrates how PV power and grid current change during partial shading while the inverter control maintains the required DC-link voltage, making the model particularly useful for learning and research in solar PV, intelligent MPPT, power electronics, and grid-connected renewable-energy systems. Pasted text
Comments