Hybrid Neural Network Incremental Conductance MPPT

Hybrid Neural Network Incremental Conductance MPPT

𝐈𝐧𝐭𝐫𝐨𝐝𝐮𝐜𝐭𝐢𝐨𝐧

𝐇𝐲𝐛𝐫𝐢𝐝 𝐍𝐞𝐮𝐫𝐚𝐥 𝐍𝐞𝐭𝐰𝐨𝐫𝐤 𝐈𝐧𝐜𝐫𝐞𝐦𝐞𝐧𝐭𝐚𝐥 𝐂𝐨𝐧𝐝𝐮𝐜𝐭𝐚𝐧𝐜𝐞 𝐌𝐏𝐏𝐓 in MATLAB/Simulink is an intelligent solar energy control approach that combines a Neural Network MPPT with the conventional Incremental Conductance MPPT technique. The hybrid controller is used to extract maximum available power from a photovoltaic array while operating a boost converter, battery energy storage system, bidirectional DC–DC converter, and grid-connected inverter.

Hybrid Neural Network Incremental conductance MPPT in MATLAB

₹10,000.00₹5,000.00

Buy Now

The system demonstrates how solar power can be efficiently extracted, stored in a battery, supplied to a load, and exchanged with a single-phase grid under changing irradiation conditions.

Solar photovoltaic panels have nonlinear electrical characteristics. Their output power changes mainly with:

• Solar irradiation

• PV cell temperature

• Connected load condition

• PV operating voltage

• Environmental variations

When a PV panel is connected directly to a changing load, the operating point of the panel also changes. Therefore, the panel may not continuously operate at its Maximum Power Point (MPP).

A Maximum Power Point Tracking controller is used to identify the best operating point of the PV array and adjust the converter accordingly.

In this system, two MPPT methods are combined:

• Incremental Conductance MPPT

• Neural Network MPPT

The outputs of both methods are processed to generate the final duty cycle for the boost converter.

𝐖𝐡𝐲 𝐈𝐬 𝐚 𝐁𝐨𝐨𝐬𝐭 𝐂𝐨𝐧𝐯𝐞𝐫𝐭𝐞𝐫 𝐔𝐬𝐞𝐝?

A DC–DC converter is required between the PV array and DC bus to perform MPPT control.

Depending on the application, the converter can be:

• Buck converter

• Boost converter

• Buck–boost converter

This model uses a boost converter.

The boost converter provides two important functions:

• It increases the PV-side voltage to the required DC-bus level.

• It changes the effective electrical operating point seen by the PV array.

By controlling the IGBT duty cycle, the MPPT controller can move the PV operating point toward maximum power.

𝐒𝐲𝐬𝐭𝐞𝐦 𝐎𝐯𝐞𝐫𝐯𝐢𝐞𝐰

The complete MATLAB/Simulink system contains:

1. Solar PV array

2. PV voltage and current measurement

3. Incremental Conductance MPPT

4. Neural Network MPPT

5. Hybrid MPPT duty-cycle generation

6. Boost converter

7. DC bus

8. Battery energy storage system

9. Bidirectional DC–DC converter

10. Single-phase full-bridge inverter

11. LC output filter

12. Load

13. Single-phase utility grid

14. PLL-based synchronization

15. Current control

16. PWM generation

The basic power-flow structure is:

PV Array → Hybrid MPPT → Boost Converter → DC Bus → Battery / Inverter → Load and Grid

𝐏𝐕 𝐀𝐫𝐫𝐚𝐲 𝐏𝐚𝐫𝐚𝐦𝐞𝐭𝐞𝐫𝐬

The simulation uses a 250 W photovoltaic module.

Parameter	Value
Single PV module power	250 W
Voltage at maximum power point	30.7 V
Current at maximum power point	8.05 A
Parallel strings	1
Modules connected in series	4
Approximate PV array rating	1000 W
Standard irradiation	1000 W/m²
Standard temperature	25 °C

Four 250 W panels connected in series provide an approximately 1 kW solar PV array under standard test conditions.

𝐏𝐕 𝐂𝐡𝐚𝐫𝐚𝐜𝐭𝐞𝐫𝐢𝐬𝐭𝐢𝐜𝐬

PV output is highly dependent on solar irradiation.

The PV model can be tested at several irradiation levels while maintaining the temperature around 25 °C.

Typical operating points observed from the PV characteristics are:

Irradiation	Approx. Maximum Power
1000 W/m²	≈ 1000 W
800 W/m²	≈ 799.6 W
600 W/m²	≈ 598.5 W
400 W/m²	≈ 395 W

The PV current also decreases with irradiation.

Irradiation	Approx. Current Near MPP
800 W/m²	≈ 6.5 A
600 W/m²	≈ 4.9 A
400 W/m²	≈ 3.23 A

These characteristics clearly show why MPPT is necessary for photovoltaic systems.

𝐈𝐧𝐜𝐫𝐞𝐦𝐞𝐧𝐭𝐚𝐥 𝐂𝐨𝐧𝐝𝐮𝐜𝐭𝐚𝐧𝐜𝐞 𝐌𝐏𝐏𝐓

The Incremental Conductance algorithm continuously measures:

• PV voltage

• PV current

• Previous PV voltage

• Previous PV current

• Present PV power

• Previous PV power

Based on changes in voltage and current, the controller decides whether the PV operating point is:

• At the maximum power point

• On one side of the maximum power point

• On the other side of the maximum power point

The duty cycle is then increased, decreased, or maintained.

Incremental Conductance controller settings

Parameter	Value
Initial duty cycle	0.42
Minimum duty cycle	0.10
Maximum duty cycle	0.85
Duty-cycle step size	0.00005

The duty-cycle limits prevent the boost converter from operating outside the required control range.

𝐈𝐧𝐜𝐫𝐞𝐦𝐞𝐧𝐭𝐚𝐥 𝐂𝐨𝐧𝐝𝐮𝐜𝐭𝐚𝐧𝐜𝐞 𝐖𝐨𝐫𝐤𝐢𝐧𝐠 𝐏𝐫𝐨𝐜𝐞𝐬𝐬

The controller follows a simple sequence:

• Measure present PV voltage and current.

• Calculate the variation from the previous sampling instant.

• Determine the location of the operating point relative to MPP.

• Maintain the duty cycle when the system is close to MPP.

• Increase or decrease the duty cycle when correction is required.

• Limit the duty cycle between the specified minimum and maximum values.

• Store the present values for the next sampling instant.

• Repeat the process continuously.

This method provides good MPPT performance without requiring irradiation and temperature sensors.

𝐍𝐞𝐮𝐫𝐚𝐥 𝐍𝐞𝐭𝐰𝐨𝐫𝐤 𝐌𝐏𝐏𝐓

The second controller is based on an artificial neural network.

The Neural Network MPPT uses environmental information such as:

• Solar irradiation

• PV temperature

The neural network is trained to estimate the appropriate PV voltage corresponding to the maximum power operating point.

The estimated maximum-power voltage is compared with the actual PV voltage.

The resulting voltage error is processed through a controller to generate a duty-cycle command.

Main advantage

The neural network can learn the nonlinear relationship between:

• Irradiation

• Temperature

• Maximum-power voltage

This makes it suitable for rapidly changing PV operating conditions.

𝐇𝐲𝐛𝐫𝐢𝐝 𝐍𝐞𝐮𝐫𝐚𝐥 𝐍𝐞𝐭𝐰𝐨𝐫𝐤–𝐈𝐍𝐂 𝐌𝐏𝐏𝐓

Instead of using only one MPPT controller, the proposed model combines:

Neural Network MPPT + Incremental Conductance MPPT

The hybrid block receives:

• Duty-cycle information from the INC controller

• Duty-cycle/control information from the neural-network-based controller

The combined output is supplied to the PWM generator.

The PWM generator then produces the switching pulse for the boost-converter IGBT.

Why use hybrid MPPT?

The hybrid technique aims to combine:

• The adaptive capability of a Neural Network

• The online tracking capability of Incremental Conductance

• Improved response during environmental changes

• Accurate tracking near the maximum power point

• Reduced dependence on a single control technique

𝐁𝐨𝐨𝐬𝐭 𝐂𝐨𝐧𝐯𝐞𝐫𝐭𝐞𝐫 𝐂𝐨𝐧𝐭𝐫𝐨𝐥

The boost converter is placed between the PV array and DC bus.

The switching device is controlled using the final hybrid MPPT duty cycle.

Its main responsibilities are:

• Extract maximum available PV power

• Regulate PV operating voltage

• Transfer PV energy to the DC bus

• Support the required DC-link voltage

• Interface the PV array with the remaining energy system

𝐁𝐚𝐭𝐭𝐞𝐫𝐲 𝐄𝐧𝐞𝐫𝐠𝐲 𝐒𝐭𝐨𝐫𝐚𝐠𝐞

A battery is connected to the DC bus through a bidirectional DC–DC converter.

The battery operates according to the difference between available solar power and required power.

Main operating conditions

PV Condition	Battery Operation
PV power greater than required power	Battery charging
PV power approximately equal to demand	Battery nearly idle
PV power lower than required power	Battery discharging

The battery therefore improves power continuity during variations in solar irradiation.

𝐁𝐢𝐝𝐢𝐫𝐞𝐜𝐭𝐢𝐨𝐧𝐚𝐥 𝐃𝐂–𝐃𝐂 𝐂𝐨𝐧𝐯𝐞𝐫𝐭𝐞𝐫

The battery voltage is approximately 48 V, while the DC bus is maintained around 220 V.

Parameter	Value
Battery-side voltage	≈ 48 V
DC-bus reference	220 V
Converter type	Bidirectional DC–DC

The converter can operate in two directions.

Battery charging mode

When PV power is higher than the required system power:

• Excess PV energy becomes available.

• The bidirectional converter transfers energy toward the battery.

• The converter operates mainly in step-down mode.

• Battery SOC increases.

Battery discharging mode

When PV power becomes insufficient:

• The battery supplies the missing power.

• Battery-side voltage is converted to the higher DC-bus level.

• The converter operates in step-up mode.

• Battery SOC decreases.

𝐄𝐱𝐚𝐦𝐩𝐥𝐞 𝐨𝐟 𝐏𝐨𝐰𝐞𝐫 𝐒𝐡𝐚𝐫𝐢𝐧𝐠

Consider a load demand of approximately 600 W.

PV Power	Load Power	Approx. Battery Action
1000 W	600 W	Remaining ≈400 W available for charging
600 W	600 W	Battery approximately idle
400 W	600 W	Battery supplies approximately 200 W

This demonstrates automatic power balancing between solar generation, battery storage, and load demand.

𝐃𝐂 𝐁𝐮𝐬 𝐕𝐨𝐥𝐭𝐚𝐠𝐞 𝐂𝐨𝐧𝐭𝐫𝐨𝐥

The DC bus is regulated at approximately 220 V.

The control process is:

• Measure the actual DC-bus voltage.

• Compare it with the 220 V reference.

• Process the voltage error using a PI controller.

• Generate the required converter duty cycle.

• Produce complementary switching signals for the bidirectional converter.

Maintaining a stable DC bus is essential because both the PV boost converter and grid inverter interact through this common DC link.

𝐆𝐫𝐢𝐝-𝐂𝐨𝐧𝐧𝐞𝐜𝐭𝐞𝐝 𝐈𝐧𝐯𝐞𝐫𝐭𝐞𝐫

After the DC bus, a single-phase full-bridge inverter converts DC power into AC power.

The inverter output is connected through an LC filter before connection to the load and utility grid.

Grid-side specifications

Parameter	Value
Grid configuration	Single phase
Grid voltage	110 V RMS
Grid frequency	50 Hz
DC bus	220 V
Local load	Approximately 200 W
Inverter type	Full bridge

𝐋𝐂 𝐅𝐢𝐥𝐭𝐞𝐫

The inverter switching output contains high-frequency components.

An LC filter is therefore used to:

• Reduce switching harmonics

• Improve inverter output waveform

• Produce a near-sinusoidal voltage

• Reduce current ripple

• Improve grid-interface power quality

𝐏𝐋𝐋-𝐁𝐚𝐬𝐞𝐝 𝐆𝐫𝐢𝐝 𝐒𝐲𝐧𝐜𝐡𝐫𝐨𝐧𝐢𝐳𝐚𝐭𝐢𝐨𝐧

A Phase-Locked Loop is used to obtain synchronization signals from the AC voltage.

The PLL generates the required:

• Sine component

• Cosine component

• Grid phase information

These signals are used for reference-current generation and current control.

Accurate synchronization ensures that the inverter current follows the required grid phase.

𝐈𝐧𝐯𝐞𝐫𝐭𝐞𝐫 𝐂𝐮𝐫𝐫𝐞𝐧𝐭 𝐂𝐨𝐧𝐭𝐫𝐨𝐥

The inverter current controller determines how much current is exchanged between the PV–battery system and the grid.

The control structure includes:

• Measured inverter current

• Current reference

• Sine/cosine signals from the PLL

• Reference transformation

• Actual-current transformation

• PI current controllers

• Voltage reference generation

• PWM generation

The generated reference signal is finally used to control the four switching devices of the full-bridge inverter.

𝐂𝐮𝐫𝐫𝐞𝐧𝐭 𝐑𝐞𝐟𝐞𝐫𝐞𝐧𝐜𝐞 𝐂𝐨𝐧𝐭𝐫𝐨𝐥

The current-reference value determines the contribution of the solar PV and battery system toward the AC side.

Different reference-current values were tested.

Current Reference	Observed Effect
0.5	Lower inverter current contribution
1.0	Normal operating condition
2.0	Higher inverter current contribution

Changing the current reference changes the power-sharing relationship among:

• PV source

• Battery

• Local load

• Grid

𝐒𝐢𝐦𝐮𝐥𝐚𝐭𝐢𝐨𝐧 𝐑𝐞𝐬𝐮𝐥𝐭𝐬

Two important operating cases can be investigated.

Case 1 – Constant irradiation with changing current reference

The irradiation is maintained at approximately 1000 W/m², while the inverter current reference is changed.

Observed behavior includes:

• PV operates close to maximum available power.

• DC bus remains around 220 V.

• Inverter current changes according to its reference.

• Grid current changes according to local inverter contribution.

• Battery charging current changes as the AC-side power demand changes.

This test demonstrates controllable power exchange with the grid.

Case 2 – Variable solar irradiation

The current reference is kept constant while irradiation is changed.

The test sequence is approximately:

Time Region	Irradiation
Initial condition	1000 W/m²
After first change	500 W/m²
After second change	300 W/m²

This allows the response of the hybrid MPPT and battery-management system to be observed.

𝐑𝐞𝐬𝐮𝐥𝐭𝐬 𝐚𝐭 𝟏𝟎𝟎𝟎 𝐖/𝐦²

At high irradiation:

• PV produces high power.

• Hybrid MPPT extracts power close to the available maximum.

• DC-bus voltage remains close to 220 V.

• Excess solar power is transferred to the battery.

• Battery operates in charging mode.

• Battery SOC increases.

• Required inverter current is maintained.

𝐑𝐞𝐬𝐮𝐥𝐭𝐬 𝐚𝐭 𝟓𝟎𝟎 𝐖/𝐦²

When irradiation decreases to approximately 500 W/m²:

• PV output power decreases.

• Boost-converter power decreases accordingly.

• DC-bus voltage remains regulated around 220 V.

• Battery charging current reduces significantly.

• Battery can approach an idle operating condition.

• Grid/inverter current control continues operating.

This shows that the battery automatically responds to reduced PV generation.

𝐑𝐞𝐬𝐮𝐥𝐭𝐬 𝐚𝐭 𝟑𝟎𝟎 𝐖/𝐦²

When irradiation decreases further to approximately 300 W/m²:

• PV output power decreases considerably.

• PV generation becomes insufficient for the required power.

• Battery current changes toward the discharge direction.

• Battery starts supporting the DC bus.

• Battery SOC begins decreasing.

• DC-bus voltage remains approximately regulated.

• The load continues receiving the required power.

This demonstrates the energy-storage system's ability to compensate for solar-generation variations.

𝐎𝐩𝐞𝐫𝐚𝐭𝐢𝐧𝐠 𝐌𝐨𝐝𝐞 𝐒𝐮𝐦𝐦𝐚𝐫𝐲

Solar Condition	PV Output	Battery Mode	SOC Trend
High irradiation	High	Charging	Increasing
Medium irradiation	Moderate	Idle/light charging	Nearly constant
Low irradiation	Low	Discharging	Decreasing

𝐊𝐞𝐲 𝐅𝐞𝐚𝐭𝐮𝐫𝐞𝐬

• Hybrid Neural Network–INC MPPT

• Approximately 1 kW PV array

• Neural-network-based MPP voltage estimation

• Incremental Conductance online tracking

• Controlled boost converter

• 220 V regulated DC bus

• 48 V battery energy storage

• Bidirectional battery converter

• Automatic battery charge/discharge operation

• Single-phase grid connection

• 110 V, 50 Hz AC system

• Full-bridge inverter

• LC output filtering

• PLL-based grid synchronization

• Current-reference-based inverter control

• Variable solar irradiation testing

• PV power, voltage and current monitoring

• Battery voltage, current and SOC monitoring

• Grid and inverter waveform analysis

𝐀𝐝𝐯𝐚𝐧𝐭𝐚𝐠𝐞𝐬 𝐨𝐟 𝐭𝐡𝐞 𝐇𝐲𝐛𝐫𝐢𝐝 𝐌𝐏𝐏𝐓 𝐀𝐩𝐩𝐫𝐨𝐚𝐜𝐡

The hybrid controller provides several useful characteristics:

• Better handling of nonlinear PV behavior

• Fast response to changes in irradiation

• Continuous online correction through Incremental Conductance

• Intelligent MPP estimation using a neural network

• Improved solar-energy utilization

• Coordinated operation with battery storage

• Stable DC-bus operation

• Suitable interface for grid-connected PV systems

• Useful platform for studying intelligent renewable-energy control

𝐀𝐩𝐩𝐥𝐢𝐜𝐚𝐭𝐢𝐨𝐧𝐬

This MATLAB/Simulink model is useful for studying:

• Solar PV maximum power point tracking

• Intelligent MPPT algorithms

• Artificial Neural Network control

• Incremental Conductance MPPT

• Hybrid MPPT techniques

• Solar battery energy management

• Bidirectional DC–DC converters

• Grid-connected solar PV systems

• Renewable energy integration

• DC-bus voltage regulation

• Single-phase inverter control

• PLL-based synchronization

• Battery charge/discharge control

• Smart energy management

• Power-electronics control systems

𝐖𝐡𝐚𝐭 𝐂𝐚𝐧 𝐁𝐞 𝐌𝐨𝐧𝐢𝐭𝐨𝐫𝐞𝐝 𝐢𝐧 𝐭𝐡𝐞 𝐒𝐢𝐦𝐮𝐥𝐚𝐭𝐢𝐨𝐧?

Important simulation signals include:

• PV voltage

• PV current

• PV power

• Boost-converter current

• Boost-converter power

• DC-bus voltage

• Battery voltage

• Battery current

• Battery SOC

• Inverter voltage

• Inverter current

• Grid current

• Load power

• MPPT duty cycle

• Current reference

Monitoring these parameters helps explain the complete power flow of the PV–battery–grid system.

𝐖𝐡𝐲 𝐓𝐡𝐢𝐬 𝐌𝐨𝐝𝐞𝐥 𝐈𝐬 𝐔𝐬𝐞𝐟𝐮𝐥

This simulation combines several important renewable-energy concepts in one model:

PV generation + intelligent MPPT + boost conversion + battery storage + DC-link regulation + inverter control + grid integration

It provides a clear way to understand how changes in solar irradiation affect the entire system rather than studying the PV converter alone.

Students, researchers, and engineers can observe how MPPT, battery storage, and grid control work together in real time.

𝐂𝐨𝐧𝐜𝐥𝐮𝐬𝐢𝐨𝐧

The Hybrid Neural Network Incremental Conductance MPPT system provides an effective method for extracting maximum available solar power under changing operating conditions.

The Neural Network estimates the appropriate maximum-power operating point using environmental information, while the Incremental Conductance controller provides continuous online correction. Their combined control signal operates the boost converter to improve PV energy extraction.

The battery and bidirectional DC–DC converter provide additional energy-management capability. During high PV generation, excess energy charges the battery. When solar generation becomes insufficient, the battery discharges and supports the DC bus.

The regulated 220 V DC bus, single-phase inverter, LC filter, PLL synchronization, and current controller enable controlled integration with the 110 V, 50 Hz grid.

Overall, the MATLAB/Simulink model provides a complete platform for understanding hybrid MPPT control, PV power conversion, battery energy management, and grid-connected renewable-energy operation.