Optimal detection and classification of grid connected system using MSVM-FSO technique.

Area of Science:

• Renewable Energy Systems
• Electrical Engineering
• Artificial Intelligence in Power Systems

Background:

• Hybrid photovoltaic (PV) and wind turbine (WT) systems are crucial for renewable energy generation.
• Ensuring the reliability and power quality (PQ) of these hybrid systems requires robust fault diagnosis and detection (FDD).
• Existing FDD methods may suffer from complexity or lower efficiency in hybrid system applications.

Purpose of the Study:

• To propose a novel hybrid fault diagnosis and detection (FDD) method for hybrid PV-WT systems.
• To enhance the power quality (PQ) of hybrid systems through accurate fault identification.
• To develop a low-complexity FDD solution for improved system reliability.

Main Methods:

• A hybrid MSVM-FSO (Multiple Support Vector Machine-Firebug Swarm Optimization) method was developed.
• The MSVM approach was employed for detecting fault conditions in the grid-tied system.
• The FSO algorithm was utilized to categorize the types of faults occurring in the grid-connected system.
• Voltage and current data at system buses were analyzed to evaluate fault events.

Main Results:

• The proposed MSVM-FSO method achieved a high accuracy of 99.7% in fault diagnosis.
• The method demonstrated an efficiency of 98%, outperforming existing techniques.
• Performance evaluation was conducted using MATLAB simulations, comparing against established methods.

Conclusions:

• The MSVM-FSO hybrid method offers a highly accurate and efficient solution for fault diagnosis and detection in hybrid PV-WT systems.
• This approach contributes to improved power quality and system reliability in renewable energy integration.
• The proposed method presents a promising, low-complexity alternative for FDD in complex hybrid power systems.