PERFORMANCE ANALYSIS OF PMSG-BASED WIND TURBINE UNDER NORMAL, ABNORMAL GRID CONDITIONS AND CYBER ATTACKS

Abstract

The integration of Permanent Magnet Synchronous Generator (PMSG) wind turbine into the modern power grid demands strict the Low Voltage Ride Through (LVRT) compliance and robust cyber-physical security. Convectional Proportional-Integrator (PI) controllers exhibit limited transient response during deep voltage sags. To overcome these limitations, this dissertation proposed and implemented an Artificial Neural Network (ANN) based LVRT control strategy for Grid Side Converter (GSC). This dissertation also deeply studies the impact and analyses the severity of cyber-attacks such as Fault Data Injection (FDI), Denial of-Service (DoS), Replay and Control Parameter attacks on the system. Simulations in MATLAB/Simulink demonstrated that the data driven ANN controller effectively suppressed DC-link overvoltage and dynamically injects reactive power, extending the system’s voltage sag tolerance from 20% under conventional control to 40%. Finally, the computational feasibility, execution speed and real-world applicability of the proposed control architecture were successfully validated using an OPAL-RT real time digital simulator.