Smart Monitoring and Diagnostics for Fault Detection in Power Plant Equipment

In recent years, Fault Diagnosis and Identification (FDI) has grown significantly. Generally used methodology in these types of frameworks include model-based approaches, fault approaches, and pattern recognition techniques. However, it was essential to society's energy supply, therefore maintaining its safe and effective functioning is of utmost importance. This research used sophisticated Fault Detection and Diagnosis (FDD) methods to diagnose the problems in contemporary power plants in order to decrease maintenance shutdowns and expenses. Therefore, this research applied deep learning to image processing using efficient Long Short-Term Memory (LSTM)-based methods for object recognition and categorization. Using the observed frequency as input, the trained LSTM is utilized to detect power variations in real time. Control devices, such as synchronous generators and Energy Storage Systems (ESSs), was maintain a steady frequency by using the detected power variations. According to the results, the suggested approach demonstrated superior performance in terms of LG faults, with 0.427%, LLG faults of 0.800%, LGG faults of 0.186 %, and LL faults of 0.741%. Compared with existing methods namely Sparrow Search Algorithm (SSA) and Deep Long Short-Term Memory (LSTM) it shows that suggested methods are achieved better performances with less faults respectively.