Enhanced Optimization Model for Inverter Short Circuit Prediction Using Machine Learning Techniques

Abstract

Short circuits are common faults that occur in inverters, which can lead to device damage, safety hazards, and downtime. Early detection of short circuits can help prevent these issues and improve the reliability of inverters. Suggest a machine learning method in this research approach short circuit prediction in inverters. Collected data from various sensors installed in the inverter system, such as voltage, current, and temperature sensors, and used this data to train several machine learning models, such as the Multilayer Perceptron (MLP), Support Vector Machine (SVM), and K-Nearest Neighbor (KNN). Also utilized artificial intelligence algorithms such as Firefly Algorithm (FA) to optimize the model parameters. One could assess the effectiveness of the models by measuring their performance using different metrics such as accuracy, specificity, and convergence curve, and found that our proposed approach achieved high accuracy and robustness in predicting short circuits. Our results demonstrate the potential of using machine learning and artificial intelligence techniques for early detection of short circuits in inverters, which can contribute to improved system reliability and safety.