AC series arc fault detection with stacked autoencoders
Abstract
Series arc fault can occur in domestic electrical network and lead to fire accident. Many conventional detection algorithms based on time or frequency analysis have been published. Recently, machine learning technics have been adapted to the arc fault detection task and give promising results. However, the use of machine learning is frequently limited to classification part. Manual feature extraction part which requires time and effort is always needed before obtain the optimize features. In this paper we used stacked autoencoder in order to replace feature extraction part. The method presented can distinguish normal state and series arc fault with good accuracies.