Failure Prediction for Large Anti-drone System Clusters
摘要
In this study, we aimed to predict the failure of a system that includes multiple anti-drone systems, namely radar and jammer devices. It is anticipated that predictive maintenance, which has been attempted in many areas before, in the anti-drone field will reduce the costs. In this study, we developed a feature selection that suits the requirements of the problem. Afterward, Support Vector Machines (SVM), and Multi-Layer Perceptron Neural Network (MLPNN) models are adapted to our problem. Created models are fed with one hand-crafted single device data set, one synthetic single device data set, and one synthetic multi-device data set and reported the success rates. It is observed that the results provide successful predictions that could used in the field.