A novel radar operating mode identification approach based on variational relevance vector machine with chaotic gravitational search optimization

【摘要】：Radar operating mode identification is one of the top priorities in the electronic countermeasure field for evaluating threat and behavior prediction.However,most of the existing identification methods usually extract features based on prior knowledge manually,which have the drawbacks of lagging inference time and unsatisfactory identification accuracy.An efficient identification approach for radar operating mode is therefore proposed without prior knowledge,where the variational relevance vector machine(VRVM) model of operating modes identification is therefore first presented based on the Bayesian probability framework.Then an improved chaotic gravity search algorithm(CGSA) is developed to increase the search breadth and effectively optimize the hyperparameters of VRVM.Finally,a novel filter method named Threshold-One-Versus-One(TOVO) is further proposed to screen the identification results to form the final radar operating mode identification model.The experimental results demonstrate that the proposed approach can accurately distinguish operating mode in real-time without any prior knowledge.