Automatic Target Recognition (ATR) is an important capability for defense application. ATR removes the human operator from the process of target acquisition and classification, reducing the reaction time to possible threats and can be used to gun target engagement. This paper presents one technique used to solve the automatic target recognition problem in Synthetic Aperture Radars (SAR) images, that is independent of target pose in the images. The classification is performed by a combination of three different classifiers the Minimum Distance Classifier (MDC), the Quadratic Gaussian Classifier (QGC) and a Multilayer Perceptron (MLP) neural network, using a voting architecture.

No abstract

No abstract

The ability of multiple manned and unmanned aircraft systems to cooperatively engage and disable an aerial threat plays a decisive role in modern warfare scenarios. In this paper, we apply key methods to enable the so-called cooperative threat engagement capability among multiple networked agents, e.g., a swarm of drones, with combat and communication capabilities. In particular, this research combines AI-based decision-making and control techniques for a swarm of loyal wingman drones to coordinate efficient defense actions in a cooperative and autonomous manner. We apply these concepts in a defense scenario that is modeled to analyze the loyal wingman concept, which we consider an interesting testbed for cooperative decision-making and low-level control techniques. The investigated methods were implemented in a realistic 3D UAV simulator for demonstration and evaluation. INDEX TERMSCooperative engagement capability, loyal wingman UAV, decision making, manned-unmanned teaming, sliding mode control.

Tracking agile aircraft under high accelerations generally demands sophisticated models for determining trajectories with desirable accuracy. Often this raises complexity of the estimation algorithm as it gives rise to more elaborated methods for both taking model nonlinearities into account and handling a greater number of state variables that describe the model. The approach of this work recalls a 3D model based on flight dynamics of a point of mass for which augmentation to the Extended Kalman-Bucy filter (EKBF) is proposed. Two methods of augmentation to the EKBF filter are studied: (i) use of second-order terms to approximate the model according to the Daum's theory; (ii) deployment of a neural network coupled to the filter for compensation of modeling and calculation errors. The evaluation of the filters performance is accomplished by measuring nonlinearities, biasness, accuracy and robustness. The designed filters are suitably accurate and robust for tracking targets in air combat scenario.