Target Detection in synthetic aperture radar(SAR) image is critical for military and national defense. In this paper, we propose YOLOv4-Attention architecture which adds attention modules to YOLOv4 backbone architecture to complement the feature extraction ability for SAR target detection with high accuracy. For training and testing our framework, we present new SAR embedding datasets based on MSTAR SAR public datasets which are about poor environments for target detection such as various clutter, crowded objects, various object size, close to buildings, and weakness of signal-to-clutter ratio. Experiments show that our Attention YOLOv4 architecture outperforms original YOLOv4 architecture in SAR image target detection tasks in poor environments for target detection.

Different from the field of electro-optical(EO) image analysis, there has been less interest in similarity metrics between synthetic aperture radar(SAR) target images. A reliable and objective similarity analysis for SAR target images is expected to enable the verification of the SAR measurement process or provide the guidelines of target CAD modeling that can be used for simulating realistic SAR target images. For this purpose, this paper presents a similarity analysis method based on the siamese network that quantifies the subjective assessment through the distance learning of similar and dissimilar SAR target image pairs. The proposed method is applied to MSTAR SAR target images of slightly different depression angles and the resultant metrics are compared and analyzed with qualitative evaluation. Since the image similarity is somewhat related to recognition performance, the capacity of the proposed method for target recognition is further checked experimentally with the confusion matrix.

Synthetic dynamic infrared image generation from the given virtual environment is being the primary goal to simulate the output of the infra-red(IR) camera installed on a vehicle to evaluate the control algorithm for various search & reconnaissance missions. Due to the difficulty to obtain actual IR data in complex environments, Artificial intelligence(AI) has been used recently in the field of image data generation. In this paper, CycleGAN technique is applied to obtain a more realistic synthetic IR image. We added the Structural Similarity Index Measure(SSIM) loss function to the L1 loss function to generate a more realistic synthetic IR image when the CycleGAN image is generated. From the simulation, it is applicable to the guided-missile flight simulation tests by using the synthetic infrared image generated by the proposed technique.

Supercavitation is a phenomenon in which the cavity covers the entire underwater vehicle. The purpose of this paper is to compare and analyze the thermal effect on the cavity characteristics by changing the ventilated gas temperature through computational analysis. For this study, a homogeneous mixture model based on the 3D Navier-Stokes equation was used. As a phase change model, it is its own code considering both pressure change and temperature change. A dimensionless number T_m was presented to analyze the numerical results, and as the T_m increased, the cavity length increased by about 3.6 times and the cavity width by about 3.3 times at 393.15 K compared to room temperature. Analyzing these thermal effects, it was confirmed that rapid heat exchange and heat transfer between the gas phase and the liquid phase occurred at the location where the ventilated gas was sprayed, affecting the cavity characteristics. In addition, it can be confirmed that the initial cavity surface becomes unstable as the ventilated gas temperature increases, and it can be confirmed based on the numerical analysis results that the critical temperature at which the cavity surface becomes unstable is 373.15 K.

The purpose of the weapon live test during the phase of development is to provide essential information to decision makers that verify and validate the performance capabilities of weapons. Due to varying allocation and high variance of test resources with an increase in the weapon system's capability, the test environment can get highly complex, which can lead to a decrease in the reliability of test results. This issue can be addressed by applying a decision support system that provides various timely information collected by resources during the test process. The decision support system can be designed by applying the concept of digital twins, that are defined as digital replicas of components, systems and processes. This paper describes a design methodology of the decision support system that consists of digital models and service functions using digital twin architecture. A case study illustrates the feasibility of the proposed methodology in supporting the weapon live test process.

This study deals with an AI architecture model for collecting battlefield data using the tactical C4I system. Based on this model, the artificial staff can be utilized in tactical echelon. In the current structure of the Army's tactical C4I system, Servers are operated by brigade level and above and divided into an active and a standby server. In this C4I system structure, the AI server must also be installed in each unit and must be switched when the C4I server is switched. The tactical C4I system operates a server(DB) for each unit, so data matching is partially delayed or some data is not matched in the inter-working process between servers. To solve these issues, this study presents an operation concept so that all of alternate server can be integrated based on virtualization technology, which is used as an source data for AI Meta DB. In doing so, this study can provide criteria for the AI architectural model of the ground tactical echelon.

In this paper, an experimental study on the control of the plume deflection panel(PDP) with a support jack of the ground launching platform using a tilt angle sensor is described. To overcome the disadvantages of the existing PDP control without a support jack using a limit sensor, the control algorithm using a tilt angle sensor and the ferroelectric RAM in the Launcher Control Unit for recognizing the contact with the ground in an abnormal operation is proposed to control the PDP in various operation environments. Finally, the proposed algorithm can be well applied for not only heavy-load launching platforms but also any other similar systems.

Recently, the R&D of weapon systems has been strengthened in terms of economic cost management throughout the entire life cycle from performance. This study proposes the method for setting the optimal maintenance concept based on RAM-C in weapon system acquisition stage by calculating the operation & maintenance cost as well as reliability, availability, and maintainability. First, we design a simulation model for analysis of weapon system logistic supportability. In addition, information such as weapon system Part Breakdown Structure, operation & maintenance system, cost, and etc for simulation analysis, is applied. Based on the obtained simulation results, the optimal plan is selected among alternatives designed with various maintenance concepts through normalization and weight setting. It is expected to be of technical help in the application of RAM-C in the weapon system acquisition stage.

This research shows the test performance of a 6 kW-scale hybrid electric power system for the high endurance drone. The power system is composed of a two-stroke reciprocal engine, starter-generator and battery, and they are integrated as one power unit. The engine is designed to provide the house for holding the starter-generator at the end of a crankshaft in turn the engine and starter-generator can maintain the same speed during the operational period. In this way, the generated power is readily controlled by just manipulating an engine throttle movement. Moreover, the starter-generator can initiate an engine operation with an aid of battery power until the combustion process becomes stabilized. In consequence, integration mechanism between an engine and generator is simplified, which results in weight reduction achieved. The duty of back-up battery is to provide a starting power to generator via a system controller in addition to covering momentarily power shortage. Therefore, the electric power system is vindicated to provide 6 kW power through a ground test.

We proposed a design method for constant current pyro squib circuit. The current method using N MOSFET for the stability problem has a weakness of the current change, requiring a new design. This paper identified the problem with conventional squib circuit where the current is reduced by 25 % when maximum resistance is 3 ohms. Thus, we proposed a stable constant current driving circuit using P MOSFET and PNP BJT. We confirmed stable constant circuit operation through simulations and measurements of the proposed circuit design where the current did not change until the resistance reached 3 ohms.