This paper investigates the air–ground cooperative time-varying formation-tracking control problem of a heterogeneous cluster system composed of an unmanned ground vehicle (UGV) and an unmanned aerial vehicle (UAV). Initially, the structure of the UAV–UGV formation-control system is analyzed from the perspective of a cooperative combat system. Next, based on the motion relationship between the UAV–UGV in a relative coordinate system, the relative motion model between them is established, which can clearly reveal the physical meaning of the relative motion process in the UAV–UGV system. Then, under the premise that the control system of the UAG is closed-loop stable, the motion state of the UGV is modeled as an input perturbation. Finally, using a linear quadratic optimal control theory, a UAV–UGV formation-maintenance controller is designed to track the reference trajectory of the UGV based on the UAV–UGV relative motion model. The simulation results demonstrate that the proposed controller can overcome input perturbations, model-constant perturbations, and linearization biases. Moreover, it can achieve fast and stable adjustment and maintenance control of the desired UAV–UGV formation proposed by the cooperative combat mission planning system. Full article

With the improvement of the living standards of the residents, it is a very important and challenging task to continuously improve the accuracy of PM2.5 (particulate matter less than 2.5 $\mathsf{\mu }$m in diameter) prediction. Deep learning-based networks, such as LSTM and CNN, have achieved good performance in recent years. However, these methods require sufficient data to train the model. The performance of these methods is limited for the sites where the data is lacking, such as the newly constructed monitoring sites. To deal with this problem, an improved deep learning model based on the hybrid transfer learning strategy is proposed for predicting PM2.5 concentration in this paper. In the proposed model, the maximum mean discrepancy (MMD) is used to select which station in the source domain is most suitable for migration to the target domain. An improved dual-stage two-phase (DSTP) model is used to extract the spatial–temporal features of the source domain and the target domain. Then the domain adversarial neural network (DANN) is used to find the domain invariant features between the source and target domains by domain adaptation. Thus, the model trained by source domain site data can be used to assist the prediction of the target site without degradation of the prediction performance due to domain drift. At last, some experiments are conducted. The experimental results show that the proposed model can effectively improve the accuracy of the PM2.5 prediction at the sites lacking data, and the proposed model outperforms most of the latest models. Full article

Essential oils (EOs) have a long tradition of use in the medical and cosmetic fields based on their versatile properties, including fungicidal, antiparasitic, and bactericidal effects. Nowadays, with the development of industry and electronics, EOs are increasingly being used in the agricultural and food industries; health industries, including pharmacy and dental medicine; and as cosmetic enhancements. The purpose of this study is to develop a compact and portable platform for the detection of EO type and the concentration levels using knitted silver threads. The method is based on measuring the variation in values of the electrical parameters of the silver threads using electrochemical impedance spectroscopy (EIS). The impedance of the solutions applied on the testing platform was measured in the frequency range from 1 Hz to 200 kHz. The platform was tested using three types of essential oils: tea tree; clary sage; and cinnamon bark oil. Increasing the concentration of essential oils resulted in increasing the electrical resistance of the platform, decreasing the capacitance, and consequently increasing the impedance. The proposed cost-effective platform can be used for the fast determination of the type and quality of essential oils. Full article

The cost assessment of bridge maintenance is a difficult topic to study, but it is critical for a bridge life cycle cost analysis. The maintenance costs sample database was established in this study according to actual engineering data, and a bridge maintenance cost prediction model was developed using a fully-connected artificial neural network (ANN) and convolutional neural network (CNN), respectively. First, eight main factors affecting maintenance costs were evaluated based on the random forest method, and the evaluation results were verified by an exploratory data analysis. The original data were then screened based on the isolation forest principle, and the recent gross domestic product (GDP) growth rate was used to illustrate the relationship between economic development and bridge maintenance costs. Finally, these two neural networks were used to establish maintenance cost prediction models, respectively. The results from the two models were compared and their prediction accuracies were analyzed. The prediction performance of the CNN model for bridge maintenance costs was found to be better than that of the traditional fully-connected ANN model. The results of this study will enhance the opportunity for bridge managers to balance lifecycle maintenance costs. Full article

Discontinuing ad creatives at an appropriate time is one of the most important ad operations that can have a significant impact on sales. Such operational support for ineffective ads has been less explored than that for effective ads. After pre-analyzing 1,000,000 real-world ad creatives, we found that there are two types of discontinuation: short-term (i.e., cut-out) and long-term (i.e., wear-out). In this paper, we propose a practical prediction framework for the discontinuation of ad creatives with a hazard function-based loss function inspired by survival analysis. Our framework predicts the discontinuations with a multi-modal deep neural network that takes as input the ad creative (e.g., text, categorical, image, numerical features). To improve the prediction performance for the two different types of discontinuations and for the ad creatives that contribute to sales, we introduce two new techniques: (1) a two-term estimation technique with multi-task learning and (2) a click-through rate-weighting technique for the loss function. We evaluated our framework using the large-scale ad creative dataset, including 10 billion scale impressions. In terms of the concordance index (short: 0.896, long: 0.939, and overall: 0.792), our framework achieved significantly better performance than the conventional method (0.531). Additionally, we confirmed that our framework (i) demonstrated the same degree of discontinuation effect as manual operations for short-term cases, and (ii) accurately predicted the ad discontinuation order, which is important for long-running ad creatives for long-term cases. Full article

This paper focuses on the opportunity to use multiple star trackers to help space situational awareness and space surveillance. Catalogs of space debris around Earth are usually based on ground-based measurements, which rely on data provided by ground-based radar observations and ground-based optical observations. However, space-based observations offer new opportunities because they are independent of the weather and the circadian rhythms to which the ground system is subjected. Consequently, space-based optical observations improve the possibility of space debris detection and cataloging. This work deals with a feasibility study of an innovative strategy, which consists of the use of a star sensor with a dedicated algorithm to run directly on board. This approach minimizes the impact on the original mission of the satellite, and on this basis, it has also the function of space debris monitoring. Therefore, theoretically, every satellite with a star tracker can be used as a space surveillance observer. In this paper, we propose a multi-star space debris detecting and positioning method with constant geocentric observation. Using the multi-star tracker joint positioning method, the angle measurement data of the star tracker is converted into the spatial coordinates of the target. In addition, the Gaussian MMSE difference correction algorithm is used to realize the target positioning of multiple optical observations, and the spatial target position information of the multi-frame images is fused, thus completing the solution of the orbit equation. The simulation results show that the proposed method is sufficient to detect and position space debris. It also demonstrates the necessity and feasibility of cooperative network observation by multiple star trackers. Full article

Convolutional Neural Networks (CNNs) have recently been proposed as a solution in texture and material classification in computer vision. However, inside CNNs, the internal layers of pooling often cause a loss of information and, therefore, is detrimental to learning the architecture. Moreover, when considering images with repetitive and essential patterns, the loss of this information affects the performance of subsequent stages, such as feature extraction and analysis. In this paper, to solve this problem, we propose a classification system with a new pooling method called Discrete Wavelet Transform Pooling (DWTP). This method is based on the image decomposition into sub-bands, in which the first level sub-band is considered as its output. The objective is to obtain approximation and detail information. As a result, this information can be concatenated in different combinations. In addition, wavelet pooling uses wavelets to reduce the size of the feature map. Combining these methods provides acceptable classification performance for three databases (CIFAR-10, DTD, and FMD). We argue that this helps eliminate overfitting and that the learning graphs reflect that the datasets show learning generalization. Therefore, our results indicate that our method based on wavelet analysis is feasible for texture and material classification. Moreover, in some cases, it outperforms traditional methods. Full article

Stereolithography is known as one of the best Additive Manufacturing technologies in terms of geometrical and dimensional precision for polymeric materials. In recent years, a lot of studies have shown that the creation of ceramic resins, through a particular combination of monomeric components and ceramic powders, allows to obtain complex shape geometries thanks to the photopolymerization process. This review highlights the characteristics and properties of ceramic resins, peculiarities of the ceramic stereolithography processes, up to the relationship between the composition of the ceramic resin and the complexity of the post-processing phases. The comparison of different studies allows outlining the most common steps for the production of ceramic resins, as well as the physical and chemical compatibility of the different compounds that must be studied for the good feasibility of the process. Full article

Due to the aggravating situations regarding climate change, resource supply, and land consumption by the landfilling of residual materials, it is necessary to develop recycling processes that allow the recovery of valuable metals from industrial residues with significantly reduced CO₂ emissions. In this context, it is conceivable that processes using chlorination reactions will be of importance in the future. The simultaneous selective chlorination and evaporation of nine valuable metals was evaluated theoretically and experimentally in small-scale STA trials; then, it was tested practically on six different iron precipitation residues from the zinc and nickel industries. The metal chlorides FeCl₃∙6H₂O and MgCl₂∙6H₂O were identified as the most effective reactants, resulting in high extraction rates for the metals In, Ag, Zn, Pb, Au, and Bi, while lower yields are achievable for Sn, Cu, and Ni. Iron, which is predominant in volume in the residual materials, shows lower chlorination tendencies which allows the effective separation of the valuable elements of interest from the iron containing matrix. Full article

In this contribution, an unsteady numerical simulation of the flow in a microfluidic oscillator has been performed. The transient turbulent flow inside the device is described by the Unsteady Reynolds Averaged Navier–Stokes equations (URANS) coupled with proper turbulence models. The main characteristics of the complex fluid flow inside the device along one oscillation cycle was analyzed in detail, including not only velocity contours but also the pressure and turbulent kinetic energy fields. As a result, two-dimensional simulations provided good estimations of the operating frequency of the fluidic actuator when compared with experimental measurements in a range of Reynolds numbers. Moreover, with the objective of altering the operating frequency of the apparatus and, in order to adapt it to different applications, geometrical modifications of the feedback channels were proposed and evaluated. Finally, a fully three-dimensional simulation was carried out, which allowed for the identification of intricate coherent structures revealing the complexity of the turbulent flow dynamics inside the fluidic oscillator. Full article

In this study, a novel feature learning method for synthetic aperture radar (SAR) image automatic target recognition is presented. It is based on spatial pyramid matching (SPM), which represents an image by concatenating the pooling feature vectors that are obtained from different resolution sub-regions. This method exploits the dependability of obtaining the weighted pooling features generated from SPM sub-regions. The dependability is determined by the residuals obtained from sparse representation. This method aims at enhancing the weights of the pooling features generated in the sub-regions located in the target and suppressing the weights of the background. The feature representation for SAR image target recognition is discriminative and robust to speckle noise and background clutter. Experiments performed on the Moving and Stationary Target Acquisition and Recognition public dataset prove the advantageous performance of the presented algorithm over several state-of-the-art methods. Full article

Small object detection becomes a challenging problem in computer vision due to low resolution and less feature information. Making full use of high-resolution features is an important factor in improving small object detection. In this paper, to improve the utilization of high-resolution features, this work proposes the Bidirectional Multi-scale Feature Enhancement Network (BFE-Net) based on RetinaNet. First, this work introduces a bidirectional feature pyramid structure to shorten the propagation path of high-resolution features. Then, this work utilizes residually connected dilated convolutional blocks to fully extract high-resolution features of low-feature layers. Finally, this work supplements the high-resolution features lost in the high-level feature propagation process by leveraging the high-level guided lower-level features. Experiments show that our proposed BFE-Net achieves stable performance gains in the object detection task. Specifically, the improved method improves RetinaNet from 34.4 AP to 36.3 AP on the challenging MS COCO dataset and especially achieves excellent results in small object detection with an improvement of 2.8%. Full article

Soil salinization has become a major problem for agriculture worldwide, especially because this phenomenon is continuously expanding in different regions of the world. Salinity is a complex mechanism, and in the soil ecosystem, it affects both microorganisms and plants, some of which have developed efficient strategies to alleviate salt stress conditions. Currently, various methods can be used to reduce the negative effects of this problem. However, the use of biological methods, such as plant-growth-promoting bacteria (PGPB), phytoremediation, and amendment, seems to be very advantageous and promising as a remedy for sustainable and ecological agriculture. Other approaches aim to combine different techniques, as well as the utilization of genetic engineering methods. These techniques alone or combined can effectively contribute to the development of sustainable and eco-friendly agriculture. Full article

The seismic risk assessment of existing urban areas provides important information for the process of seismic risk reduction in different phases of planning and emergency management. Between different large-scale assessment approaches, a vulnerability index method is often used for the first screening of the buildings and vulnerability classification. However, this method cannot fully predict the effects of a specific seismic action on buildings. This paper fully extends the scale of the settlement and properly upgrades a methodology previously proposed by authors to predict seismic damage and the risk to a restricted number of masonry buildings in the Croatian settlement Kaštel Kambelovac located along the Adriatic coast. The proposed approach is based on a hybrid empirical-analytical procedure that combines seismic vulnerability indices with critical peak ground accelerations for different limit states computed through a non-linear pushover analysis. The procedure’s outcomes are the computation of a relationship linking vulnerability indices to peak ground acceleration for a series of states, corresponding to damage limitation, significant damage, and near collapse. The described methodology is used to estimate seismic risk in terms of damage and the index of seismic risk for selected return periods. The general methodology has allowed a full seismic vulnerability assessment of the whole Croatian settlement of Kaštel Kambelovac. Full article

Embedded systems are increasingly applied in numerous security-sensitive applications, such as industrial controls, railway transports, intelligent vehicles, avionics and aerospace. However, embedded systems are compromised in the execution of untrusted programs, where the instructions could be maliciously tampered with to cause unintended behaviors or program execution failures. Particularly for remote-controlled embedded systems, program execution monitoring and instruction fault self-repair are important to avoid unintended behaviors and execution interruptions. Therefore, this paper presents a hardware-enhanced embedded system with the integration of a Security Processing Unit (SPU) in which integrity signature checking and checkpoint-rollback mechanisms are coupled to achieve real-time program execution monitoring and instruction fault self-repairing. This System-on-Chip (SoC) design was implemented and validated on the Xilinx Virtex-5 FPGA development platform. Based on the evaluation of the SPU in terms of the performance overhead, security capability, and resource consumption, the experimental results show that, while the CPU executes different benchmarks, the average performance overhead of the SPU lowers to 1.92% at typical 8-KB I/D caches, and it provides both program monitoring and fault self-repairing capabilities. Unlike conventional hardware detection technologies that require manual handling to recovery program executions, the CPU–SPU collaborative SoC is a resilient architecture equipped with instruction tampering detection and a post-detection strategy of instruction fault self-repairing. Moreover, the embedded system satisfies a good balance between high security and resource consumption. Full article