Search

Connected vehicles will change the modes of future transportation management and organization, especially at an intersection without traffic light. Centralized coordination methods globally coordinate vehicles approaching the intersection from all sections by considering their states altogether. However, they need substantial computation resources since they own a centralized controller to optimize the trajectories for all approaching vehicles in real-time. In this paper, we propose a centralized coordination scheme of automated vehicles at an intersection without traffic signals using reinforcement learning (RL) to address low computation efficiency suffered by current centralized coordination methods. We first propose an RL training algorithm, model accelerated proximal policy optimization (MA-PPO), which incorporates a prior model into proximal policy optimization (PPO) algorithm to accelerate the learning process in terms of sample efficiency. Then we present the design of state, action and reward to formulate centralized coordination as an RL problem. Finally, we train a coordinate policy in a simulation setting and compare computing time and traffic efficiency with a coordination scheme based on model predictive control (MPC) method. Results show that our method spends only 1/400 of the computing time of MPC and increase the efficiency of the intersection by 4.5 times. [ABSTRACT FROM AUTHOR]

Metallic wires are used in many industrial applications, where the diameter and conductivity are key parameters to be measured. In eddy currents testing, the inductance of the sensor is affected by both the diameter and conductivity of sample wire. Consequently, wire diameter and conductivity cannot be determined from a single inductance measurement. In this paper, a novel combined inductive and capacitive sensor is designed, in which the diameter can be inversed from the capacitance measurement of the sensor. Based on the diameter value, the inductance measurement can be used to accurately determine the conductivity of the wires using a look-up table approach. In addition, a model for a wire inside of the capacitive sensor is derived. Measurements have been carried out for 16 samples and the results have verified the proposed approach. This technique could contribute to real-time in-situ quality control of wire + arc additive manufacture processes. [ABSTRACT FROM AUTHOR]

Due to more or less deviations in the imaging system, there will be noise in the image, which makes the image segmentation inaccurate. To divide a natural image into a more accurate binary image, the target and background of the image are effectively separated to achieve a more effective segmentation result. Therefore, this paper proposes an image segmentation algorithm combining a saliency map and neutrosophic set (NS) theory. First, to overcome the problem of weak edges in the image, we highlight the details and use the guided filter to filter the various channels of the natural image. Then, the initial saliency map is generated. After the weighted superposition of the initial saliency map, the local entropy map and the gray scale map, the final saliency map can be generated using the nonlinear function, and it can effectively highlight the foreground information of the image. Second, the saliency map is transformed to the NS domain and interpreted by three subsets: true (T), indeterminate (I), and false (F). According to NS theory, the indeterminacy is reduced, and the segmentation results are finally obtained by using the method of threshold. Various experiments were done and compared with other state-of-the-art approaches. These experiments demonstrate the effect of the proposed work, which is fast and effective for de-noising. [ABSTRACT FROM AUTHOR]

Positron emission tomography (PET) imaging is a noninvasive technique that makes it possible to probe biological metabolic processes in vivo. However, PET image reconstruction is challenging due to the ill-posedness of the inverse problem. Many image reconstruction methods have been proposed over the past few years to improve diagnostic performance. However, most of these methods can compromise the reconstruction of important high-frequency structural details after aggressive denoising. To address this problem, in this work, we propose a novel deep learning method that reconstructs PET images using a cascading back-projection neural network (bpNet). This network consists of a domain translation operation, which acts as prior knowledge, cascaded with a modified encoder-decoder network. The image reconstruction pipeline ranges from the sinogram to the back-projection image and then to the PET image. Quantitative results from simulation data and Derenzo phantom experiments with the small animal PET prototype system developed in our laboratory clearly demonstrate that our proposed method provides favorable reconstructed image quality, especially for low-count PET image reconstruction. [ABSTRACT FROM AUTHOR]

Fuzzy systems have achieved great success in numerous applications. However, there are still many challenges in designing an optimal fuzzy system, e.g., how to efficiently optimize its parameters, how to balance the trade-off between cooperations and competitions among the rules, how to overcome the curse of dimensionality, how to increase its generalization ability, etc. Literature has shown that by making appropriate connections between fuzzy systems and other machine learning approaches, good practices from other domains may be used to improve the fuzzy systems, and vice versa. This article gives an overview on the functional equivalence between Takagi–Sugeno–Kang fuzzy systems and four classic machine learning approaches—neural networks, mixture of experts, classification and regression trees, and stacking ensemble regression—for regression problems. We also point out some promising new research directions, inspired by the functional equivalence, that could lead to solutions to the aforementioned problems. To our knowledge, this is so far the most comprehensive overview on the connections between fuzzy systems and other popular machine learning approaches, and hopefully will stimulate more hybridization between different machine learning algorithms. [ABSTRACT FROM AUTHOR]

This paper studies the fault detection problem for a class of discrete-time switched linear systems under dwell-time (DT) constraints, using the parity space-based approach. The DT-dependent Lyapunov function is employed to investigate the asymptotic stability with less conservatism and to solve the constant ${l}_{2}$ -gain performance analysis problem, and its advantage is verified compared to the time-independent Lyapunov function approach. The corresponding switching residual generation is made to carry out the desired fault detection in the framework of parity space-based model. Then, by means of solving a generalized eigenvalue–eigenvector problem, the parity space matrices design is implemented. A quantitative relationship is established between the optimization performance and the choice of the parity space order. Two numerical examples are utilized to demonstrate effectiveness of the developed fault detection approach, including an application to switched RLC circuits. [ABSTRACT FROM AUTHOR]

In this paper, the event-triggered $\boldsymbol {H}_{\boldsymbol \infty }$ state estimation problem is investigated for a two-degree-of-freedom quarter-car suspension system operated over a switching-channel network environment. First, the channel-switching is governed by a nonhomogeneous Markov chain whose probability transition matrix is time-varying. Then, a Markov jump linear system model is adopted to represent the overall networked system in view of the event-triggered communication scheme, signal quantization and random packet losses on account of the limited network bandwidth. By virtue of the Lyapunov functional and linear matrix inequality method, the event-triggered $\boldsymbol {H}_{\boldsymbol \infty }$ state estimation problem is transformed into an optimization problem that switching-channel-dependent estimators are designed such that the estimation error system is exponentially stable in the mean square sense and achieves a desired performance level. Finally, a simulation example is used to demonstrate the validity of proposed design method. [ABSTRACT FROM AUTHOR]

Loaded walking with a rucksack results in both gravitational and inertial forces of the load that must be borne by human carriers. The inertial force may be the source of metabolic burden and musculoskeletal injuries. This paper presents a lightweight backpack with a disturbance observer-based acceleration control to minimize the inertial force. The backpack was evaluated by seven participants walking on a treadmill at 5 km h−1 with a 19.4 kg load. Three experimental conditions were involved, including walking with a locked load (LOCKED), with an acceleration-controlled load (ACTIVE) using the designed backpack and walking with the same load using a rucksack (RUCKSACK). Our results showed that the ACTIVE condition reduces the load acceleration by 98.5% on average, and reduce the gross metabolic power by 8.0% and 11.0% as compared to LOCKED and RUCKSACK conditions respectively. The results demonstrate that the proposed active backpack can improve the loaded walking economy compared with a conventional rucksack in level-ground walking. [ABSTRACT FROM AUTHOR]

Multi-view learning improves the learning performance by utilizing multi-view data: data collected from multiple sources, or feature sets extracted from the same data source. This approach is suitable for primate brain state decoding using cortical neural signals. This is because the complementary components of simultaneously recorded neural signals, local field potentials (LFPs) and action potentials (spikes), can be treated as two views. In this paper, we extended broad learning system (BLS), a recently proposed wide neural network architecture, from single-view learning to multi-view learning, and validated its performance in decoding monkeys’ oculomotor decision from medial frontal LFPs and spikes. We demonstrated that medial frontal LFPs and spikes in non-human primate do contain complementary information about the oculomotor decision, and that the proposed multi-view BLS is a more effective approach for decoding the oculomotor decision than several classical and state-of-the-art single-view and multi-view learning approaches. [ABSTRACT FROM AUTHOR]

Photosynthetically active radiation (PAR) incident at the surface is crucial for understanding and modeling the Earth’s climate and ecosystems. In this article, a general parameterization scheme suitable for PAR estimation under cloudy skies is proposed based on an elaboration on the PAR radiative transfer (RT) processes above, in, and beneath cloud layers. Its most important novel property is that all RT processes in cloudy atmospheres are explicitly explained, including ozone absorption, Rayleigh scattering, cloud single scattering, cloud multiple scattering, aerosol scattering, and cloud reflection. Theoretical accuracy evaluations show over 95% of errors (against rigorous RT calculations) lie within ±20 W/m2, and an operational application with multisource satellite products as inputs shows the root-mean-square error (RMSE) of ≤42 W/m2 on the hourly timescale. Therefore, the parameterization scheme is accurate both in theory and actual applications. [ABSTRACT FROM AUTHOR]

Moving target defense (MTD) has emerged as a proactive defense mechanism aiming to thwart a potential attacker. The key underlying idea of MTD is to increase uncertainty and confusion for attackers by changing the attack surface (i.e., system or network configurations) that can invalidate the intelligence collected by the attackers and interrupt attack execution; ultimately leading to attack failure. Recently, the significant advance of software-defined networking (SDN) technology has enabled several complex system operations to be highly flexible and robust; particularly in terms of programmability and controllability with the help of SDN controllers. Accordingly, many security operations have utilized this capability to be optimally deployed in a complex network using the SDN functionalities. In this paper, by leveraging the advanced SDN technology, we developed an attack graph-based MTD technique that shuffles a host’s network configurations (e.g., MAC/IP/port addresses) based on its criticality, which is highly exploitable by attackers when the host is on the attack path(s). To this end, we developed a hierarchical attack graph model that provides a network’s vulnerability and network topology, which can be utilized for the MTD shuffling decisions in selecting highly exploitable hosts in a given network, and determining the frequency of shuffling the hosts’ network configurations. The MTD shuffling with a high priority on more exploitable, critical hosts contributes to providing adaptive, proactive, and affordable defense services aiming to minimize attack success probability with minimum MTD cost. We validated the out performance of the proposed MTD in attack success probability and MTD cost via both simulation and real SDN testbed experiments. [ABSTRACT FROM AUTHOR]

This paper investigates a heterogeneous cellular network (HCN) topology, where multiple sub-6 GHz bands are available for access and device-to-device (D2D) links with the full-duplex (FD) mode, and an integrated millimeter-wave (mm-wave) band is added to the D2D links’ sub-channel resource options. D2D links share spectrum resources of cellular users or dedicated mm-wave sub-channel. Densely deployed small cells scenario provides an ideal platform to implement mm-wave, FD and D2D communications. With large bandwidths available, mm-wave has the potential to meet the expected extreme data rate demands. Leveraging highly directional beamforming, severe transmission loss and the blockage phenomenon of mm-wave can be alleviated. Theoretically, FD communications can double the spectral efficiency due to bi-directional communications with the same spectral and temporal resources. However, such benefit comes at the cost of residual self-interference (RSI). Multi-user interference (MUI) and RSI among network uplink, downlink and D2D links are the main challenges. In the paper, we formulate an optimization problem for FD access and D2D links sub-channel allocation underlaying HCNs combining sub-6 GHz and mm-wave bands to make the system transmission rate maximized. Then, a scheme based on the coalition formation game is proposed to deal with this challenging NP-complete optimization. Finally, performance evaluation investigates the suitable conditions for FD and HD operations. Under different network parameter settings, we quantify associated performance of proposed scheme. [ABSTRACT FROM AUTHOR]

Since the manual inspection of analog instruments is inefficient, many computer vision-based automatic reading systems have been proposed recently. However, most of them use fixed cameras that are costly due to a large number of used cameras. Although some other systems adopting the pan–tilt–zoom camera and the movable inspection robot can avoid using many cameras, they have to overcome high computational cost in aligning the camera to the tested instrument. Meanwhile, most existing systems are instrument type-dependent and, hence, cannot handle multiple types of instruments simultaneously. In this article, first, based on an inspection robot, an automatic reading system equipped with a pan–tilt–zoom camera is designed for different types of round-shape analog instruments. Then, a fast camera alignment algorithm based on visual servo is proposed, in which YOLOv3 is applied and improved to locate the instrument and guide the camera to iteratively align with the instrument. Finally, a monocular-vision pointer reconstruction algorithm is proposed to accurately read the instrument. Experimental results demonstrated that our proposed system is fast and reliable in the camera-alignment process and is effective in reading different types of analog instruments during the robot-based inspection. [ABSTRACT FROM AUTHOR]

The overwhelming increase of ubiquitous data, connections, and services brings serious challenges, in particular facing the demanding requirements of the Internet of Things (IoT). In order to seek better solutions and achieve more efficient information retrieval, artificial intelligence (AI) serves as a strong technical earthquake and contributes a lot to data analysis and decision making. It plays a compelling role in prompting digital and intelligent services. In this article, we focus on and emphasize the great significance pertaining to the convergence of IoT and AI. We first elaborate two typical forms of AI, namely knowledge-enabled AI and data-driven AI, with a comparison between respective advantages and disadvantages. Then we survey recent progress relating to the convergence of AI throughout the IoT architecture, from the sensing layer through the network layer to the application layer. In addition, a case study of a smart city is presented illustrating the convergence between IoT and AI. Furthermore, we point out open issues worth further research. The convergence of IoT and AI marries the merits of both and enables strong capability of resolving a broad range of problems. [ABSTRACT FROM AUTHOR]

Heterogeneous vehicle routing problem with time windows (HVRPTW) aspect of hazardous material (HazMat) transportation is studied in this article. Given the multiobjective nature of HazMat transportation, a three-objective optimization model is defined for HVRPTW in HazMat transportation. The objective is to determine fleet size and routes so as to meet all given constraints as well as to minimize the objectives of the total traveling cost, the transportation risk, and the average vehicle redundancy. A load-variant HazMat transportation risk assessment model considering vehicle type and waiting time is presented to describe the transportation risk. A variable neighborhood search-based hybrid multiobjective evolutionary algorithm (VN-HMOEA) is proposed for solving the problem. The proposed VN-HMOEA integrates a two-phase push forward insertion heuristic (TP-PFIH) for initial population construction, specialized evolutionary operators for optimizing different objectives and a VNS metaheuristic for local search exploitation. The algorithm is tested on the modified Solomon benchmark instances for HVRPTW. Experimental results show that the proposed VN-HMOEA is competitive in terms of convergence and diversity. We also find that multiobjective fashion is of great significance for transportation risk mitigation to provide a set of nondominated solution rather than a single solution. [ABSTRACT FROM AUTHOR]

The processing errors for columns of heavy-duty machine tools are affected by deformation induced by a variety of force loads and thermal loads. In general, error reduction usually uses method of error compensation. However, the precondition of error compensation is to accurately obtain the machine full-field displacement. The inverse finite element method (iFEM) is an efficient tool for the real-time reconstruction of deformation. It obtains nodal degree of freedoms (DOFs) by solving the minimum value of the error least-squares function, which constitutes the relationship between the DOFs and analytical strain. Although double-sided strain rosettes are widely used in the iFEM to obtain strain information, the strain rosettes have some shortcomings, including the internal installation being difficult and measurement size being large. Fiber Bragg grating (FBG) sensors have the advantages of being lightweight, small geometric profile and being easy to installation. Therefore, this study utilizes distributed single-sided and unidirectional strain data obtained from FBG sensors. To further improve the possibility of this practical application, cubic spline interpolation is used to reduce the number of sensors and further extend the practical usefulness. Considering coordinate transformation and boundary conditions, the total deformation of the column is determined. The superior accuracy and practical applicability of the improved formulation is demonstrated after comparing the results of analytic solutions and that of the FEM. Therefore, new direction method for supporting the machining precision of heavy-duty machine tools is proposed. [ABSTRACT FROM AUTHOR]

In the presence of prominent gate oxide trapping, the conventional technique for channel mobility extraction in MOSFETs based on I – V/C – V measurements becomes inadequate. This is the consequence of two different effects associated with oxide traps: gate voltage stretch-out and electron trapping and detrapping in the oxide at the megahertz (MHz)-range frequencies that are commonly utilized. In thin-channel planar InGaAs MOSFETs, both effects are observed and found to result in a severe overestimation of mobile charge and subsequently an underestimation of mobility using I – V/C – V. To address this issue, we demonstrate a new mobility extraction technique (RF- ID) based on concurrent I – V and S-parameter measurement in the gigahertz (GHz) regime that is largely immune to oxide trapping. Excellent agreement with Hall measurements as well as with theoretical predictions from Poisson–Schrodinger simulations gives confidence to the new technique. Importantly, the new technique is not limited to InGaAs planar MOSFETs, but applies to any device geometry and any material system. Promising mobility ~1100 cm2/V⋅s is found in quantum-well planar InGaAs MOSFETs with a 4-nm-thick channel. [ABSTRACT FROM AUTHOR]

Distributed stabilization problem is investigated in this brief for a class of heterogeneous multi-agent systems (MASs) in the presence of exogenous disturbances and cooperation-competition interactions. Specifically, the dynamics of the heterogeneous agents are governed by the second-order systems with the nonlinear intrinsic dynamics where the heterogeneity of agents is mainly reflected in the following three aspects: the intrinsic dynamics of agents, the heterogeneous velocity damping terms, and the exogenous disturbances. With the assumption that the exogenous disturbances are bounded, a discontinuous distributed protocol associated with a time-varying estimator of the agent are designed. Then, with the help of the LaSalle-Yoshizawa theorem for nonsmooth system and the Barbalat’s lemma, it is proved that no matter whether the network topology is structurally balanced or not, the heterogeneous MASs can achieve robust distributed stabilization asymptotically as long as the control parameters of the distributed protocol are chosen appropriately. Finally, the derived analytical results are demonstrated by performing a numerical example. [ABSTRACT FROM AUTHOR]

In pneumatic muscle actuators (PMAs)-driven robotic applications, there might exist unpredictable shocks which lead to the sudden change of desired trajectories and large tracking errors. This is dangerous for physical systems. In this article, we propose a novel adaptive proxy-based robust controller (APRC) for PMAs, which is effective in realizing a damped response and regulating the behaviors of the PMA via a virtual proxy. Moreover, the integration of the APRC and the nonlinear disturbance observer further handles the system uncertainties/disturbances and improves the system robustness. According to the Lyapunov theorem, the tracking states of the closed-loop PMA control system are proven to be globally uniformly ultimately bounded through two motion phases. Extensive experiments are conducted to verify the superior performance of our approach, in multiple tracking scenarios. [ABSTRACT FROM AUTHOR]

There has been a burgeoning prospect in developing a spaceborne P-band synthetic aperture radar (SAR) mission for its stronger penetrability through foliage and subsurface than the higher-frequency system. However, the transionospheric signals operating at P-band are more susceptible to scintillation impacts, which may bring about the decorrelation of the signal amplitude, phase, and frequency introduced by ionospheric irregularities. In this article, a comprehensive numerical model of the generalized ambiguity function (GAF) is established to evaluate SAR imaging deterioration. On the one hand, an improved two-frequency and two-position coherence function (TFTPCF) is integrated in the GAF to include the amplitude scintillation derived from the diffraction. On the other hand, the anisotropic characteristics of the irregular structure is introduced into TFTPCF by adopting the Rino’s 2-D spectrum. Furthermore, the ambiguous resolution is redefined for a more strict numeration. Numerical analyses about the ionospheric coherence and an ambiguous resolution are performed to investigate their sensitivity to scintillation parameters. Results show that this model is capable of depicting more comprehensive effects of the anisotropic irregular ionosphere, which may exhibit a rod-like structure, including elongation by anisotropic scale, rotation by geomagnetic heading, and projection by geomagnetic inclination. At last, the signal-level simulations are operated to verify the effectiveness of numerical conclusions, which further confirm that the structural configuration of anisotropic irregularities has a complicated effect on spaceborne P-band SAR image resolution. [ABSTRACT FROM AUTHOR]

The long-term storage of nuclear-spent fuel in geological repositories has introduced the need to develop new technologies for safeguarding the fuel. Continuity of knowledge about the fuel’s location must be maintained during transport of the copper canisters that contain the fuel from the encapsulation plant to the final repository. Among the possible different containment and surveillance measures are the identification and authentication of each canister. The authors propose an innovative solution for this purpose, which is based on the ultrasonic acquisition of two fingerprints: an artificial code realized by chamfers machined in the inner surface of the copper lid (identification) and a natural signature due to the weld between the lid and the tube (authentication). Several prototypes of ultrasonic acquisition devices have been developed and tested on real copper samples to validate the identification and authentication concepts. This article describes the design of a new, optimized version of the ultrasonic acquisition device. The aim of this new design is to provide a solution for the identification and authentication of copper canisters using a portable device that is cost-effective and easy to use, which does not require the presence of an inspector in the field. The mechanical design of the reader has been upgraded with the introduction of a stepper motor and a new probe holder that includes a beam splitter to acquire two fingerprints simultaneously. The new device has been tested on copper samples, both with and without chamfers, and the results are reported in this article. The implementation of the seal fingerprint acquisition device (SFAD) within the new portable acquisition system is discussed at the end of this article with the goal of improving the electronic performance of the acquisition device in the field. [ABSTRACT FROM AUTHOR]

The unique and random physical-layer properties of optical fiber channels have been exploited for a simple and cost-effective secure key generation and distribution (SKGD). However, state-of-the-art SKGD approaches focused mainly on initial key extraction, while the post-processing protocol has to be explored to guarantee error-free operation, effective key generation rate (KGR) and high key secrecy. Here we propose a full post-processing protocol for SKGD in fiber networks. A lossy quantizer is optimized to achieve the trade-off between the KGR and key disagreement rate. The distributed source coding based information reconciliation is used to remove quantization errors. The privacy amplification is applied to improve the randomness and secrecy of the generated key using hashing functions. The post-processing protocol provides a SKGD solution of error-free, high randomness and high secrecy in fiber networks. [ABSTRACT FROM AUTHOR]

In this article, we investigate the potential of empirical wavelet transform (EWT) as a moving ground target detection and classification technique using seismic signal modality. EWT gives the opportunity of adaptive wavelet technique to construct basis functions based on the information contained in the seismic signal. A seismic dataset is created by recording the seismic signature of moving ground targets, i.e., bus and tractor using an array of geophones. Statistical features computed using EWT based time-frequency coefficients, and subsequently, SVM and KNN classify the moving ground targets on the basis of these features. The performance of the proposed technique is analyzed using the parameters: accuracy, true positive rate (TPR), and AUC (Area under the curve) -ROC (Receiver Operating Characteristics), and subsequently, the comparison performed with an existing technique, i.e., STFT. The results are satisfactory with accuracy and TPR of the order of 90% and AUC ~ 95% for classification results between bus and noise. Similarly, for classification between tractor and noise, the accuracy, TPR, and AUC are 84%, 82%, and 90%, respectively. The performance of the EWT for ground target detection and classification is also presented in terms of the F-Score and confusion matrix. The classification using EWT and SVM as a classifier provides F-Score as 78%, 67%, and 86% for bus, tractor, and noise, respectively, which is having the average relative enhancement of about ~8% in comparison with STFT based technique. [ABSTRACT FROM AUTHOR]

Scalable High Efficiency Video Coding (SHVC) is the extension of High Efficiency Video Coding (HEVC). In intra prediction for quality SHVC, a Coding Unit (CU) is recursively divided into a quadtree-based structure from the largest ${64}{\times }{64}$ CU to the smallest ${8}{\times }{8}$ CU, in which 35 intra prediction modes and Inter-Layer Reference (ILR) mode are checked to determine the best possible mode. This leads to very high coding efficiency but also results in an extremely high coding complexity. To improve coding speed while maintaining coding efficiency, in this paper, we propose a new efficient algorithm for fast intra prediction for enhancement layer in SHVC. First, temporal and spatial correlations, as well as their correlation degrees, are combined in a Naive Bayes classifier to predict depth probabilities and skip depths with low likelihood. Second, for a given depth candidate, we combine ILR mode probability with Partial Zero Blocks (PZBs) based on the Sum of Squared Differences (SSD) to determine whether the ILR mode is the best one. In that case, we can skip intra prediction, which requires very high complexity. Third, initial Intra Modes (IMs) are obtained through Sobel operator, and are combined with the relationship between IMs and their corresponding Hadamard Cost (HC) values to predict candidate IMs in Rough Mode Decision (RMD). Then, an analytical criterion of early termination is developed based on the HC values of two neighboring IMs in the Rate-Distortion Optimization (RDO) process. Finally, we combine depth probabilities and the distribution of residual coefficients at the current depth to early terminate depth selection. The proposed scheme can significantly decrease the complexity of depth determination while reducing the complexity of mode decision for a depth candidate. Our experimental results demonstrate that the proposed scheme can achieve a speed up gain of more than 80% in average, while maintaining coding efficiency. [ABSTRACT FROM AUTHOR]

Human parsing and matting play important roles in various applications, such as dress collocation, clothing recommendation, and image editing. In this paper, we propose a lightweight hybrid model that unifies the fully-supervised hierarchical-granularity parsing task and the unsupervised matting one. Our model comprises two parts, the extensible hierarchical semantic segmentation block using CNN and the matting module composed of guided filters. Given a human image, the segmentation block stage-1 first obtains a primitive segmentation map to separate the human and the background. The primitive segmentation is then fed into stage-2 together with the original image to give a rough segmentation of human body. This procedure is repeated in the stage-3 to acquire a refined segmentation. The matting module takes as input the above estimated segmentation maps and produces the matting map, in a fully unsupervised manner. The obtained matting map is then in turn fed back to the CNN in the first block for refining the semantic segmentation results. [ABSTRACT FROM AUTHOR]

Picture Wise Just Noticeable Difference (PW-JND), which accounts for the minimum difference of a picture that human visual system can perceive, can be widely used in perception-oriented image and video processing. However, the conventional Just Noticeable Difference (JND) models calculate the JND threshold for each pixel or sub-band separately, which may not reflect the total masking effect of a picture accurately. In this paper, we propose a deep learning based PW-JND prediction model for image compression. Firstly, we formulate the task of predicting PW-JND as a multi-class classification problem, and propose a framework to transform the multi-class classification problem to a binary classification problem solved by just one binary classifier. Secondly, we construct a deep learning based binary classifier named perceptually lossy/lossless predictor which can predict whether an image is perceptually lossy to another or not. Finally, we propose a sliding window based search strategy to predict PW-JND based on the prediction results of the perceptually lossy/lossless predictor. Experimental results show that the mean accuracy of the perceptually lossy/lossless predictor reaches 92%, and the absolute prediction error of the proposed PW-JND model is 0.79 dB on average, which show the superiority of the proposed PW-JND model to the conventional JND models. [ABSTRACT FROM AUTHOR]

The temporal flicker distortion is one of the most annoying noises in synthesized virtual view videos when they are rendered by compressed multi-view video plus depth in Three Dimensional (3D) video system. To assess the synthesized view video quality and further optimize the compression techniques in 3D video system, objective video quality assessment which can accurately measure the flicker distortion is highly needed. In this paper, we propose a full reference sparse representation-based video quality assessment method toward synthesized 3D videos. First, a synthesized video, treated as a 3D volume data with spatial (X-Y) and temporal (T) domains, is reformed and decomposed as a number of spatially neighboring temporal layers, i.e., X-T or Y-T planes. Gradient features in temporal layers of the synthesized video and strong edges of depth maps are used as key features in detecting the location of flicker distortions. Second, the dictionary learning and sparse representation for the temporal layers are then derived and applied to effectively represent the temporal flicker distortion. Third, a rank pooling method is used to pool all the temporal layer scores and obtain the score for the flicker distortion. Finally, the temporal flicker distortion measurement is combined with the conventional spatial distortion measurement to assess the quality of synthesized 3D videos. Experimental results on synthesized video quality database demonstrate our proposed method is significantly superior to the other state-of-the-art methods, especially on the view synthesis distortions induced from depth videos. [ABSTRACT FROM AUTHOR]

This article presents a physical model of the current-assisted photonic demodulator (CAPD) for the time-of-flight (TOF) CMOS image sensor. An analytical expression relating the optical photogenerated current to the input bias and incident light intensity is derived by solving the continuity and drift-diffusion equations along the optical window of the CAPD. The model has been extensively verified with numerical simulations, as well as with published experimental data. Based on the developed model, the CAPD efficiency with respect to peak modulation voltage is discussed. [ABSTRACT FROM AUTHOR]

The insulation material applied on the stator windings of automotive electric machines are susceptible to the excessive temperature rise. Typically, several thermocouples are embedded in the end-winding and the active winding in the slot to detect the peak temperatures in real-time to avoid thermal damage and adjust cooling. However, in the slot windings, the uneven distribution of power loss will change with frequency. This phenomenon causes the movement of the hotspot in the slot winding at different operating states, while it is not so obvious in the end-winding. As a result, the peak temperature in the slot winding cannot always be detected exactly by the fixed thermocouples, which has never been realized and valued. In this article, we analyze the movement of the hotspot in both of the slot winding and end-winding. In order to improve the detecting precision, we use an offline theoretical thermal-magnetic model to deduce the relationship between the temperatures at the measured position and real hotspot at various operating states. Then the thermocouple reading can be corrected in real-time according to the deduced relationship. The experimental test validates the performance of the proposed method in elevating the precision of peak temperature detection. [ABSTRACT FROM AUTHOR]

This article proposes single-objective/multiobjective cat swarm optimization clustering algorithms for data partition. The proposed methods use the cat swarm to search the optimal. The position of the cat tightly associates with the clustering centers and is updated by two submodes: the seeking mode and the tracing mode. The seeking mode uses the simulated annealing strategy to update the cat position at a probability. Inspired by the quantum theories, the tracing mode adopts the quantum model to update the cat position in the whole solution space. First, the single-objective method is proposed and adopts the cohesion of clustering as the objective function, in which the kernel method is applied. For considering more objective functions to reveal diverse aspects of data, the multiobjective method is proposed and adopts both the cohesion and the connectivity as the objective functions. The Pareto optimization method is applied to balance the objectives. In the experiments, three kinds of data sets are used to examine the effectiveness of the proposed methods, which are three synthetic data sets, four data sets from the UCI Machine Learning Repository, and a field data set. Experimental results verified that the proposed methods perform better than the traditional clustering algorithms, and the proposed multiobjective method has the highest accuracy. Note to Practitioners—This article presents single-objective/multiobjective cat swarm optimization clustering analysis methods for data partition. Through automatically extracting meaningful or useful classes, clustering analysis could help the practitioners or the intelligent devices find the specific meanings of data, natural data structure, the data relationships, or other characteristics. The proposed methods use the cat swarm to search the optimal clustering result. One or more criterion functions could be selected as the optimization objectives. The time complexity of the multiobjective type is higher than that of the single-objective type. Therefore, in the industrial field, engineers should choose the number of the optimization objectives based on the actual requirements. The proposed methods could be widely used into industrial applications to deal with complex data sets. Future research could consider some more progressive optimization schemes to improve the effectiveness. [ABSTRACT FROM AUTHOR]

Distributed optical vibration sensors (DOVS) have attracted much attention recently since it can be used to monitor mechanical vibrations or acoustic waves with long reach and high sensitivity. Phase-sensitive optical time domain reflectometry ($\Phi $ -OTDR) is one of the most commonly used DOVS schemes. For $\Phi $ -OTDR, the whole length of fiber under test (FUT) works as the sensing instrument and continuously generates sensing data during measurement. Researchers have made great efforts to try to extract external intrusions from the redundant data. High signal-to-noise ratio (SNR) is necessary in order to accurately locate and identify external intrusions in $\Phi $ -OTDR systems. Improvement in SNR is normally limited by the properties of light source, photodetector and FUT. But this limitation can also be overcome by post-processing of the received optical signals. In this context, detailed methodologies of SNR enhancement post-processing algorithms in $\Phi $ -OTDR systems have been described in this paper. Furthermore, after successfully locating the external vibrations, it is also important to identify the types of source of the vibrations. Pattern classification is a powerful tool in recognizing the intrusion types from the vibration signals in practical applications. Recent reports of $\Phi $ -OTDR systems employed with pattern classification algorithms are subsequently reviewed and discussed. This thorough review will provide a design pathway for improving the performance of $\Phi $ -OTDR while maintaining the cost of the system as no additional hardware is required. [ABSTRACT FROM AUTHOR]

This article presents an integrated high-efficiency 28 GHz power amplifier (PA) employing a combination of Chireix outphasing and Doherty architectures in order to simultaneously achieve power back-off efficiency and linearity with a single RF input signal and no predistortion. The amplifier consists of a dual-input high-efficiency outphasing PA and a simple input network which serves as a power splitter and feeds the same signal to the inputs of the main and auxiliary PA cells that are biased in class-AB and class-C regions, respectively, similar to the Doherty architecture. The operation of the PA cells together with the Chireix combiner result in back-off efficiency enhancement plus systematic AM-AM and AM-PM variations which are used to correct the distortions caused by transistors, resulting in a linear response. The implemented PA demonstrates 19 dBm saturation power (Psat) with 34.4% peak power-added efficiency (PAE) and 6 dB back-off PAE of >23% at 27.5 GHz. The modulated signal performance using a 100 MHz 64QAM OFDM signal shows an average output power of 11.9 dBm with PAE >20%, EVM < 5%, and ACLR < −33 dBc without using predistortion. [ABSTRACT FROM AUTHOR]

This article presents a low-dropout regulator (LDO), with analog-proportional (AP) and digital integral (DI) controls. The design concerns are discussed at first, on how to improve the load transient response, enhance the power supply rejection (PSR), and reduce the limit cycle oscillation (LCO). For a good output dc accuracy, the DI section is implemented with shift-register-based coarse- and fine-tuning loops. Meanwhile, the AP section, based on a low-supply flipped-voltage follower (FVF), can respond fast to the load step and input supply ripple. A replica loop is used to define the steady-state output current of AP, allowing a sufficient dynamic swing against the supply ripple. To lower the load current range with no LCO, the AP section will output all the current at very light load. An error amplifier (EA) with moderate gain is added to improve the light-load output accuracy. This EA also improves the PSR by approximately 6 dB. Fabricated in a 65-nm CMOS process, a 65-mV undershoot is measured with a 0–10-mA load current step under 0.6-V supply voltage and 50-mV dropout. Due to the fast AP, a 5-MHz operation clock is applied to the digital section, reducing the overall quiescent current to 29 $\mu \text{A}$. A 0.37-ps figure of merit (FoM) is then achieved. A −22-dB PSR at 1 MHz is measured at 0.6-V supply, 100-mV dropout, and 10-mA load current. [ABSTRACT FROM AUTHOR]

Recent efforts have been made on acoustic scene classification in the audio signal processing community. In contrast, few studies have been conducted on acoustic scene clustering, which is a newly emerging problem. Acoustic scene clustering aims at merging the audio recordings of the same class of acoustic scene into a single cluster without using prior information and training classifiers. In this study, we propose a method for acoustic scene clustering that jointly optimizes the procedures of feature learning and clustering iteration. In the proposed method, the learned feature is a deep embedding that is extracted from a deep convolutional neural network (CNN), while the clustering algorithm is the agglomerative hierarchical clustering (AHC). We formulate a unified loss function for integrating and optimizing these two procedures. Various features and methods are compared. The experimental results demonstrate that the proposed method outperforms other unsupervised methods in terms of the normalized mutual information and the clustering accuracy. In addition, the deep embedding outperforms many state-of-the-art features. [ABSTRACT FROM AUTHOR]

Railway communications has attracted significant attention from both academia and industries due to the booming development of railways, especially high-speed railways (HSRs). To be in line with the vision of future smart rail communications, the rail transport industry needs to develop innovative communication network architectures and key technologies that ensure high-quality transmissions for both passengers and railway operations and control systems. Under high mobility and with safety, eco-friendliness, comfort, transparency, predictability, and reliability. Fifth-generation (5G) technologies could be a promising solution to dealing with the design challenges on high reliability and high throughput for HSR communications. Based on our in-depth analysis of smart rail traffic services and communication scenarios, we propose a network slicing architecture for a 5G-based HSR system. With a ray tracing-based analysis of radio wave propagation characteristics and channel models for millimeter wave (mmWave) bands in railway scenarios, we draw important conclusions with regard to appropriate operating frequency bands for HSRs. mymargin Specifically, we have identified significant 5G-based key technologies for HSRs, such as spatial modulation, fast channel estimation, cell-free massive multiple-input–multiple-output (MIMO), mmWave, efficient beamforming, wireless backhaul, ultrareliable low latency communications, and enhanced handover strategies. Based on these technologies, we have developed a complete framework of 5G technologies for smart railways and pointed out exciting future research directions. [ABSTRACT FROM AUTHOR]

An L-band geosynchronous synthetic aperture radar (GEO SAR) system has to be confronted by an intractable issue of the decorrelations imposed by ionospheric irregularities. On the one hand, the phase and amplitude scintillations will bring about the decorrelation within the synthetic aperture and result in azimuth-imaging degradation. On the other hand, the imposed scintillation history is spatially decorrelated across the ultra-large GEO SAR scene. In this article, a signal model of the GEO SAR acquisitionis established with the two-way ionospheric transfer function (ITF) modulation to incorporate these two types of decorrelations. This model meanwhile takes the anisotropic and flowing irregularities into account. By using this model, the L-band GEO SAR azimuth-imaging is evaluated in terms of five indexes, whose performances are dependent on nine ionospheric parameters. Furthermore, the spatial correlation of the phase and intensity scintillation histories is investigated for the L-band GEO SAR scene, both in simulation and statistics. The statistical result implies a sized scene, in which the phase scintillation history tends to be consistent. Finally, the interferometric performance is investigated between the pure and contaminated GEO SAR images. The simulation result shows that the degradation of the interferometric coherence results from the in-aperture decorrelation. [ABSTRACT FROM AUTHOR]

Science and technology innovation is a critical determinant of a nation's competitiveness in the global economy and a key driver of wellbeing for its residents. Building research partnerships with other nations can create and sustain a lasting competitive advantage and facilitate the international transfer of knowledge and skills. While it is tempting to continue to collaborate with established partners, emerging partners may represent a source of untapped benefit. In this article, we discuss how modeling approaches and techniques may be used to identify innovation partnerships. We describe a number of modeling methods that can be applied in isolation or in combination to inform the partnership identification and selection process. A multiplicity of available modeling tools exist, which can be applied depending on how the problem is scoped with respect to partners, science domains, time frames, and objectives. We develop an objectives hierarchy that can be used to direct future decision making efforts, and we contrast the assumptions and requirements underlying various methodological approaches. [ABSTRACT FROM AUTHOR]

The global science and technology landscape is rapidly changing. In an environment where collaboration is necessary for scientific and technological innovation, it is nonetheless challenging to identify and establish relationships between international partners, especially in emerging technological fields where no country has yet demonstrated a clear advantage. This article describes a model to facilitate a quantitative assessment and ranking of national innovation potential based on the integration of established measures (e.g., patents and peer-reviewed publication counts) as well as nonobvious or subjective measures (e.g., national attitudes toward particular veins of research) to predict future performance. Such an assessment can provide a helpful step toward improving projections of scientific performance, particularly for those countries with a relatively limited footprint in the scientific landscape. Modeling can help identify and engage newcomers early in the innovation process to maximize benefits from a potential bilateral or multilateral relationship. [ABSTRACT FROM AUTHOR]

The main challenging task in moving ground target detection with seismic signals is the extraction of robust feature vectors for accurate localization of a seismic event. Toward this goal, a new feature extraction technique is proposed for seismic event detection based on time-frequency analysis implemented with the Smooth pseudo Wigner–Ville distribution (SPWVD). The time-frequency coefficients are utilized to constitute a Renyi entropy vector exploiting a probability distribution function. This entropy vector gives information on the localized measurement of a possible seismic event resulting from moving ground vehicles. The likelihood of an actual seismic event is then ensured by using a constant false alarm rate (CFAR) detector which dynamically adjusts the threshold to minimize false alarms. The proposed algorithm is tested on the benchmarked data set SITEX02 consisting of seismic signals generated by moving tracked and wheeled vehicles. Additionally, the algorithm is also applied on a data set generated by collecting the seismic signature of a moving wheeled vehicle, i.e., bus within the campus. The results are also compared with pseudo Wigner–Ville distribution (PWVD) and short-time Fourier transform (STFT)-based seismic event detection technique. Significantly improved detection results are observed for SPWVD technique. An F-score improvement of ~24% as well as lead time enhancement of twofold is observed in comparison with the PWVD for relatively weak seismic signals generated by wheeled Dragon Wagon (DW) vehicles. Similarly, the F-score improvement is of ~13% and the lead enhancement is approximately 10 s in comparison with the STFT method. [ABSTRACT FROM AUTHOR]

In this article, we experimentally demonstrate a novel expanded nonorthogonal multiband super-Nyquist carrier-less amplitude and phase (m-ESCAP) modulation for bandlimited organic visible light communication (VLC) systems. The proposed scheme has the same bandwidth requirement as the conventional m-CAP while breaking the orthogonality between subcarriers by purposely overlapping them. We compare m-ESCAP with the conventional m-CAP and a compressed nonorthogonal version of m-CAP (m-SCAP) in terms of measured bit error rate (BER) performance, bit rates, and spectral efficiencies. We show that the m-ESCAP system offers improvement in the bit rate of $\sim$ 10% and 20% compared to the m-CAP and m-SCAP, respectively, and in the spectral efficiency of $\sim$ 20% compared to m-CAP. These gains are achieved at the cost of increased BER, which, however, remains below the 7% forward error correction limit. [ABSTRACT FROM AUTHOR]

Manufactory load allocation can be used as an effective industrial demand response scheme to reduce operating costs for industrial multienergy microgrids (iMEMGs). In addition, combined cooling, heat, and power (CCHP) plants with auxiliary devices can provide low-cost multiple energies for industrial plants. However, uncertain power generation from renewable energy sources impairs the iMEMG's operation, leading to challenges such as increased operating costs and energy supply deficiency. To conquer these challenges, this paper proposes a multitimescale coordinated adaptive robust operation approach where manufactory load allocation and iMEMG operation are optimally coordinated on different timescales. In the weekly scheduling stage, industrial loads and CCHP units are scheduled for the following week and the hourly iMEMG operation is optimized within the week. Besides, this paper applies an adaptive robust optimization method where the uncertain renewable power generation is fully addressed. The proposed approach is tested on an iMEMG with various industrial manufactories, and it is compared with conventional methods. The simulation results indicate that compared to the conventional ones, the proposed approach can guarantee a robustly optimal operation solution for the iMEMG against any uncertainty realization. [ABSTRACT FROM AUTHOR]

Traditionally, regularly spaced antenna arrays follow the spatial Nyquist criterion to guarantee an unambiguous analysis. We present a novel technique that makes use of two sparse non-Nyquist regularly spaced antenna arrays, where one of the arrays is just a shifted version of the other. The method offers several advantages over the use of traditional dense Nyquist-spaced arrays, while maintaining a comparable algorithmic complexity for the analysis. Among the advantages we mention: an improved resolution for the same number of receivers and reduced mutual coupling effects between the receivers, both due to the increased separation between the antennas. Because of a shared structured linear system of equations between the two arrays, as a consequence of the shift between the two, the analysis of both is automatically paired, thereby avoiding a computationally expensive matching step as is required in the use of so-called co-prime arrays. In addition, an easy validation step allows to automatically detect the precise number of incoming signals, which is usually considered a difficult issue. At the same time, the validation step improves the accuracy of the retrieved results and eliminates unreliable results in the case of noisy data. The performance of the proposed method is illustrated with respect to the influence of noise as well as to the effect of mutual coupling. [ABSTRACT FROM AUTHOR]

In this paper, we present a comprehensive framework for the performance analysis of dual-hop relaying communications with variable gain amplify-and-forward (AF) relays and operating in the presence of both multipath fading and shadowing, modeled by the Fisher-Snedecor $\mathcal {F}$ -distribution. Novel closed-form expressions for the probability density function (PDF) and the cumulative distribution function (CDF) of the end-to-end signal-to-noise ratio (SNR) of the considered system subject to hardware impairments are first derived. Single-integral expressions for the numerical evaluation of the $n$ -th moment of the end-to-end SNR, the outage probability (OP), the ergodic capacity under different adaptive transmission schemes, the effective capacity and the average bit error rate (ABER) of $M$ -ary modulation schemes are further presented. The proposed analytical expressions are valid for most of the well-known fading distributions, provided that the moment generating function (MGF) of the inverse SNR of each hop is readily available. For the special case of ideal hardware, it is shown that the above performance metrics can be expressed in closed-form. It is worth pointing out that the proposed analysis is valid even when the destination node is equipped with multiple antennas and maximal ratio combining (MRC) is employed. The correctness of the proposed mathematical analysis is validated through extensive numerically evaluated results accompanied with Monte-Carlo simulations. [ABSTRACT FROM AUTHOR]

The scalable high efficiency video coding (SHVC) is an extension of high efficiency video coding (HEVC). It introduces multiple layers and inter-layer prediction, thus significantly increases the coding complexity on top of the already complicated HEVC encoder. In inter prediction for quality SHVC, in order to determine the best possible mode at each depth level, a coding tree unit can be recursively split into four depth levels, including merge mode, inter2N×2N, inter2N×N, interN×2N, interN×N, inter2N×nU, inter2N×nD, internL×2N and internRx×2N, intra modes and inter-layer reference (ILR) mode. This can obtain the highest coding efficiency, but also result in very high coding complexity. Therefore, it is crucial to improve coding speed while maintaining coding efficiency. In this research, we have proposed a new depth level and inter mode prediction algorithm for quality SHVC. First, the depth level candidates are predicted based on inter-layer correlation, spatial correlation and its correlation degree. Second, for a given depth candidate, we divide mode prediction into square and non-square mode predictions respectively. Third, in the square mode prediction, ILR and merge modes are predicted according to depth correlation, and early terminated whether residual distribution follows a Gaussian distribution. Moreover, ILR mode, merge mode and inter2N×2N are early terminated based on significant differences in Rate Distortion (RD) costs. Fourth, if the early termination condition cannot be satisfied, non-square modes are further predicted based on significant differences in expected values of residual coefficients. Finally, inter-layer and spatial correlations are combined with residual distribution to examine whether to early terminate depth selection. Experimental results have demonstrated that, on average, the proposed algorithm can achieve a time saving of 71.14%, with a bit rate increase of 1.27%. [ABSTRACT FROM AUTHOR]

A codesigned hepta-band antenna with a large frequency ratio is presented by using a mode-composite method. The method is based on an idea that apart from a quasi-TE10 mode transmitting in a substrate integrated waveguide (SIW) structure at millimeter-wave frequencies, additional radiation modes of the wire antennas can be excited simultaneously on the outer surface of a well-designed SIW structure at the microwave frequencies much lower than its cut-off frequency. In this design, a combination of a meander SIW section and an L-shaped SIW section that are connected with a metallic ground plane constructs the mode-composite structure. Two slot antenna arrays with a size of $2\times4$ are implemented in the SIW sections for the radiation in the 28 GHz and 38 GHz bands. Unidirectional features with low backward radiation can be achieved with the aid of parasitic dipole elements close to the slots. A T-shaped probe operating as a coupled feed is applied to excite five radiation modes existing on the outer surface of the mode-composite structure, which can cover five sub-6 GHz bands for the LTE and 5G mobile communications and the WLAN applications. Stable radiation characteristics are maintained at the microwave frequencies. The work paves a new means to design the compact multiband antenna with a large frequency ratio, which would be attractive for future portable devices operating in both microwave and millimeter-wave bands. [ABSTRACT FROM AUTHOR]

A novel millimeter-wave aperture-coupled magnetoelectric (ME) dipole antenna with a low-profile metallic geometry is proposed based on the 3-D printing technology. An impedance bandwidth of 53.7%, a gain up to 10.8 dBi, and stable unidirectional radiation patterns with the cross polarization of less than −38 dB are achieved by the antenna. Its operating mechanism and design process are studied in detail as well. In order to widen the bandwidth of the millimeter-wave full-corporate H-plane air-filled waveguide feed network, a wideband H-plane stepped waveguide T-junction with a compact size is then investigated, which has a bandwidth of about 50% for the reflection coefficient of less than −20 dB. By combining the presented radiating element with the feed network, an $8 \times 8$ ME dipole array is designed, printed, and tested in the Ka-band. An enhanced bandwidth of 31%, symmetric radiation patterns with the cross polarization of less than −35 dB, and a gain varying from 25.5 to 28.5 dBi are experimentally obtained. The proposed design demonstrates the advantage of utilizing the 3-D printing method to improve the operating characteristics of the millimeter-wave antenna array and is attractive for the emerging wideband millimeter-wave wireless applications. [ABSTRACT FROM AUTHOR]

Real-time detection of sonar target plays a vital role in the underwater research field. Conventional deep learning methods need large quantities of sonar images as the sample for model training, and they cannot ensure detection speed and feature extraction ability simultaneously. For sonar dataset with small effective sample and low Signal-to-Noise Ratios (SNR), an improved YOLOv3 algorithm for real-time detection called as YOLOv3-DPFIN is proposed. The objective of the proposed YOLOv3-DPFIN is to accomplish the accurate detection of noise-intensive multi-category sonar targets with minimum time consumption. The proposed model conducts efficient feature extraction via the Dual-Path Network (DPN) module and the fusion transition module, and adopts a dense connection method to improve multi-scale prediction, which can complete precise object classification and location. The experimental results show that the algorithm achieves 84.4% ${mAP}_{75}$ with 56fps on a Nvidia Titan Xp, when testing on the sonar dataset and using the new VOC2012 mAP standard, which can meet the requirement of robust real-time detection for both raw and noised sonar targets. Moreover, the precision and speed of the proposed YOLOv3-DPFIN are superior to the original YOLOv3 model and state-of-the-art improved SSD models. [ABSTRACT FROM AUTHOR]

Penetration of renewable energy sources (RESs) and electrical energy storage (EES) systems in distribution systems is increasing, and it is crucial to investigate their impact on systems’ operation scheme, reliability, and security. In this paper, expected energy not supplied (EENS) and voltage stability index (VSI) of distribution networks are investigated in dynamic balanced and unbalanced distribution network reconfiguration, including RESs and EES systems. Furthermore, due to the high investment cost of the EES systems, the number of charge and discharge is limited, and the state-of-health constraint is included in the underlying problem to prolong the lifetime of these facilities. The optimal charging/discharging scheme for EES systems and optimal distribution network topology are presented in order to optimize the operational costs, and reliability and security indices simultaneously. The proposed strategy is applied to a large-scale 119-bus distribution test network in order to show the economic justification of the proposed approach. [ABSTRACT FROM AUTHOR]

This paper is devoted to the stability and stabilization for Takagi–Sugeno fuzzy systems with time-varying delays. First, an improved matrix inequality is presented to bound both strictly and nonstrictly proper rational functions, which is more general than the existing versions of reciprocally convex lemmas. Second, by suitable operations on parameter-dependent polynomial multiplied by state rate, a couple of novel intermediary-polynomial-based functions (IPFs) are developed in delay-product types. Benefitting from slack matrices of IPFs, a certain degree of flexibility is furnished. More importantly than that, by feat of adjustment of the variable parameter, the resulting conditions will be further endowed with additive freedom, which relaxes the feasible space in a distinctive manner. Third, by utilizing IPFs along with triple integrals, the stability criteria and the controller design approach are derived by some advanced integral inequalities. Resorting to elaborate construction of IPFs, the strengths of bounding techniques are sufficiently exploited, and the information on delay derivative is adequately reflected. Consequently, more desirable performances are achieved, while without excessive computational complexity. Finally, the effectiveness of the proposed methods is verified by numerical examples. [ABSTRACT FROM AUTHOR]

The extensive use of simulation-based serious games (SSGs) has made a revolution in educational techniques. As a potentially significant feature for SSG design and evaluation, the term fidelity (the similarity between an SSG and its real reference) emerges and attracts increasing attention. The study of fidelity not only benefits the design, development, and analysis of an SSG with the consideration of improving the learning effect but also contributes to the investment reduction of an SSG. However, the term fidelity is used inconsistently in the current literature. The introduction of new technologies (e.g., virtual reality) and the blend of multiform SSGs also facilitate the extension of fidelity with new connotations. All lead to confusing concepts and vague measure metrics. Besides, the relationship between fidelity and learning effect is still uncertain. A new vision and a comprehensive conceptual framework of fidelity for more general applications are in need. In this paper, further exploration and discussion of these issues in relation to fidelity of SSGs are presented through a systematic review. A general conceptual framework considering both aspects of the SSG system itself and the learners is developed and applied to analyze fidelity in SSGs. Based on that, a discussion on fidelity-related issues of SSG design and development is presented. [ABSTRACT FROM AUTHOR]