Search

Transient recovery characteristics of DC voltage are critical to analyze the stability criterion and the fault ride-through capability of voltage source converter (VSC)-based DC power systems. In this paper, the dynamic model of a flexible DC power grid was initially established, and then, the dynamic operation mechanism of the system was presented by analyzing the movement locus of the operating point when disturbances occur in the DC grid. Subsequently, the DC transient voltage limit and the DC fault clearing time were calculated in order to obtain the transient stability criterion of the DC power system. In addition, a novel voltage transient recovery control strategy was proposed by improving the droop control of the DC voltage to extend the transient stability margin of the power system. Finally, a typical multi-terminal flexible DC power grid was developed on a hardware-in-the-loop platform. The results demonstrated that the voltage recovery ability of the DC power system can be quantitatively estimated by the stability criterion, while the system transient stability can be improved by the proposed control strategy.

In view of the dynamic instability of rock mass in high geostress areas during underground engineering excavation, the comprehensive weight and extension methods are adopted to research the rockburst prediction. Firstly, five main influencing factors including uniaxial compressive strength, stress coefficient, brittleness coefficient, elastic energy index, and integrity of rock mass are used as the evaluation indexes of rockburst prediction according to the conditions required for rockburst occurrence. The assessment index system of rockburst intensity is constituted. Secondly, the analytic hierarchy process (AHP) and variation coefficient methods are used to determine the comprehensive weight of evaluation index, and the rockburst prediction model is established based on the extension evaluation method. Thirdly, the parameter programming and numerical calculation of the proposed prediction model are carried out in the MATLAB software. The user visualization execution window and software system of rockburst prediction model are realized. Finally, the software system is applied to the rockburst prediction in the water diversion tunnels of Jiangbian hydropower station and Jinping secondary hydropower station. The prediction results are compared with the actual situation and other evaluation methods. The results show that (i) the establishment of the user visualization window realizes the visualization and systematization of rockburst prediction model, which improves the data import rate and calculation efficiency. (ii) The prediction results of the proposed software system agree well with the actual situation, and they are more accurate than other evaluation methods. (iii) The proposed software system of rockburst prediction can also be used in coal mine, metro, and other underground projects, which has good engineering application values.

Hybrid dc circuit breakers (HCBs) usually require a large number of insulated-gate bipolar transistors (IGBTs) with high turn- off capability for series and parallel connections. This requirement entails excessive cost. Compared with IGBTs, integrated gate-commutated thyristors (IGCTs) are another kind of fully controlled power device with high surge current and voltage capability but are low cost. However, the weak turn- off capability of IGCTs hinders their application to high-powered HCBs. On the basis of the complementary advantages of IGBTs and IGCTs, this paper proposes a mixture solid-state switch (MSS) for HCBs that uses the switching action of several IGBTs to create a low-current condition that favors the turning off of large-scale series-connected IGCTs. The MSS topology can significantly reduce cost while maintaining the excellent performance of traditional solid-state switches. This study also introduces two different turn- off modes for MSS, namely, single oscillation and multiple oscillations. Finally, 25 kV/20 kA prototypes in both modes are developed to verify the effectiveness of the MSS topology.

High-voltage dc (HVdc) circuit breakers (CBs) are a key technology of multiterminal dc (MTdc) systems and grids. In this article, a state-of-the-art 500-kV hybrid CB (HCB) topology is proposed and compared with other existing HCB solutions. Implementing the topology includes three important components: a modular designed high-current bidirectional solid-state switch (SS), an ultrafast mechanical switch (MS) based on a Thomson (CCDC) equipment with no conduction loss, whose corresponding functional verification test results have been presented in detail. On this basis, the overall 500-kV HCB and a comprehensive experimental platform are developed. Detailed experiments, including 25-kA turn-off and reclosing experiments, are conducted. The HCB was found to be capable of breaking a 25-kA current in 2.7 ms with 800-kV overvoltage, which is on par with the level of leading CBs worldwide.

A dc chopper (DCC) is one of the key equipment in voltage-sourced converter-based high-voltage dc transmission systems with offshore wind power access. Combining the complementary advantages of concentrated and modular DCCs, a novel DCC with modular solid-state switches and a concentrated resistor (MC-DCC) is proposed. The non-long-term work attribute of the DCC provides opportunities to apply metal oxide varistors in modularizing series solid-state switches, thereby removing the capacitors for voltage balance. A method of designing the MC-DCC is introduced on the basis of the analysis of system requirement. Finally, prototype experiments are conducted to verify the effectiveness of the MC-DCC.

At present, it is usually necessary to use fast recovery type diodes to cooperate with the fast switching actions of fully controlled devices such as insulated-gate bipolar transistors (IGBTs). However, the specific nature of the working conditions of the diode-bridge bidirectional solid-state switch (DBSS) provides an opportunity to apply standard recovery diodes for this purpose, which helps to reduce the cost of the bidirectional switch. This letter presents the self-oscillation phenomenon that is caused by the recovery processes of the diodes in a DBSS. Then, the recovery effects of these diodes in different topologies are analyzed from the perspective of the DBSS and optimized designs for both the snubber circuit and the energy-absorbing circuit are proposed. A 500-kV dc breaker is realized based on the proposed DBBS using the standard recovery diodes, which can work reliably with IGBTs, and the cost of the resulting dc circuit breakers is greatly reduced.

Different from the continuous-switching behavior of series power electronic devices in converters, the single-switching attribute of dc circuit breakers (DCBs) provides opportunities to integrate metal oxide varistors (MOVs), balancing circuits, and series power devices. This paper proposes a modular design methodology based on discrete MOVs of a DCB with series power electronic devices. The modular dc circuit breaker (MDCB) topology and voltage balance issues, the steep front effect and solution, and the design method are proposed and analyzed in this paper. Compared with the traditional DCB, the MDCB has a self-balancing ability, thereby removing the voltage-balancing circuit and reducing the number of series modules. The standard designed cell can be operated as an independent DCB or compose a high-voltage DCB in series flexibly, which decreases the difficulty of design and commission. Finally, a series experimental platform based on an insulated-gate commutated thyristor, an insulated-gate bipolar transistor, and an injection-enhanced gate transistor is built to verify the correctness and effectiveness of the proposed design methodology.

Fault monitoring of multiphase batch process is a difficult problem in multivariate statistical process monitoring. It needs to consider not only the process monitoring under stable mode, but also the transition mode with strong dynamic nonlinearity. Since the data has different correlations under different operating modes, it is necessary to establish different monitoring models for each process mode, especially the transition process between stable modes. The biggest feature is the dynamic characteristics of the variables. This feature can be better reflected in this transition using a time-varying covariance instead of a fixed covariance during the transition phase. In this paper, a new strategy for batch process sub-phase partition and process monitoring is proposed. Firstly, the three-dimensional data matrix is expanded into a new two-dimensional data according to the time slice expansion strategy. Secondly, the data of each time slice is transformed by Kernel Entropy Component Analysis (KECA), and then the production process is divided into phases according to the spatial angle of the kernel entropy. The production operation process is divided into a stable phase and a transition phase, and monitoring models are respectively established to monitor the production process; Finally, the application of the penicillin fermentation simulation platform shows that the Sub-MKECA phase partition results can reflect the mechanism of the batch process well, and the fault monitoring of the process shows that it can detect faults in time and accurately, and has high practicality value.

The real-time requirements of tool wear states monitoring are getting higher and higher, at the same time, tool wear monitoring lacks a modeling data comprehensive carrier, which hinders its application in the actual machining process. In order to solve this problem, combining the high fidelity real-time behavior simulation characteristics of digital twin (DT) and the powerful data mining capabilities of artificial intelligence, an online tool wear monitoring method based on DT and Stack Sparse Auto-Encoder-parallel hidden Markov model (SSAE-PHMM) was proposed. First, a DT which can reflect the real state of the tool was established, and the tool wear state was predicted by visual display and analysis in the virtual space; Second, a tool wear state recognition model based on SSAE-PHMM was established, which can adaptively complete time domain feature extraction. And for each tool wear state, multiple HMM models were combined into a PHMM model to realize accurate recognition of tool wear state. PHMM overcome the defects of poor convergence and long training time of artificial neural network, and greatly improved the performance of classifier. Through the deep integration of DT and artificial intelligence, real-time data-driven tool wear qualitative and quantitative online monitoring was realized, and the effectiveness of this method was verified by experiments.

Under the background of increasingly serious environmental pollution, virtual power plants have become an effective way to solve environmental pollution due to the characteristics of integrating a large number of clean distributed energy generation. At the same time, electric vehicles with dual attributes of power supply and load bring opportunities for the further development of virtual power plants. In this paper, the ideal matter-element comprehensive benefit evaluation model of virtual power plants is constructed by constructing the index system and weighting model of virtual power plants considering electric vehicles. The virtual power plants connected with different proportion of electric vehicles are taken as an example analysis. The results of the example analysis show that the comprehensive benefit of virtual power plants is the highest when the proportion of electric vehicles access reaches 120%.

The inertia support of DC microgrid system is less, the instability of output power of distributed generation and the sudden change of load cause the sharp fluctuation of DC voltage, which endangers the safe and stable operation of DC microgrid. In order to solve the problem of small inertia of DC microgrid, a virtual inertial control strategy with exponential change is proposed based on U-I droop control to slow down the voltage shock caused by disturbance of the system. The DC microgrid system is built in MATLAB/Simulink. The simulation results show that the proposed strategy can increase the inertia of DC system for a short time, suppress DC voltage oscillation, effectively improve the transient response of DC voltage disturbance and improve the operation stability of the system when the power shortage suddenly occurs.

Significant advances in face recognition tasks have been made with the development of convolutional neural networks. With large scale convolutional neural networks we are now able to achieve 99% accuracy on many public face test sets. However, the face recognition networks commonly used for mobile are limited by their narrow network width which makes it difficult to learn better distributed features in face recognition feature embedding learning. In this work, based on MobileFaceNets we propose an attention-guided hierarchical feature fusion model. Compared with the traditional cascading convolutional neural network classification model, our approach can explicitly consider both global and local features. And our approach requires only a small increase in computation to obtain better results than directly increasing the feature dimension.

We present joint multi-dimension pruning (abbreviated as JointPruning), an effective method of pruning a network on three crucial aspects: spatial, depth and channel simultaneously. To tackle these three naturally different dimensions, we proposed a general framework by defining pruning as seeking the best pruning vector (i.e., the numerical value of layer-wise channel number, spacial size, depth) and construct a unique mapping from the pruning vector to the pruned network structures. Then we optimize the pruning vector with gradient update and model joint pruning as a numerical gradient optimization process. To overcome the challenge that there is no explicit function between the loss and the pruning vectors, we proposed self-adapted stochastic gradient estimation to construct a gradient path through network loss to pruning vectors and enable efficient gradient update. We show that the joint strategy discovers a better status than previous studies that focused on individual dimensions solely, as our method is optimized collaboratively across the three dimensions in a single end-to-end training and it is more efficient than the previous exhaustive methods. Extensive experiments on large-scale ImageNet dataset across a variety of network architectures MobileNet V1&V2&V3 and ResNet demonstrate the effectiveness of our proposed method. For instance, we achieve significant margins of 2.5% and 2.6% improvement over the state-of-the-art approach on the already compact MobileNet V1&V2 under an extremely large compression ratio., Accepted to IEEE Transactions on Image Processing (TIP) 2021

In order to improve the tracking accuracy of surface target with multiple cruise missiles, an adaptive fusion tracking algorithm based on multi-missile network is proposed. For this method, optimized arrangement of chief missiles and attack missiles is performed to improve the penetration capability of multiple cruise missiles firstly. Then time alignment and space alignment is adopted to reduce the difference multiple cruise missiles. Finally, adaptive fusion tracking algorithm is performed to further improve the tracking accuracy of multiple cruise missiles based on the statistic threshold metric. Simulation results show that, target tracking can be finished more effectively than existing methods by using the proposed algorithm.

Aiming at the complex form of complex deception jamming signal and the difficulty of recognition, a hybrid deception jamming recognition and suppression algorithm based on multi-dimensional feature extraction is proposed. Firstly, using the uniqueness of the spatial position coordinates of the false target in the radar network, the spatial characteristics of the composite interference are extracted, and the distance false target, the angular vacation target and the distance-angle compound deception jamming are identified and suppressed. Then, the motion feature is further introduced. Using the difference characteristics of radial velocity estimation and Doppler velocity measurement, the distance-velocity deception jamming, angle-speed deception jamming and distance-angle-speed deception jamming are identified and suppressed. Finally, by extracting multiple times of compound deception jamming Features to further improve the effectiveness of the identification of compound fraud interference. The simulation results show that the algorithm can effectively identify and suppress the composite deception jamming.

In order to alleviate the impact of wind power randomness on the safe operation of the system, the wind power penalty cost is introduced, and the joint scheduling optimization model of wind power and pumped storage power station is established. The minimum total power generation cost is taken as the objective function in the proposed model with considering the network security constraints under N-1 conditions and a series of constraints such as wind power, thermal power, and pumped storage. The simplified N-1 safety verification method is adopted to improve the efficiency of the solution. The improved IEEE30 node system is used as the simulation system to analyze the system scheduling results under different scenarios. The simulation results show that the model can prevent the problem of line over-limit caused by the untimely adjustment of the unit's output. Under the condition of meeting the corresponding safety constraints, the operating cost of the system is reduced.

When a turn to turn fault takes place in a dry-type iron-core reactor, the circulating current in the shorted turns is ten times or tens of times larger than the rated phase current, which could cause insulation and structure damage of windings. However, the changes of terminal voltage and phase current are very small, thus they cannot be regarded as the reliable criterion of turn to turn fault protection. In this paper, the characteristics of the magnetic field distribution in the reactor are analyzed as a criterion for the detection of turn to turn fault and its detection algorithm is further proposed. To investigate the effects of different numbers of shorted turns and fault regions on the spatial magnetic field distribution of the iron-core reactor, various finite element analysis simulations were performed by Maxwell. It provides a theoretical basis for the study of turn to turn fault detection algorithm. Simulation results verify the effectiveness and feasibility of the proposed detection algorithm in this paper.

Rectangular harmonic model is a common model in UAV geomagnetic navigation. In order to reduce the navigational error caused by the model factors and achieve high-precision navigation, the original Earth’s main magnetic field model is replaced by rectangular harmonic model. In creating a rectangular harmonic model, inertia weight adaptive particle swarm algorithm is adopted to solve the problem that too many undetermined coefficients can easily fall into local optimum in the central region. The ability of generalization to inhibit the boundary effect of radial basis function neural network is used in the border region. Meanwhile, the design of the error box to a more reasonable distinction between center and boundary is designed. The simulation results show that the optimized combination of the two intelligent algorithm makes the accuracy of the model further higher.

Light stripe centerline extraction is the basic and key procedure in line-structured laser three-dimensional (3D) scanner. Based on the fact that light stripe's contour is approximately parallel to its centerline, a novel contour polygon segmentation method is proposed for extracting and optimizing centerline. Different light stripe segments are identified in images by contour tracking, and then each of them is segmented into several parts using contour polygonization. Interior angle is defined to trim open light stripe polygon and contour to make sure that centerlines extracted from open light stripes do not include superfluous points. Taking advantage of polygon segmentation, piecewise polynomial fitting method and self-adaptive interpolating strategy are adopted to acquire smoother and evenly spaced centerline points. Simulated experiments show that the proposed method can calculate centerlines from images robustly with a 0.309 pixel average accuracy. Point clouds and surface models of different objects acquired by a line-structured laser scanner demonstrate that the proposed method can produce more complete and smoother 3D models compared to other classical methods. Processing time for the proposed method is approximately positive proportional to the number of pixels in image. Quantitative analysis of time used for each sub-procedure puts forward an improvement direction for the proposed method. Based on that light stripe's contour is approximately parallel to its centerline.Segment light stripe into parts by contour polygon representation.Interior included angle is defined to trim open light stripe's polygon and contour.Piecewise polynomial fitting and self-adaptive interpolating strategy are adopted to generate optimized centerline points.Extensive experiments validate the method's performance including accuracy, robustness and speed.

In the power system with high penetration of renewable power generators (RPG), although frequency performance can be improved by introducing virtual inertia control, the swing stability of the rotor angle is likely to be reduced due to the changed inertia distribution. The motion equation of an interconnected power system with RPGs is established firstly, and the transient energy function is then derived to analyse the impacts of virtual inertias on transient energy conversion. In this paper, the transient process of the rotor angle after the disturbances is divided into two periods of first swing and multi-swing, and the effects of virtual inertia on the system stability are then analysed in the two periods, respectively. Under the virtual inertia control, the RPGs are utilized not only to share accelerating energy with synchronous generators (SG), but also to provide additional decelerating energy for the reduction of the rotor angle variation during the first swing period. In addition, based on the eigenvalues analysis, the effects of virtual inertias on system damping are evaluated theoretically, and the virtual energy of RPGs is then regulated to damp the oscillations during the subsequent multi-swing period. Experimental tests on a hardware-in-the-loop platform of a two-area interconnected power system with high wind power penetration are carried out to validate the proposed control strategy. The results demonstrate that in addition to the basic frequency support function, the swing stability of rotor angle is also improved, thus greatly improving the reliability of the virtual inertia support.

DC fault protection is one of the biggest challenge to multi-terminal direct current (MTDC) systems, because DC current has no natural zero crossing point. To overcome this problem, hybrid DC circuit breaker (DCCB), which serves as a promising solution to the isolation of DC faults, has been developed. However, hybrid DCCB contains a huge number of power semiconductors, which is costly and inefficient especially in MTDC. This paper proposed a novel hybrid DCCB for nodes in MTDC system. By optimizing the topology and corresponding control strategy of traditional hybrid DCCB, the requirement of power semiconductors is reduced without degrading the performance. The novel hybrid DCCB is verified by simulations of a four-terminal radial DC grid with MATLAB/Simulink.

State of the art glasses-free multi-view 3D displays suffer from the number of views/resolution tradeoff due to the dependence on spatial image multiplexing only. To overcome this trade-off we are developing spatiotemporal multiplexed display systems where the fast frame rate potential of organic light emitting diode (OLED) displays is utilized. By producing narrow viewing zones in the viewing field and deflecting these laterally over time with the use of a special optical screen dense viewing zones are formed by rapidly updating the image displayed. The operation of the system is being verified on a 9-view display based on a 144 Hz liquid crystal display (LCD) monitor. In parallel with this a fast OLED display and driver is under development which should be capable of running in excess of 1 kHz.

Sentiment analysis is an important task in natural language processing, which has promises great value to areas of interests such as business, politics and other fields. The prevalence of the internet has caused people to prefer expressing their opinion and sentiment on the Internet via methods such as tweeting on social media and commenting on products. However, the discourse of users on social media are usually short either due to limitations on post size or time limitations of the users. What emerges from these features is that the grammar used in these posts and the word meanings are flexible, which makes sentiment analysis difficult. Therefore, sentiment analysis is an important yet challenging task. In this paper, we propose a context-based regularization classification method for short text sentiment analysis. Specifically, we use contextual knowledge obtained from the data to improve performance of the sentiment classification. In this paper, the contextual knowledge includes two parts: word-sentiment knowledge and word-similarity knowledge. Moreover, we propose methods to calculate the sentiment of the words and the similarity between words on semantic level. Specifically, on the one side, we use a TRSR method based on the TextRank algorithm to rank words in each sentiment sample to determine the sentiment polarity of each word. On the other, we calculate the similarity between words according to word-embedding. In this way, we can determine the similarity between words and the sentiment polarity of a word. We then incorporate the contextual knowledge as a regularization into a supervised classification framework, which then converts into an optimization problem to train a more accurate model. Experiments on both Chinese and English datasets outperform than other baseline approaches, which demonstrates our method to be stable and effective.

In view of tracking problems of hypersonic target in near space, such as large manoeuvring and difficulty to measure and so on, a new Sine tracking model based on characteristics analysis of gliding trajectory is proposed. Firstly, target of gliding trajectory has characteristics such as long distance, low energy costs and strong defending, and flight forces and motion are mainly analyzed, to reduce measure errors for model lost correlation. Then based on the full analysis of target hypersonic characteristics, non-periodic random, Sine model is built scientifically, to solve the problem that routine tracking model is only able to track low dynamic changing target. Finally, Sine model tracking on ECEF (Earth-centred Earth-fixed) coordinate is used to further improve on the hypersonic target tracking. Compared with other simulation models such as Singer and Jerk, Sine model will further improve the precision of locating and tracking.

Track initiation algorithm based on Hough transform is an effective method on condition of strong clutter environment. But due to the influence of measurement error and the selection of parameter space threshold, the problems of clutter points mix in target tracks and multiple tracks initialized by only one target are existed. In order to solve these problems, this paper proposes a track initiation algorithm based on Hough transform and space accumulation. The algorithm firstly using Hough transform for the preliminary screening of the original data. And then the obtained primary measurement data would be accumulated on data space. Finally, the threshold detection is used to judge whether to initial track. Algorithm effectively reduces the clutter points mixed in target tracks and solves the problem of multiple tracks initialized by only one target. Simulation results show that the algorithm can initial the target track effectively.

Aiming at the range ambiguity problem of radar for near space hypersonic target, the existing track before detect (TBD) algorithm cannot detect the target effectively through the long time accumulation, a kind of Hough transform (HT) TBD (HT-TBD) algorithm based on multiple frequency and multiple hypotheses is proposed. Firstly, several pulse repetition frequencies (PRF) are worked alternately to detect the target and get the initial measurement data. At this time, the distance measurement of the initial measurement is ambiguous. Then, the multiple hypotheses extending is made on all the possible ambiguous intervals to get the multiple hypotheses extending measurement data. Finally, the extended measurement data are used for time-radial distance plane HT. The target track could be obtained after HT-TBD algorithm, and the range ambiguity could be resolved at the same time. The validity of the algorithm is verified by simulation, and the algorithm performance under different SNR and clutter density is detailedly analyzed. The simulation results show that the near space hypersonic target under the condition of range ambiguity can be effectively detected according to the algorithm in this paper.

The management of teaching ability evaluation is the main part of the personnel training target, which is formulated by the school, this article unifies the practice of classroom teaching ability evaluation, research has analyzed the key technologies about the design of the evaluation index system, the theory of support vector machine, incremental learning and so on; it has established the Model of Teaching Ability Evaluation based on the TSVM, and applies it to the School Teaching Ability Evaluation System. The experimental studies have shown that this model can obtain more desirable results.

With the rapid development of Internet and Web technology, Web page has become a main carrier of information publishing. In connection with the problems of current complex implementation, high error rate and low extraction speed of Web information extraction technology, this paper proposes a new method of Web extraction based on the characteristics of structure of Web page. This method is to use tree structure of DOM (Document Object Model) when analyzing web page, parsing the Web page into DOM tree to sort the scattered web pages, by the using of the characteristics of Chinese web pages similar in information structure and aggregated distribution to achieve simply with good versatility. At the same time, this method can reduce the complexity when dealing with the structure of web page and increase the speed of the Web information extraction. At present, the method has been applied to the news page automatic classification system, which is good to meet the system's requirements.

The paper introduced the present insufficient situation of the frequency spectrum resource. Several effective techniques to raise the utilization efficiency of the spectrum and corresponding developments of their applications, such as the Cognitive Radios, the Ultra-Wide Band radio, the Water Filling Principle and the Detection & Avoidance Mechanism, are analyzed and compared, which are increasingly significant to the spectrum regulation and management research.

Ultra wideband (UWB) system operates by transmitting signals over a very wide band of spectrum, so there must be the problem of coexistence of UWB with other radio services besides the spread spectrum communication system. By designing the simulation models of UWB system and spread spectrum system which is used more broadly, the simulation experiments has been taken on the performance of spread spectrum under the interference from UWB. The results prove the validity of simulation model. And the main factors are analyzed on UWB interference to the spread spectrum communication, and the corresponding results are given.

As Virtual Reality(VR), Augmented Reality(AR), Mixed Reality(MR) technology becomes more accessible, it is important to explore VR/AR/MR technologies that can be used for remote collaboration on physical tasks. Previous research has shown that gesture-based interaction is intuitive and expressive for remote collaboration, and using 3D CAD models can provide clear instructions for assembly tasks. In this paper, therefore, we describe a new MR remote collaboration system which combines the use of gesture and CAD models in a complementary manner. The prototype system enables a remote expert in VR to provide instructions based on 3D gesture and CAD models (3DGAM) for a local worker who uses AR to see these instructions. Using this interface, we conducted a formal user study to explore the effect of sharing 3D gesture and CAD models in an assembly training task. We found that the combination of 3D gesture and CAD models can improve remote collaboration on an assembly task with respect to the performance time and user experience. Finally, we provide some conclusions and directions for future research.

Most of the code in typical Node.js applications comes from third-party libraries that consist of a large number of interdependent modules. Because of the dynamic features of JavaScript, it is difficult to obtain detailed information about the module dependencies, which is vital for reasoning about the potential consequences of security vulnerabilities in libraries, and for many other software development tasks. The underlying challenge is how to construct precise call graphs that capture the connectivity between functions in the modules. In this work we present a novel approach to call graph construction for Node.js applications that is modular, taking into account the modular structure of Node.js applications, and sufficiently accurate and efficient to be practically useful. We demonstrate experimentally that the constructed call graphs are useful for security scanning, reducing the number of false positives by 81% compared to npm audit and with zero false negatives. Compared to js-callgraph, the call graph construction is significantly more accurate and efficient. The experiments also show that the analysis time is reduced substantially when reusing modular call graphs.

This paper provides a review of research into using Augmented Reality (AR) and Mixed Reality(MR) for remote collaboration on physical tasks. AR/MR-based remote collaboration on physical tasks has recently become more prominent in academic research and engineering applications. It has great potential in many fields, such as real-time remote medical consultation, education, training, maintenance, remote assistance in engineering, and other remote collaborative tasks. However, to the best of our knowledge there has not been any comprehensive review of research in AR/MR remote collaboration on physical tasks. Therefore, this paper presents a comprehensive survey of research between 2000 and 2018 in this domain. We collected 215 papers, more than 80% of which were published between 2010 and 2018, and all relevant works are discussed at length. Then we elaborate on the review from typical architectures, applications (e.g., industry, telemedicine, architecture, teleducation and others), and empathic computing. Next, we made an in-depth review of the papers from seven aspects: (1) collection and classification research, (2) using 3D scene reconstruction environments and live panorama, (3) periodicals and conducting research, (4) local and remote user interfaces, (5) features of user interfaces commonly used, (6) architecture and sharing non-verbal cues, (7) applications and toolkits. We find that most papers (160 articles, 74.4%) are published in conferences, using co-located collaboration to emulate remote collaboration is adopted by more than half (126, 58.6%) of the reviewed papers, the shared non-verbal cues can be mainly classified into five types (Virtual Replicas or Physical Proxy(VRP), AR Annotations or a Cursor Pointer(ARACP), avatar, gesture, and gaze), the local/remote interface is mainly divided into four categories (Head-Mounted Displays(HMD), Spatial Augmented Reality(SAR), Windows-Icon-Menu-Pointer(WIMP) and Hand-Held Displays(HHD)). From this, we can draw ten conclusions. Following this we report on issues for future works. The paper also provides an overall academic roadmap and useful insight into the state-of-the-art of AR/MR remote collaboration on physical tasks. This work will be useful for current and future researchers who are interested in collaborative AR/MR systems. Refereed/Peer-reviewed

In this paper, we describe a new Mixed Reality (MR) remote collaborative platform making use of 3D CAD models for training in the manufacturing industry. It enables a remote expert in Virtual Reality (VR) to train a local worker in a physical assembly task. For the local site, we use Spatial Augmented Reality (SAR) to enable the local worker see virtual cues without wearing any AR devices, leaving their user hands free to easily manipulate the physical parts. For the remote expert, we construct a 3D virtual environment using virtual replicas of the physical parts. We also report on the results of a usability study of the prototype. Refereed/Peer-reviewed

In AR/MR remote collaboration, a remote expert often has to demonstrate tasks for a local worker. To make this easier, we have developed a new adaptive MR remote collaborative architecture which enables a remote expert to guide a local user in physical assembly and training tasks. A remote user can activate the instructions through simple and intuitive interaction, then clear instructions are shown in both AR (local) and VR (remote) views, enabling a local worker to operate the tool following the instructions. In the demonstration we use a hammer operation as a physical task, showing the benefits of the adaptive MR remote collaborative platform Refereed/Peer-reviewed