最新刊期
卷 21 ， 期 1 ， 2022
- 摘要：In educational environments, as knowledge points are interrelated, a bunch of negative effects emerges without timely revision. Confusions will be accumulated and learning efficiency gets lower, so it is essential for student to solve their questions in time. Therefore analyze three key elements by up-to-date AI and deep learning technology, fully considering the subtle impact on knowledge digestion subjected to students’concentration, teaching quality and knowledge comprehension hardness. Integrating the implication of the three key elements above, self learning materials are generated and recommended considering students’personality, interests and habits, so as to acquire the knowledge efficiently and effectively through stylization. The experimental results show that the accuracy of text recognition is up to 92%and face recognition reaches 99%, which meet project requirements. Through survey in our university after tentative application of our platform, the average satisfaction rate is up to 83.4%. Students thought that the recommended materials were in accordance with their confusions and easy to accept.关键词：Smart Class;learning status analysis;deep learning;personalized recommendation
- 445
- |
- 77
- |
- 0
发布时间：2022-01-28 - 摘要：Data sparsity and cold start are the problems in recommendation system, which will lead to bad recommendation effect. Therefore, based on the end-to-end deep learning framework, propose a deep convolutional neural network (Course Knowledge Graph Convolutional Networks, KGCN-CR) that integrates course knowledge graphs (KG), KG enhances entity representation by gathering information about the neighborhood of course entities, which, the individualized potential interests of students are obtained. As an example, students and course interaction data and course attributes of the major in Computer Science and Art are crawled on the MOOC platform. The course attributes data crawled is aimed to be constructed a course KG. The course KG as the auxiliary information enhance the performance of course recommendation, by using 18, 135 pieces of interaction data respectively and 44, 600 course attribute for experimentation. The experiment results shows that the ACC and AUC of KGCN-CR are reached 82.3%and 78.2%, respectively, which are 15%higher than SVD, and the accuracy rate, recall rate and F1 value are also optimal. Even when the interactions between students and the coursesare extremely sparse, the good performance is maintained. The KG as an auxiliary information can effectively improve the performance of online courses recommendation, can solve the problem of data sparsity and cold start, and has better recommendation interpretability.关键词：knowledge graph;online course recommendation;neural network
- 223
- |
- 59
- |
- 0
发布时间：2022-01-28 - 摘要：To stimulate learner's interest in computational thinking training, gamification methods are introduced, and a scheme for comprehensively using development games and playing games for computational thinking training is proposed. In case of recursive thinking training, on the one hand, it cultivates computational thinking through game design.A Scratch game design tutorial suitable for the recursive thinking training is given. Learners develop Scratch game under the guidance of the tutorial, break down the problems, compare the similarities and differences, and conclude to complete the design. On the other hand, it cultivate computational thinking by game playing.A Tower of Hanoi game was developed using Unity game engine to cultivate recursive thinking. Complex and abstract problems are made clear and simple with the principle of recursion and regression. Through step-by-step guidance, the problem is broken down and the thinking process of computational thinking is repeated, allowing players to exercise computational thinking while playing games. The survey shows that the scheme proposed have significantly improved the effect of computational thinking training and learning interest.关键词：computational thinking;serious game;recursion;Tower of Hanoi
- 67
- |
- 14
- |
- 0
发布时间：2022-01-28 - 摘要：April 2021, EDUCAUSE released《2021 EDUCAUSE Horizon Report (Teaching and Learning Edition) 》. The report analyzes five macro trends, six key technologies, four future scenarios and five development cases of informatization in higher education. The report has the salient features. On the basis of the above analysis, puts forward the enlightenment and suggestions for the development of higher education informatization in China under the background of Education 4.0.关键词：Education 4.0;horizon report;higher education;education informationization
- 62
- |
- 15
- |
- 0
发布时间：2022-01-28 - 摘要：The advantages of AI teachers in knowledge retrieval, data storage, intelligent computing, interactive feedback have made their application research in the training of new engineering disciplines talents a hot spot. On the one hand, AI teacher can reduce teachers’pressure and expand teachers’teaching capabilities. On the other hand, it can assist students to carry out independent learning and inquiry, stimulate their sense of innovation. To this end, first analyzed the feasibility of applying AI+teacher to develop new engineering disciplines talents, then built its key technical framework. Finally, took the "Introduction to Artificial Intelligence" course as an example, the pre-class, in-class, and after-class were used as granular to carry out the staged teaching activity design in layers, which verify its promotion effect on the teaching of new engineering disciplines. Based on the above research, this paper has given some thoughts and suggestions on the application of AI+teacher to develop curriculum-based new engineering disciplines talents training.关键词：AI+teacher;instructional design;new engineering disciplines;talents cultivation
- 40
- |
- 15
- |
- 0
发布时间：2022-03-03 - 摘要：In the context of deepening the construction of smart campus, cloud computing, Internet of things, big data and other emerging technologies are constantly applied to school management and teaching. The unified identity authentication system is located in the platform support layer in the overall framework of smart campus, which provides identity authentication, application authorization, role management and other functions for the smart campus system. The login mode of unified identity authentication system needs to balance ease of use and security. This paper proposes an improved scheme of "WeCom+" unified identity authentication system, which effectively combines the OAuth identity authentication ability in WeCom platform with CAS protocol, unifies the identity authentication entrance of each system in smart campus, simplifies the login operation, and improves the security of user authorized login. After the application of the improved scheme in the author's company, it gradually replaced the authentication method of account password (accounting for more than 80%). Practice shows that the unified identity authentication scheme of "WeCom+" has a high promotion value for similar universities that use WeCom as the smart campus mobile portal.关键词：smart campus;WeCom;unified identity;OAuth2;CAS;single sign on;senseless login
- 32
- |
- 9
- |
- 0
发布时间：2022-01-28 - 摘要：The research of affective feedback in computer-supported collaborative learning, which combines affective computing, machine learning, and learning science research, is devoted to studying how to promote effective collaborative learning and provide affective support for learners. First, analyze the affective feedback system framework of CSCL, which consists of three parts: affective state acquisition, affective state analysis, and affective feedback processing. Then, this paper describes the representation of affective state and analyzes how to use multi-modal emotion recognition technology in CSCL. After that, this article introduces the key technologies of affective state analysis, which includes text-based sentiment analysis, affective visualization analysis, and interactive behavior analysis. Finally, it analyzes the key technologies of affective feedback processing from affective feedback strategies, affective reasoning models and related technical support, to provide clear research clues and directions for affective feedback research in computer-supported collaborative learning.关键词：computer-supported collaborative learning;affective recognition;affective state;affective feedback
- 33
- |
- 9
- |
- 0
发布时间：2022-01-28 - 摘要：Classroom debate, as one of important forms of classroom activity in a flipped classroom, can improve students'thinking ability, communication skills, and learning enthusiasm. Taking the teaching practice of Internet finance as an example, the teaching design of classroom debate in flipped classroom is carried out. According to the knowledge point of the definition of electronic currency in the MOOC, set up debate questions, discuss the key and difficult points of classroom debate in the flipped classroom, design the specific teaching steps of classroom debate and compares the distribution of the scores of two different teaching classes. The results show that the implementation of classroom debate in flipped classroom debate can improve students’performance and comprehensive ability.关键词：classroom debate;hybrid of online and offline teaching;flipped classroom
- 31
- |
- 14
- |
- 0
发布时间：2022-01-28 - 摘要："Internet+education" and "artificial intelligence+education" have become an important development direction of higher education in the world, and the improvement of technological level has also brought key breakthroughs to the learning revolution. The abundance of learning resources, the diversity of learning objects and the diversity of learning cultures brought about by the era of artificial intelligence provide many important opportunities for the transformation of learning styles of college students. In the strong base generation, the transition generation and the future generation, the iterative upgrading of artificial intelligence makes the transformation of students'learning style show typical gradual characteristics, and encourages students to change their learning concepts, break the barriers, and gradually move towards deep learning, boundary-free learning, mixed learning and community learning.关键词：artificial intelligence;learning mode;college student;reform
- 39
- |
- 14
- |
- 0
发布时间：2022-01-28 - 摘要：Augmented reality (AR) is an emerging educational technology, and its promotion and application in education and teaching has attracted the attention of scholars at home and abroad, but whether this technology can improve students'learning effects has always been a controversial issue. In view of this, this research adopts meta-analysis methods to comb and quantitatively analyze the international empirical literature on the impact of augmented reality technology on learning effects in the past ten years, and systematically examine the actual impact of augmented reality technology on students’learning effects. The study found that: on the whole, AR technology has a significant role in improving learning effects (ES=0.63) ; from the perspective of learning environment, AR has the best application effect in the school laboratory (ES=0.80) ; from the school stage In view of this, AR has the most significant impact on the learning effect at the higher vocational and high school level (ES=1.01) ; from the subject point of view, AR has an upper-middle-level impact on the learning of physics, science, medicine and mathematics (ES=0.64) ; From the perspective of teaching methods, the combination of AR and discovery method has the strongest moderating effect on learning effects (ES=1.13). Based on the above conclusions, this research proposes the following suggestions for the effective application of AR technology in education and teaching: First, design and develop appropriate AR applications based on different learning environments; second, select scientific and reasonable teaching methods based on subject characteristic; Third, research and develop AR teaching resources and learning support systems for different stages of learning.关键词：augmented reality technology;learning effect;empirical literature;meta-analysis;moderating effect
- 25
- |
- 6
- |
- 0
发布时间：2022-01-28 - 摘要：Taking literatures related to education artificial intelligence in CNKI and Wanfang database as data sources, adopting bibliometrics analysis and content analysis, using CiteSpace knowledge mapping analysis tool and Excel software, conduct a visual analysis of the literature related to educational artificial intelligence from 2011 to 2021. There are four characteristics of education AI research in China in the last 10 years. Firstly, there is an build-uping drift in the number of papers on educational AI research. It can be divided into germination and rapid development period; Secondly, from the perspective of publishing journals, periodicals of educational AI research can be roughly divided into three categories: Educational comprehensive journals, educational technology journals and primary and secondary school journals. Thirdly, from the perspective of the types of researchers, mainly including university teachers and primary and secondary school teachers, the research units are mostly concentrated in Beijing, Shanghai and other regions; Fourthly, from the perspective of keyword popularity, "artificial intelligence education" has the highest degree of centrality, followed by "deep learning" "artificial intelligence" "intelligent education" and other words. On this basis, combined with the development status of domestic educational artificial intelligence, it promulgates the frontier plumes and development trend of domestic educational artificial intelligence, in order to provide reference for the follow-up research and application.关键词：educational intelligence;artificial intelligence;knowledge map;CiteSpace
- 31
- |
- 13
- |
- 0
发布时间：2022-01-28 - 摘要：In recent years, a large number of program online evaluation systems for the teaching of computer courses are launched in universities to improve students'programming ability. In order to better clearly understand the impact of data in such systems on teaching, analyze the ontology contained according to the data of the program online evaluation system, and then use neo4j to construct the knowledge map after data cleaning. Through the visual analysis of learning path, learning situation analysis and teaching path, it is concluded that the knowledge map of program online evaluation system can express the teaching information in the form of knowledge map, so that learners can clearly understand their own level, improve their learning interest, and realize the high efficiency of teachers'teaching and students'autonomy of learning.关键词：knowledge graph;program online evaluation system;Neo4j graph database;visualization
- 22
- |
- 11
- |
- 0
发布时间：2022-01-28 - 摘要：For higher education, blockchain is a new technology that can promote education reform and reshape the ecology of education. Chain technology because of the limitation of technology blocks while facing changes in the field of education lack of motivation, education data storage bottlenecks, privacy, security hidden danger, virtual education data problem in aspects of property right is not yet clear, but does not hinder the chain block and the coupling of the higher education development trend of it in education information management, education resource sharing, lifelong learning system, and promote higher education democratization four aspects for the development of higher education brings new opportunity.关键词：blockchain;higher education;education informationization;education reform
- 46
- |
- 10
- |
- 0
发布时间：2022-01-28 - 摘要：In view of the existing recipe recommendation algorithms that do not fully utilize the interaction between users and recipes, a recipe recommendation algorithm based on heterogeneous network embedding of attention attributes was proposed. Firstly, one-hot coding is used to convert recipes into node features. Secondly, feature is transformed into feature embedding and base embedding of nodes through transformation function; at the same time, attention mechanism is used respectively in the same type of edge and between different types of edge to learn the preference weight of different users under different types of interaction during node aggregation, and the edge embedding of nodes is generated. The three kinds of embeddings are added to obtain the complete embedding of nodes. Finally, negative sampling optimization is carried out through random walk based on meta-path and skip-gram model to dig out the potential interaction possibility between nodes and finally complete the diet recommendation. Experiments are carried out with 10000 interactive data from the kitchen dataset. The results show that the ROC, PR and F1 values of the algorithm are improved by 2.5%, 2.8%and 2.3%respectively compared with the current leading algorithm.关键词：recipe recommendation;recommendation algorithm;network embedding;attention mechanism
- 29
- |
- 6
- |
- 0
发布时间：2022-01-28 - 摘要：Thyroid nodule is a common disease in clinical practice, the incidence of thyroid cancer caused by nodules has increased year by year. Thyroid ultrasound images have complex tissue structure, unclear edges, severe speckle noise, and low contrast, which brings great difficulties to doctors in diagnosing thyroid diseases. In this article, using the Mask R-CNN algorithm, combined with the migration learning method, pre-training the networks ResNet50, SENet and SE-ResNet50 on the ImageNet data set to obtain the pre-training weight parameters, which are used as the initialization parameters of the experimental model. And the optimized loss function is used in the method of fusion residual attention mechanism network SE-ResNet50 for feature extraction under the backbone network, the accuracy rate is 0.936, the recall rate is 0.851, the specificity is 0.948, the mAP is 0.824. The proposed method can assist doctors in diagnosing thyroid ultrasound diseases and has certain reference significance.关键词：Mask R-CNN;transfer learning;residual attention mechanism;optimized loss function;medical imaging
- 20
- |
- 7
- |
- 0
发布时间：2022-01-28 - 摘要：Word segmentation is the basic work of structural representation and deep mining of ancient Chinese medicine books. The supervised word segmentation of ancient Chinese medicine literature is simple and feasible, but it needs a lot of manpower and material resources, and has high professional threshold, strong subjectivity and poor expansibility. Therefore, this paper improves the textrank algorithm, and proposes the connectrank algorithm. According to the character connectivity, the unsupervised segmentation of ancient Chinese medicine books can be realized. Based on the data of 700 ancient Chinese medicine books, such as treatise on febrile diseases, Huangdi Neijing and Nanjing, the experiment is designed, which is 11.2%higher than the traditional unsupervised word segmentation method. The results show that the model has better word segmentation effect than other unsupervised word segmentation models.关键词：traditional Chinese medicine ancient literature;ConnectRank algorithm;unsupervised word segmentation;character connectivity;domain adaptability
- 14
- |
- 3
- |
- 0
发布时间：2022-01-28 - 摘要：Vehicle lane changing is a very frequent driving behavior, so it is necessary to predict the reliability of unmanned vehicle lane changing behavior. In order to better predict the reliability, UML activity diagram is used to model the lane changing behavior of unmanned vehicle based on long term memory network (LSTM). Considering the robustness of LSTM neural network, 6000 sets of experimental data are used to calculate the error permeability of LSTM modules, so as to measure the fault propagation probability between modules. The UML model is transformed into a discrete-time Markov chain (DTMC) model, and the reliability prediction value is obtained by prism. The experimental results show that the error permeability of LSTM module is 0.3025. When the DTMC model based on UML model mapping and component transfer diagram is used for prediction, the reliability values are 76.47%and 90.19%respectively. The results show that the error permeability of LSTM module can not be ignored in the reliability calculation of lane changing behavior of unmanned vehicle based on LSTM. The DTMC model obtained by mapping is more detailed to describe the modules, which is more suitable for the reliability prediction of unmanned vehicle lane changing.关键词：unmanned vehicle lane change;reliability;UML;error permeability;discrete-time Markov chain
- 16
- |
- 4
- |
- 0
发布时间：2022-01-28 - 摘要：The depth of myometrial invasion (MI) affects the treatment and prognosis of patients with endometrial cancer, commonly evaluated using MRI, which is greatly influenced by subjective factors. Based on MRI, propose a computer-aided diagnosis method for the depth of MI. This method only requires the corpus uteri region provided by humans or computers as input, which is easier to identify, and then it estimates the depth of MI automatically. First, the tumor region is segmented based on Otsu and morphological processing. Then the first order texture features and GLCM features are extracted. Finally, SVM is trained for the depth of MI classification. This method achieved an accuracy of 86.1%, sensitivity of 68.4%and specificity of 91.7%, which outperformed the commonly used classifiers. The results show that the proposed method is feasible for the auxiliary determination of the depth of MI and helpful to extract and fuse more kinds of features from tumor and myometrium to improve the classification performance in future work.关键词：SVM;feature extraction;depth of myometrial invasion;MRI;computer-aided diagnosis
- 14
- |
- 6
- |
- 0
发布时间：2022-01-28 - 摘要：The region proposal generation method (ie, PRN) based on convolutional neural networks (CNNs) is trained through instance-level annotations, and is also an important part of the current fully supervised target detection (FSOD). Because instance-level annotations are time-consuming and labor-intensive, while image-level annotations are easier to collect, so weakly supervised object detection (WSOD) that only uses image-level annotations has attracted the attention of many researchers. The current WSOD relies on standard region proposal generation methods such as selective search. These methods are prone to generate a large number of noisy proposal boxes, resulting in their existence that cannot fit the real target object. This paper is based on the multi-layer fusion of convolutional features and the segmentation guidance strategy to obtain high-quality proposal boxes. Specifically, the deep information of the convolutional network is used for multi-layer fusion, and the edge information is used to obtain the initial candidate proposal boxes, and then through weakly supervised semantic segmentation The consistency criterion divides the segmentation map into two variables, horizontal and vertical, to obtain the target consistency representation, thereby extracting high-quality proposal boxes. The experimental results on the PASCAI VOC2007 dataset show that the method in this paper exhibits excellent performance in classification and localization detection, with mean of average precision (mAP) and localization (CorLoc) reaching 51.0%and 71.2%accuracy rates, respectively.关键词：feature fusion;segmentation guidance;object consistency;weakly supervised object detection
- 14
- |
- 11
- |
- 0
发布时间：2022-01-28 - 摘要：With the rapid development of track transportation, the safe and effective track operation is an important guarantee. The emerging omni-directional intelligent techniques are being used to replace the previous detection performed by manual checks in rail transit operation.A real-time detection method based on the machine vision is improved for the interval spacing between basic rail and sharp rail. The turnout edge feature can be extracted and the accurate parameter space may be calculated. The approach contains four parts: image preprocessing, image edge detection, edge extraction, and edge straight line fitting. The preprocessing part uses the grayscale transformation and grayscale enhancement technology to simply process the image taken by the camera, so that the image edge features are more prominent. The image edge detection adopts the method based on canny edge detection operator, which can automatically calculate the threshold. Next, the improved Hough transform extract the switch feature. Finally, the edge fitting is carried out by the least squares method and convert the gap in the parameter space to complete the switch fitting detection. By comparing manual measurement with the machine vision method, it is obvious that the machine vision method can achieve 91.25%recall within 0.1mm tolerance.关键词：intelligent techniques;track transportation;track switch edge detection;edge straight line fitting
- 13
- |
- 8
- |
- 0
发布时间：2022-01-28 - 摘要：Text sentiment analysis extracts text features and classifies them according to the sentiment tendencies in the text. Research show that models based on recurrent neural network (RNN) and convolutional neural network (CNN) have good performance. In order to improve the performance of text sentiment classification, a hybrid network sentiment analysis model (BGC-ACES) combining attention and comparative reinforcement learning mechanism is proposed, and the comparison reinforcement learning mechanism is used for classification instead of a large number of complex calculations. After the model is vectorized through the embedding layer, CNN and bidirectional gated recurrent unit (BiGRU) are used to extract text features with different characteristics, and the features extracted by the two are fused, After feature fusion, the attention mechanism is introduced to judge the importance of different words to the meaning of the sentence; And then use the comparative enhancement learning mechanism to score the feature vector by comparing with the sample vector. The correct rate of BGC-ACES on the two sentiment analysis data sets reached 92.2%and 94.1%, respectively, increased 0.6%and 0.9%compared with the traditional models.关键词：sentiment analysis;convolutional neural networks;bidirectional gated recurrent unit;attention mechanism;comparison enhancement learning mechanism
- 12
- |
- 4
- |
- 0
发布时间：2022-01-28 - 摘要：Aiming at the problem of slow convergence and low recognition rate caused by using utterance-level feature parameter matrix as input of convolutional neural network.A speaker recognition based on 2DPCA (two dimensional principal component analysis) feature dimension reduction and CNN (convolutional neural network) was proposed in this paper. Firstly, each speech was divided into several frame-level speech, and the frame-level features of the same size were extracted to form the feature matrix. Then, 2DPCA was used to reduce the dimension of the feature matrix, and the principal component feature vectors were combined into a new feature matrix as the input of CNN. Finally, the speaker model was created through adaptive feature learning of CNN. The experimental results with CNN model based on Alexnet show that the running time is reduced by 57%and the recognition rate is improved after using the speaker recognition method.关键词：two dimensional principal component analysis;frame-level features;convolutional neural networks;speaker recognition
- 11
- |
- 3
- |
- 0
发布时间：2022-01-28 - 摘要：A reliable intelligent guided physical examination path optimization model was established based on greedy algorithm, and the physical examination time was optimized according to the physical examination sequence and time node information of multiple examination items. The physical examination data of physical examination center of Jiangsu Hospital of Traditional Chinese Medicine in March 2021 were taken as the initial data samples, and 29 126 data were selected as the research objects after data cleaning, and the least square method was used to fit the data to construct the functional relationship between the physical examination client arrival rate and time. The sum of squares of error (SSE) and R-Square were used as evaluation indexes. An intelligent guided physical examination model was established based on the sum of service time, queuing time and walking time.the research results show that the error sum of squares (SSE) of physical examination customer arrival rate were 5.115, 0.787 8, 4.541, etc., and the R-Square values were 0.979 6, 0.996 7, 0.898 6, etc., respectively. The fitting effect of the test model was significant. The intelligent guided physical examination model based on greedy algorithm can complete the planning of physical examination sequence and predict the physical examination path.关键词：intelligent guided physical examination;interpolation fitting;greedy algorithm;evaluation index
- 14
- |
- 4
- |
- 0
发布时间：2022-01-28 - 摘要：Optical music recognition is a key part of the development of intelligent music, and has important value in the fields of music teaching and creation. Aiming at the cumbersome steps and low recognition accuracy of the existing guitar tablature recognition methods, a guitar tablature recognition method based on deep learning is proposed. First, by analyzing the characteristics of the guitar tablature, the guitar tablature is divided into fret note image, minus time line image, rest image and increased time line image; then, the fret note image is superimposed with the minus time line image and input to the firstCRNN model for recognition, and superimposes the reduced time line image, the rest image and the increased time line image, and inputs them to the second CRNN model for recognition; finally, the recognized symbols are globally associated to obtain the complete musical score semantics. Through experimental analysis, guitar tablature recognition method based on deep learning can achieve 98.3%accuracy of character note recognition and 99.1%accuracy of time value note recognition, compared with the traditional guitar tablature recognition method, this method has faster recognition speed and higher recognition accuracy.关键词：optical music recognition;guitar tablature recognition;deep learning;CRNN;image recognition
- 12
- |
- 8
- |
- 0
发布时间：2022-01-28 - 摘要：Simultaneous optimization of multiple conflicting goals is called a multi-objective optimization problem. Multi-objective evolutionary algorithms are developed to solve these multi-objective problems. In the iterative process of evolutionary algorithm, the algorithm uses constant crossover factor and mutation factor, which obviously does not meet the characteristics of the iterative evolution of the population, so it is necessary to adjust the evolution direction of the population according to the convergence of the initial and late population solutions. At the same time, when using the boundary and cross aggregation algorithm, the clustering algorithm dominated by θ only refers to the vertical distance of the weight vector, but the linear distance from the projection point of the weight vector to the ideal point is not considered, which directly affects the individual. The convergence problem on the Pareto frontier. Using the improved adaptive population generation strategy to dynamically change the crossover probability and mutation probability, and adjust the evolution direction according to the current iteration of the population; by increasing the distance between the projection point and the ideal point and the penalty factor, calculate the individual to the center of aggregation The distance of the individual is randomly selected in the cluster, which effectively improves the convergence of the algorithm. By testing on the multi-objective problem test sets ZDT and DTLZ, NSGA-ACM has a better convergence and distribution effect on the solution set.关键词：multi-objective optimization;clustering algorithm;self-adaptive
- 16
- |
- 6
- |
- 0
发布时间：2022-01-28 - 摘要：In order to solve the problem of linear path tracking heading deviation and slow correction speed of heading deviation of underactuated water quality detection ship under the condition of external wind and wave disturbance and uncertain parameters of model system, a finite time control theory is proposed to design the heading control system. Firstly, the mathematical model of the course control of the underactuated water quality detection ship is established, and the complex nonlinear course control is simplified as a second-order system, which simplifies the design complexity. Secondly, the finite time course control law is designed based on the homogeneous theory, which improves the convergence speed and robustness of the system by fractional power, and estimates the error disturbance by combining with the extended state observer, which improves the control efficiency Performance. The simulation results show that the course angle error can be controlled within 0.1°by using the finite time course control, which can effectively enhance the anti-interference ability of the water quality detection ship and ensure the accuracy of operation path tracking.关键词：water quality detection ship;linear path tracking;course control;finite time control
- 8
- |
- 6
- |
- 0
发布时间：2022-01-28 - 摘要：The mixed-criticality systems' schedulability analysis is normally on the basis of worst case execution time (WCET), which leads to the problem of over-provisioning for resources and causes quite pessimistic analysis in low-criticality mode, In order to solve the problem of over-provisioning for resources and simplify the schedulability analysis method in low critical level mode. In this paper we propose a probabilistic Demand Bound Function (pDBF) model for the sporadic task sets scheduled by EDF scheduling policy on one processor.we present how to analyze and calculate the pDBF of the mixed-criticality system with an example, and the corresponding schedulability conditions are derived. An algorithm of schedulability testing with respect to sporadic task systems is designed with concerning maximizing the execution-budget of low-criticality task. Evaluations illustrate that our model can significantly improve the schedulability, meanwhile, reducing the complexity of the testing algorithm through the existing method. The experimental results show that the schedulability acceptance rate can be improved by 32%compared with the previous deterministic analysis. The proposed method can reduce the complexity of the analysis algorithm in pDBF model, and also improve the schedulability significantly.关键词：mixed-criticality;demand bound function;probabilistic analysis;schedulability;scheduling algorithm
- 8
- |
- 4
- |
- 0
发布时间：2022-01-28 - 摘要：In order to improve the instability of the edge network caused by the energy exhaustion of the nodes.A hybrid offloading strategy based on node energy in edge computing is proposed. This strategy mainly adopts a hybrid offloading mode of uplink and parallel, and transforms the offloading energy minimization problem into the residual energy maximization problem. The task is offloaded to a suitable node for execution by comparing the demand characteristics of the source offloading node (computing power, energy, etc.) with the resource characteristics of the offload node. Additionally, particle swarm optimization and simulated annealing algorithms are used to optimize HOS-NE and improve network stability, throughput, and load balance. Experimental results reveal that compared with random execution algorithm and greedy algorithm, HOS-NE based on intelligent algorithm optimization increases the network throughput and remaining energy by 16.7%and 28.6%, respectively, and reduces the number of packet loss to 5%~10%.关键词：mobile edge computing;residual energy;offloading strategy;particle swarm optimization algorithm;simulated annealing algorithm
- 14
- |
- 15
- |
- 0
发布时间：2022-01-28 -
Research on Local Discharging Positioning Method of Switchgear Based on Linear Array UHF Sensor AI导读
摘要：On-line partial discharge monitoring and positioning of the switchgear can detect faults and hidden dangers in time, which is of guiding significance for inspection personnel to perform further maintenance. The UHF method used for online monitoring of high voltage switchgear has obvious advantages such as high detection frequency and small external interference signals. However, the UHF method still has the problem that the number of equipment channels is limited in the partial discharge detection and positioning of the switchgear, and the multi-sided switchgear cannot be monitored at the same time. Based on this, propose a new linear array UHF sensor switchgear partial discharge online monitoring method. The linear array method is adopted to realize the simultaneous partial discharge detection of the multi-sided switchgear, and the envelope detection is used to shift the spectrum of the partial discharge signal to realize the function of dual-channel data acquisition. Experiments show that the maximum positioning error of this method for partial discharge signals is 13.4cm, and the average positioning error is 8.1cm. This method not only satisfies the detection and positioning of partial discharge in multi-sided switchgear, but also has good scalability and economy.关键词：UHF sensor array;switchgear;partial discharge location;online test;envelope detection;two-channel data acquisition;scalability- 8
- |
- 3
- |
- 0
发布时间：2022-01-28 - 摘要：To explore the process and effect of the diffusion of knowledge domain visualization software, and predict its future development trend; Based on the "diffusion of innovation" theory, take "data+graph" methods such as bibliometric method and main path analysis method to explore the diffusion law of knowledge domain visualization software in terms of diffusion curve, diffusion paths, and diffusion network. The results show that the knowledge domain visualization software’s diffusion process conforms to the S-curve rule. With the software’s update and iteration, the diffusion path of knowledge domain visualization software is divided into two key main paths, there are four core characters (Garfield E, Merigo J M, Van Eck, N J and Chen C M) and three important events (HistCite, CiteSpace and VOSviewer).关键词：knowledge domain visualization software;diffusion of innovation;diffusion curve;main path analysis
- 8
- |
- 4
- |
- 0
发布时间：2022-01-28 - 摘要：In view of the large number of single processing in the self-service germ rice machine system on the market, it is unable to ensure that citizens can eat fresh and healthy germ rice every meal. Based on the concept of "freshly milled rice, freshly eaten", a self-service rice milling system is designed based on the existing germ rice milling machine. The system realizes the function of storing brown rice and quantitative rice milling at any time by designing a quantitative rice pickup mechanism; designing a quantitative rice pickup motor driving circuit and a rice milling spindle motor driving circuit; designing a cloud server system, realizing remote control and local data networking based on MQTT IOT protocol by using WIFI module, realizing remote control of quantitative rice milling and data feedback. The system can meet the demand of "rationing each meal and eating now" and help to improve people's health.关键词：germ-remained rice machine;drive circuit;cloud server;MQTT protocol;IoT system
- 20
- |
- 7
- |
- 0
发布时间：2022-01-28 - 摘要：In order to improve the experience dominated by negative emotions of commuters around the world, and enhance the happiness of commuters on the subway commute. It is proposed that by effectively integrating advanced technologies in ambient intelligence and gamification into subway carriage, the commuters’sense of participation and immersive experience can be increased to increase the sense of joy on the road. In response to the different travel needs of commuters, a subway carriage plan for the five functions of relaxation, entertainment, humanities, business and learning was created, and specific implementation methods were given. When evaluating the implementation of the program, it is proposed that the use of analysis techniques such as measuring human body temperature, body movements and facial expressions can objectively measure the satisfaction and happiness of commuters. After the transformation of subway carriage, the quality of subway operation services can be effectively improved, and the happiness of commuters can be improved.关键词：ambient intelligence;gamification technology;subway commute;intelligent transportation
- 10
- |
- 3
- |
- 0
发布时间：2022-01-28 - 摘要：In order to find a way for users to reduce their negative waiting emotions through interactive design in the use of mobile applications, and to explore the design strategies to improve and optimize the user waiting experience. Based on the literature research and with the help of waiting psychology theory, take the user interaction waiting phenomenon in mobile application software as the research object, and explores the design ideas and ways to optimize the user interaction waiting experience by analyzing the subjective and objective factors affecting the user waiting psychology. The comparative experimental results show that the targeted interactive waiting design can effectively reduce the user’s perception of waiting time in the waiting process, so as to improve the user’s waiting experience, reduce the negative emotions of waiting and improve the user retention.关键词：mobile interaction;interactive waiting;interaction design;waiting experience
- 21
- |
- 6
- |
- 0
发布时间：2022-01-28 - 摘要：In order to analyze the cause of indoor mold pollution simply and effectively, numerical methods can be used to obtain the temperature and humidity data of indoor walls, and then use WUFI-Bio to predict the growth characteristics of mold spores on indoor walls. Result showed that the initial moisture content of mold spores only affected the water absorption rate before spore germination, the initial moisture content had no effect on the critical moisture content for spore germination, and had little effect on germination time. In autumn and winter, due to the influence of outdoor environment temperature on indoor temperature, the critical moisture content for spore germination is high, and spores are not easy to germinate. Because the temperature in spring and summer is higher than that in autumn and winter, it promotes the germination of spores and the growth of hyphae. The more nutrients the substrate contains, the lower the critical moisture content required for spore germination. In spring and summer, by lowering indoor temperature and humidity, reducing wall pollution can effectively reduce the chance of spore germination.关键词：coupled heat and moisture transfer;biological heat and moisture model;mold contamination
- 9
- |
- 3
- |
- 0
发布时间：2022-01-28 - 摘要：With the continuous growth of mobile application and the popularity of micro-service architecture, it becomes complicated to troubleshoot system faults. Therefore, it is imperative to establish a system that can quickly locate and solve problems. According to the characteristics of the Internet industry and current situation of the enterprise business, design and implement a mobile APM system. In this paper, key technical points such as the optimization of original data analysis algorithm, data storage aggregation strategy and symbol parsing system are explained. The data of half a year after the system was put into operation shows that the failure rate of business application has been reduced by more than 50%.关键词：APM;dail testing;choppy;variable grid;density biased samping
- 23
- |
- 5
- |
- 0
发布时间：2022-01-28 - 摘要：In the context of intelligent manufacturing, in order to improve the intelligence of a mobile phone assembly workshop, improve workshop production efficiency, use self-developed production line simulation software to build virtual workshops and export projects, and use 3D Max and Unity3D to achieve model optimization and construction plant, embed the above Unity3D executable program into the production line simulation software framework to realize view optimization, import the above virtual workshop project, realize model drive through the collection, storage and analysis of physical workshop data, and finally develop based on the original production line simulation software framework. It has realized the digital workshop of mobile phone assembly workshop with good view effect and real-time data drive and historical data backtracking function. The workshop has good use effect and can effectively improve the intelligence degree and efficiency of the workshop.关键词：mobile phone assembly;digital workshop;data-driven;intelligent manufacturing
- 9
- |
- 3
- |
- 0
发布时间：2022-01-28 -
Design and Implementation of Network Element Automatic Inspection System for Large-Scale Cluster AI导读
摘要：How to bring the newly built and upgraded network elements into the network element inspection system quickly, and how to control the number and priority of inspection tasks on the network element or network element group, is an urgent problem in the existing network element inspection system. Therefore, based on the technology of SpringCloud and Redis, a unified collection platform of network element data is designed. On the basis of the collection platform, the scheduling and execution of network element patrol task is realized. The system running test results show that, through the patrol task scheduling configuration and the deployment of micro services in the collection platform, the system can quickly complete the smooth expansion of the system, and solve the problem of current limitation and priority of network element patrol task.关键词：network element patrol;task scheduling;current limiting;logical template;microservice;distributed memory database- 17
- |
- 5
- |
- 0
发布时间：2022-01-28 - 摘要：In order to test the performance of openEuler, an open source operating system, in a high quality and efficiency way in a distributed environment, we have built the experimental platform based on Raspberry Pi (RPi). In aspect of the hardware: three pieces of RPi build the LAN with a switch, composing a cluster with one master node and two slave nodes. The SD card and HDD are respectively used as the main and secondary memory. Also, in terms of software, the OS of the entire cluster is supported by openEuler, above which installed Ceph, Seafile and Frp. Ceph is used to manage the cluster storage resource allocation, while Seafile is used as the cloud storage management. By opening intranet IP, it is supported that members of the open source learning community access the experimental platform via the Internet and do some test work. The experimental platform is simple in composition and low in cost. Through promotion in the openEuler open source learning community and large-scale student experiments, distributed testing of various performances of openEuler can be realized.关键词：open source learning community;openEuler;operating system test;cloud disk;continuous integration
- 21
- |
- 6
- |
- 0
发布时间：2022-01-28 - 摘要：With the development of searchable encryption technology, users can enter multiple query keywords to retrieve data in the cloud server. However, as the amount of data increases, the retrieval efficiency of the cloud server continues to decrease, and its security is difficult to guarantee. To this end, this paper proposes a multi-keyword encrypted sort search method in cloud storage environment. Firstly, by clustering the keywords of the documents, an index vector with more concentrated features is obtained. Second, build tags for the index and query vector, filter irrelevant documents based on the location of the query tags, and reduce search time. Finally, the index vector is grouped according to the category of the corresponding mark, and the high-dimensional encryption key is reduced to multiple low-dimensional keys, which further reduces the encryption time of the index. As the number of document groups increases, query time will be reduced by more than 50%. Experiments show that the scheme can improve query efficiency while ensuring safety and query accuracy.关键词：cloud storage environment;searchable encryption;keyword clustering;index grouping;dimensionality reduction
- 15
- |
- 3
- |
- 0
发布时间：2022-01-28 - 摘要：Bad comments will have an impact on APP downloads. Through the text analysis of the bad evaluation of the captured APP users, explores whether the value-added services launched by developers will have an impact on the consumer’s bad evaluation behavior, and further finds out the factors that cause the consumer’s bad evaluation behavior, providing advice to developers. Using Python crawling comments, using logical regression model to classify, mining bad information in the comments, using LDA model to pre-processed data subject analysis, and the results of visualization and interpretation. Research results show that after a period of time, the behavior of pushing value-added service will cause users to make bad evaluation behavior, and the quality, function and frequency of developers pushing paid version of value-added service will play a certain role in regulating its influence process. Developers need to take a reasonable way to pay attention to push frequency when introducing value-added behavior.关键词：APP;text mining;LDA model;negative evaluation behavior
- 9
- |
- 3
- |
- 0
发布时间：2022-01-28 - 摘要：To solve the problem that the current deepfake image detection methods need to use complex neural networks and a large amount of calculation, propose a deepfake image detection method using low-level features. Firstly, we process images by using error level analysis (ELA) and SRM filter to extract the ELA image and noise image, and convert them into low-level features according to the number of pixels in their grayscale images. Finally a three-layer fully connected network and softmax are used to classify the low-level features. The experimental results are compared with the current popular methods. Compared with the current popular methods, the proposed method in this paper has a simple network structure and is better than other methods in terms of AUC value (99.55%), EER value (2.5%) and computation amount. Extracting low-level features from image preprocessing can effectively improve the classification accuracy and reduce the amount of calculation.关键词：deepfake;fake image detection;error level analysis;SRM filter
- 11
- |
- 4
- |
- 0
发布时间：2022-01-28 - 摘要：Aiming to study the real-time performance and robustness of several common binary feature extraction algorithms in complex environment, four mainstream binary feature-based algorithms, ORB, SIFT, SURF and BRISK, were matched and statistically compared with the image pairs in the Euroc dataset and real scene to analyze their real-time performance. Mikolajczyk's data set is selected to compare the robustness of the algorithm in the scene of image rotation transformation, blur transformation, illumination transformation and perspective transformation. The evaluation index are the algorithm running time and the number of matched feature points. The experimental results show that ORB can achieve enough matched number, and the running time of the algorithm is not more than 24.6ms; SURF can achieve the maximum matching number of 123 pairs in the scene of blur transformation, illumination transformation and perspective transformation; ORB and BRISK can obtain enough matched number in the scene of illumination change. In terms of real-time performance, ORB algorithm has the fastest running speed and the highest real-time performance under the condition of enough matches; SURF has good robustness in the scene of blur transformation, illumination transformation and perspective transformation, and SURF algorithm has the strongest robustness.关键词：feature extraction and matching;real-time performance;robustness;algorithm test
- 10
- |
- 4
- |
- 0
发布时间：2022-01-28 - 摘要：The traditional convolution neural network is insensitive to spatial information and cannot learn the relative position relationship between different features, and the sensory field of each layer of neurons is designed to be the same size, which leads to the inaccuracy of the extracted image feature information. To address these problems, a selective convolutional kernel capsule network is proposed for image classification tasks. The convolution layer of the classical capsule network is integrated into the selective convolution kernel network with two branches, which can extract more abundant and accurate data image feature information and improve the accuracy of image classification. The experimental use CIFAR-10, Fashion-MNIST, SVHN these classical image classification data sets. The results show that the recognition accuracy of the new model is higher than that of the baseline capsule network model, especially the recognition accuracy on the CIFAR-10 data set is improved by 1.73%. The new model effectively improves the accuracy of image classification and has good image recognition ability.关键词：capsule network;dynamic routing;feature extraction;selective kernel networks;dynamic selection mechanism;image classification
- 19
- |
- 9
- |
- 0
发布时间：2022-01-28
- Technical support is provided by Beijing Founder electronics co., LTD 鄂ICP备12016542号-70
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0