Table of contents

2021 2nd International Conference on Computer Science and Communication Technology [ICCSCT2021] has been held successfully on July 29-31, 2021. Due to the COVID-19 pandemic around the world, and with the strict travelling rules in China, it is still difficult to take international travel for our attendees abroad. Therefore, ICCSCT2021 was held both in physical (Beijing, China) and online (Zoom).

The organizing committee would like to appreciate all participants in this special time. The COVID-19 brought huge difficulty in academic communications. However, with the help of internet and communication technology, researchers from different countries and areas could get together and present their latest findings and ideas. This time, we are honored to have 50 participants, including 46 presenters, involved in the forum. The conference was divided into three parts: two keynote speeches in the first day, each with 45min, and 12 oral presentations followed in the first afternoon and the second day, each with 15min. Besides, 34 e-posters have been presented in the conference website. The session chairs and keynote speakers have worked together to selected the best oral presentation and poster presentation.

This year, ICCSCT2021 has attracted about 200 submissions, and 80 of them are accepted to be included in this collection. The process of evaluation and peer-review has last for about 6 months, with around 80 TPC members and reviewers involved in. We would like to express our sincere gratitude to all experts for their professional and detailed review comments, which have contributed to the high standard of this collection.

Collected papers discuss mainly about the state-of-the art topics including Communication Technology, Algorithms, Artificial Intelligence, and Big Data, Robot and Automation, Data Processing, Computing and Analysis. The application field ranges from manufacturing industry to daily life.

In the end, we hope this collection can enhance the understanding of computer science and communication technology and encourage more applications and creations. We appreciate the more convenience life brought by technology. And look forward to a more successful conference next year.

Changbo Cheng

Hubei Zhongke Institute of Geology and Environment Technology

List of titles Organizer, General Chair, Co-chair, Publication Chair, Publication Co-chair, Program Chair, Technical Program Committee are available in this Pdf.

All conference organisers/editors are required to declare details about their peer review. Therefore, please provide the following information:

• Type of peer review:

Double-blind: While receiving a new submission, we/our editors firstly check the scope of it whether suitable for our conference. If not, it will be rejected directly. If yes, it will be sent out to check plagiarism and then peer review. We choose 2 reviewers for each of the submission and after getting the two reviews, we conclude a final result for the paper. Once the two reviews show totally different attitudes, we will find another reviewer for it. And final decisions are made by our editor team based on the reviews.

• Conference submission management system: By our own submission system

• Number of submissions received: 196

• Number of submissions sent for review: 181

• Number of submissions accepted: 80

• Acceptance Rate (Number of Submissions Accepted / Number of Submissions Received X 100):

40.8%

• Average number of reviews per paper: 2-3

• Total number of reviewers involved: Around 80

• Any additional info on review process: Each reviewer reviews 3-5 papers on average. Each paper is assigned to at least 2 peer reviewers.

• Contact person for queries (please include: name, affiliation, institutional email address)

Dr. Yingfa Lu, Hubei University of Technology, 20121063@hbut.edu.cn

The performance bottleneck problem of blockchain has become a core obstacle for its role in more important areas, especially low throughput and long transaction waiting time. Many researches have been conducted to improve the efficiency of blockchain by simplifying the consensus process or redesigning the chain structure at a higher level, but these approaches are subject to various limitations such as the type of blockchain and even pose some security problems. Meanwhile, the blockchain network layer based on the traditional TCP/IP system suffers from problems such as solidified transmission mechanisms and massive redundancy of traffic. Named-Data Networking (NDN), as a new network architecture, can well improve the efficiency of data transmission and information interaction by reshaping the network layer of blockchain. This paper proposes a blockchain data transmission structure using NDN, which will establish the separated data distribution channels among blockchain nodes for block data and transaction data while achieving active data pushing. We design a new table structure for NDN routers to implement path-specific routing and forwarding and exploit the multipath forwarding feature of NDN for fast transmission. Simulation results show that this method has better performance than both existing TCP/IP approaches and typical NDN usage.

In order to improve the communication quality between unmanned aerial vehicle (UAV) and ground users. An UAV communication system model based on full-duplex and non-orthogonal multiple access (NOMA) technology is proposed, and the capacity of the system model in urban scenarios is analyzed. First, the accurate capacity expression of the system model is given; then, the calculation problem of the exponential integral function in the formula is solved by introducing the Q function, and the approximate closed-form expression of the exponential integral function is obtained, and then the approximate closed-form expression of the capacity is obtained; second, the coefficient factor is used to obtain a more accurate approximate closed-form expression; finally, simulation and numerical results show that increasing the number of UAV or NOMA power vector can achieve better capacity performance.

In this paper, we present an Interactive multiple mode-Unscented Kalman filter (IMM-UKF) algorithm to achieve mobile node location under wireless sensor networks environments. In the IMM structure, UKF and Variational Bayesian Adaptive Method based on UKF are adapted in parallel, which can improve positioning accuracy in the process of line-of sight (LOS) and non-line-of-sight (NLOS) signal state switching. The estimated values by filtering are fused according to the weighting factors to get the estimated positions. Moreover, when NLOS measurement noise covariance change, we propose Variational Bayesian Adaptive Method based on UKF to improve robustness. Both Simulation and experiments illustrate that the propose algorithm performs can achieve competitive localization accuracy.

Telemedicine builds a patient-centered remote diagnosis and continuous monitoring service system, which can effectively alleviate the current situation of difficult and expensive medical treatment. This paper focuses on mutual authentication with privacy protection in telemedicine. First, we propose a decentralized multi-authority attribute-based signature scheme. The signature scheme can be used to verify the signer's attributes without leaking the identity information, thereby achieving privacy protection while authenticating. Combined with the proposed signature scheme, a blockchain-based mutual authentication protocol with privacy protection in telemedicine is designed. By this protocol, patients and telemedicine terminal can encrypt and store the medical data off-chain, and store the data digest and decentralized attribute-based signature on the blockchain. As a result, the confidentiality of the data and the authenticity of the source can be guaranteed. In addition, the nature of the blockchain and signature ensure that accurate liability determination can be made when medical disputes occur. Theoretical analysis shows that our protocol is secure and practical.

Due to integration of the frequency agility (FA) technology and coherent technology, Coherent FA radar has natural advantage in ECCM performance and represents one of important development trends of new radar types. Aiming at the technical characteristics of coherent FA radar, this paper proposed a jamming method to coherent FA radar based on intermittent sampling repeater. Theory models of coherent FA radar and intermittent sampling repeater jamming are established at first and the jamming effect is analyzed. The proposed jamming method is validated to be able to achieve the expected jamming effect to coherent FA radar by simulation at last.

The demodulation of digital signal plays a key role in the communication system. The traditional demodulator is usually realized by special hardware platform, which has the disadvantages of high cost and long development cycle. In this paper, we propose an end-to-end digital signal demodulator based on convolutional neural network (CNN). It consists of an encoder and a decoder, in which the encoder encodes the input symbol sequence and maps the signal features to the hidden layer space. Then, the decoder decodes the features of the hidden layer space to obtain the demodulation result of the input sequence. The proposed algorithm can automatically learn how to demodulate the received signal without manually extracting the features. Compared with the traditional demodulator, the proposed CNN demodulator has better bit error rate (BER) performance.

Learning a classifier from imbalanced data continues to be a challenging issue. The oversampling methods can improve the imbalanced classification from the perspective of data preprocessing. Different oversampling methods have been proposed. Nevertheless, most tend to generate unnecessary noise, create redundant synthetic samples in the class center and heavily rely on the parameter k. To solve the above issues, this work presents an oversampling method based on local sets and SMOTE (LS-SMOTE). First, the local sets are searched to describe the local characteristic of imbalanced data. Second, a local sets-based noise filter is designed to remove noise and smooth the class boundary. Finally, on each local set, the interpolation of SMOTE between a base sample and a selected sample closest to the majority class is employed to create the synthetic samples. Experimental results with 12 real data sets have proved that LS-SMOTE outperforms representative oversampling methods in training k nearest neighbor classifier.

The channel coding structure classification based on the traditional method usually requires the coding type and prior knowledge to be known, which is completely invalid for the unknown signals that are usually intercepted in the current communication environment. In order to solve this problem, this paper proposes a method of blind classification of channel coding structure based on one-dimensional Inception. First, we use MATLAB to generate signal datasets with different channel coding structures under different signal-to-noise ratios (SNR), and then construct a Convolutional Neural Network (CNN) model based on the one-dimensional Inception, and use the datasets to train the network model. The final classification accuracy rate is 96%. This experimental result demonstrates the effectiveness of the proposed method.

The deepening of the new round of scientific and technological revolution and the uncertainty of global politics and economy made standardization as an important component of the national innovation system. Therefore, it is necessary to pay attention to the trend of the interaction between the new generation of information technology innovation and standardization, and the trend of how the world is responding. This paper summarizes the characteristics of the measures taken by international and foreign countries to deal with the trend of standardization and information technology innovation through combing the relevant literature, and analyzed these characteristics and put forward suggestions for China.

AFDX network simulation is of great significance for evaluating and determining whether AFDX can meet the real-time requirements of avionics networks. Compared with theoretical analysis methods such as network calculus, network simulation can better understand the true behavior of the AFDX network. In this paper, we build an AFDX network simulator based on NS2, where AFDX switches, end systems and virtual links are properly modeled. Using this simulator, the end-to-end delay of an AFDX network under different scheduling strategies is studied. Some key performance measures of the system are also analyzed.

Chopping and Interleaving (C&I) interference can produce better jamming effect on pulse compression radar. Based on this, a novel radar jamming approach to utilize frequency diverse array (FDA) antenna is proposed in this paper. Compared with the traditional phased array antennas, there is a small frequency increment (δf) between each element of the FDA antenna, which provides promising prospect to develop new radar jamming techniques. The jamming approach proposed in this paper can quickly produce more false targets. Moreover, the number of false targets is manipulated by the elements and frequency offset of FDA antenna, which indicates that many false targets can be easy to control. The simulation results provide strong support for the effect of the proposed approach.

With the progress and development of science and technology, high dimensional data has become a hot topic in scientific research. Lasso is one of the most commonly used methods for solving high dimensional variable selection so far. In the process of performing Lasso solution, the least angle regression algorithm and coordinate descent algorithm are often used. In this paper, Lasso is applied to reduce the dimensionality of the gasoline octane problem, and three algorithms are used to compare, and it is found that the mean square error of the model obtained by using the generalized path search algorithm is the smallest.

Chaotic maps are characterized by randomness, unpredictability and sensitivity to initial values, and are suitable for information encryption. The natural characteristics of DNA molecules have advantages over conventional methods in the process of mass information storage and processing. In this paper, we propose an information encryption scheme based on Logistic map and hybridization chain reaction. Firstly, the plaintext is XOR-operated by chaotic sequences generated by Logistic map, and it is scrambled by hybridization chain reaction, and finally the plaintext is transformed into encrypted text. The analysis of an example shows that the information encryption scheme has good encryption effect.

The development of VR (virtual reality) technology has greatly promoted immersion, but it may cause people to worry about the safety of humans. This study aims to determine the effect of virtual reality based on 9D-VR seat (9 Dimension VR Seat) on the physiological response of users. Recruit 20 volunteers to experience the VR roller coaster and VR big pendulum. Through analysis of heart rate data with visualization and statistical, it is found that there is a significant effect on heart rate when users experience virtual reality, and the excitement points of the VR content were related to the changes in heart rate. Further, we found that the difference between the first experience and the second experience for the same VR content, and the changes in heart rate have eased. These findings provide references for the analysis of users for VR and the design of content of the VR.

A novel laser system of chaos shift parameter synchronization (CSPS) and dynamic chaos shift orbit synchronization (CSOS) is studied in-depth when the parameters of transmitter shift in time, but the parameters of receiver do not change. We find that the system can obtain CSPS between the transmitter and receiver when the transmitter's current is adjusted as chaos shift parameter (CSP) or an alterable parameter (AP). And CSOS still be achieved when the chaotic behavior and its trajectory of the emitter vary or jump in time with a real-time alterable current change in one laser. And our study proves that CSP and AP do not disturb to achieve the synchronizations. Based on CSPS and CSOS, a novel chaos shift orbit modulation (CSOM) coding scheme is studied via a real-time CSP change in the emitter for optics secret communications. The CSOM demodulation of an information is proved in two cases that a CSOM encoding technique is performed successfully. This system has real-time shift orbit performance, so it has high security and increases the difficulty for outside observers to break the information.

With the development of the complexity of the vehicle platform, the bandwidth required for video transmission on the vehicle platform is increasing. The traditional data bus transmission rate can no longer meet the needs in some application fields. The new generation of GMSL technology video format is equivalent to the serial interface video. Format, digital video and audio data are higher than the traditional bus data transmission rate, up to 3.12Gbps. In this context, this paper studies the design of GMSL technology high-speed video transmission circuit, completes the hardware development in ISE, uses FPGA platform to build a video video transmission control system, and completes the high-speed transmission of GMSL format video.

The performance of current radars is degraded in complicated E.M. environment and there is no uniform method to make evaluation to the performance degradation till now. Aiming at certain 3D radar in specific combat scenario, the paper provided the specification system for performance evaluation of radar against complicated E.M. environment, established the evaluation model and conducted normalization processing to specifications through AHP and made quantitative evaluation. The simulation results indicate that the established model by that method is effective and feasible and the evaluation result can provide reference for performance evaluation of radar against complicated E.M. environment.

With the continuous development of mobile communication technology, the market scale of smart mobile terminal industry is expanding rapidly. At the same time, due to the open characteristics of the software ecosystem and usage scenarios of smart mobile terminal, its security is facing great risks. Based on this, this paper first summarizes and classifies the security risks faced by smart mobile terminals according to the risk sources, then studies the current development status of trusted computing technology, and finally analyses the feasibility of introducing trusted computing technology into smart mobile terminals.

To discover the potential risks and to issue a warning is cost-effective to prevent or reduce the loss of geologic hazards. Remote sense technology plays an important role in the development of monitoring methods. In this paper, a monitoring system based on quantitative analysis videos and images is presented. There are three phases. Firstly, layout the monitoring marks as a grid. Videos are captured in the second phase. After that, images are extracted from the video and image processing algorithms about edge detection and center detection chosen by comparing examinations then used to get the current coordinate of the marks. The displacement of each mark is collected by calculation, and then used to display and predicate the trend of deformation.

Real-time monitoring of the aquaculture water state is essential for shrimp aquaculture, and it plays a vital role in the growth of shrimp. The main purpose of this research is to find an effective method to achieve rapid detection and classification with water state images. We obtained a water state image dataset containing four categories in the laboratory, and gained the corresponding Hough datasets through Hough transform, including blue-green and turbid water, reddish brown and turbid water, light green and clear or tawny and clear water, colorless and transparent water. We compare the performance and spatio-temporal complexity of 12 widely used convolutional neural network models in water state image classification tasks, and then optimize the best performing model on the optimizer and hyperparameters. The experimental results show that the InceptionResNet model has the best effect and the InceptionV3 model has the smallest spatio-temporal complexity, while both models realize the task of water state image classification excellently.

Deep learning (CNN) can significantly improve the accuracy of image recognition with its powerful features, but the low-level network layer also contains important feature information. In order to achieve more stable and efficient tracking in multi-target tracking, this type of deep features will also be used to make the features more expressive by integrating the data from the front and back layers. The deformable convolution is also introduced to overcome the deformation problem caused by the camera motion. And with the increase of time, we predict the position of the target by the motion model, so as to remove the position where the target is impossible to reach in physical space, and further optimize the association before multiple targets. In this paper, we use an end-to-end correlation method to reduce the complexity of the algorithm. We tested it on the open source dataset MOT17 dataset and obtained remarkable results.

Load forecasting is an important guarantee for power system design, planning and operation. In order to further improve the accuracy of power load forecasting, a combined forecasting model based on gating recurrent neural network (GRU) and limit gradient lifting (XGBoost) is proposed. Firstly, according to the load data and the input structure of GRU and XGBoost models, the data are preprocessed, and the preprocessed data are input into the corresponding model respectively. Then, the model is weighted by entropy weight method to obtain the predicted value of the final combination model. Finally, the first mock exam is used to compare the combination forecasting model with the common forecasting models. The results show that the proposed method can effectively combine the advantages of the two models, and take into account the continuous time series and discontinuous feature variables, which is more accurate than the single model and the common forecasting models.

Current malignant tumor is one of the main causes of death in Chinese residents, and the incidence of brain tumor in the population is higher. With the continuous development of modern imaging technology, especially MRI imaging technology, it provides a favorable help for doctors to assess brain tumors and choose treatment options. The accuracy of brain tumor segmentation is the key for doctors to diagnose patients' medical condition information. Brain tumor's structure is complex, shape is changeable and grayscale isn't uniform, showing considerable differences in different patients. Artificial segmentation of MRI brain tumor images is time-consuming, and it is often affected by subjective differences. For the above reasons, researchers are working to develop an MRI brain tumor recognition segmentation method. This paper presents a recognition and segmentation method for MRI brain tumors, based on an improved U-Net network model. The proposed method avoids too large or too small learning rate leading to too fast or too slow convergence, and adds the BatchNormalization module to normalize the input of the input activation function of the network, thus realizing the training length, robustness, and accurately and completely dividing the brain tumor image. This paper performs detection experiments on the Kaggle+LGG Datasets, with the average classification accuracy of 99.8309%, the average IoU is 0.9935, and the average Dice coefficient is 0.9956.

The C-band GF-3 satellite is the first fully polarimetric synthetic aperture (PolSAR) designed by China, which has multi-polarization image modes and been used in various applications. Land cover classification is an effective approach in PolSAR image interpretation and further application. However, conventional classification methods are mostly pixel-based and are easily affected by inherent speckle noise. In addition, the feature selection of PolSAR image and the amount of training samples are also critical to the performance of classifiers. To solve these problems, in this paper, we propose a region-based PolSAR image classification method, which uses reinforcement learning method altogether with statistical region merging algorithm to improve the classification performance. The contributions of our method are mainly reflected in three aspects: First, the T3 matrix is considered as the only feature set in our method, including image segmentation and classification. Second, the region is produced via statistical region merging algorithm. Finally, a deep reinforcement learning model is used to obtain PolSAR image classification result. To evaluate the performance of the proposed method, two real GF-3 images are performed in the experiments, and the experimental results illustrate that the proposed method outperforms the conventional methods (support vector machine, random forest, and convolution neural network) in terms of accuracy and achieves the state-of-art results.

This paper proposed a distributed computing environment large-scale face image retrieval method. Based on Spark platform, a face retrieval method based on machine learning is established, which takes advantage of parallel computing to improve the efficiency of face image retrieval. In the method, SIFT algorithm is used for feature coding, PCA dimension reduction processing, HBase database is used for data storage, and KD-Tree query algorithm is used to match images similar to the query images. Meanwhile, large-scale computing engine uses Spark to process data to improve the retrieval efficiency. The CelebA dataset is selected to test the method, and the experimental results show the effectiveness of the method.

In this paper, we propose a geometric generation model for the shape estimation of 3D human body. The model decouples the two processes: (1) Encoding/decoding process maps the samples between the image space X and the latent (feature) space Z. This step achieves the dimension deduction. (2) The process of the Probability measure transformation transforms a fixed distribution ζ G P(Z) to any given distribution µ G P(Z). Before the shape estimation, U-V transform is done with Densepose, in order to transform the 2D human body image to 3D shape pose image.

Empirical formulas for PTC optical efficiency calculation are difficult and costly to obtain from rigorous comparative experiments, whereas simpler optical modeling methods inadequately incorporate realistic optical effects. In this article, algorithms are respectively developed to calculate the geometric concentration ratio (C_g) of linear Cassegrainian solar concentrators (CSC) with a secondary flat mirror based on the way of edge rays from solar sources to a flat-plate receiver. On the basis of the large amount of data generated, machine learning and Python language programming methods are used to analyze and process the data, and the functional relationship between the concentration ratio and each parameter is obtained. The learning and training effect is good, and the ideal result is achieved.

Recognizing health status through human faces is a challenging research topic. Among them, facial recognition expressions can indirectly reflect the inner health status, which has very significant commercial and research value. Most research work on facial expression recognition uses traditional methods, and the accuracy of traditional methods highly depends on feature extraction. Deep learning has already promoted the research on facial expression recognition. This paper proposes a dual-branch network that uses global facial information and local information obtained by using the attention mechanism to merge and identify human facial emotional information. Use of shared pre-training module to extract low-level semantic information of global and also local images. The dual-branch network architecture utilizes the attention module to capture the relationship between different sub-images to fuse the local features of the face. Experimental results demonstrate that the accuracy of the CK+ Dataset reaches 95.96%, which is improved compared to other existing methods.

Written examination is an important way of measuring knowledge and ability, but manual scoring of papers is tedious and error-prone. Automated scoring is more fair, accurate and efficient. With the development of artificial intelligence, automatic scoring of papers through image recognition and object detection is becoming an achievable and promising technology today. The aim of this paper is to design an automatic scoring system for objective questions in the examination papers. The automatic scoring system uses YOLOv3 technique to detect and recognize handwritten numbers and characters on examination papers. It also addresses the problem of incorrect recognition due to scribbles. Compared to optical symbol recognition, it can recognize the handwritten answers without extra answer cards. In addition, there is no limit where the student can fill in the answer. The experiments show that the automatic scoring system performs satisfactorily and has good prospects for practical application in the future.

In the process of hot dip galvanization. An improved multiple linear regression model is established to predict the welding seam position of strip steel when two continuous strip steels pass through the air knife. Firstly, a multiple linear regression model is established according to the production parameters of hot dip galvanizing process. Secondly, the least square method is used to improve the multiple linear regression model. Finally, in order to verify the effectiveness of the method, the model is simulated by SPSS software, and the experimental result show that the method achieves the purpose of quantification.

An analysis and evaluation model combining Pearson correlation and logistic regression with linear regression (LR) was proposed in this paper, to screen the remote measurement stations meeting the threshold requirements by Pearson correlation analysis based on historical rainfall of the related remote measurement stations; and the Pearson correlation coefficient was used to analyze the linear mapping relationship between the target remote measurement stations and the surrounding ones; finally, a data reliability determination model was finally constructed through analyzing the difference between the forecasted rainfall and measured rainfall. With the subjects of rain gauges, and based on the historical rainfall at 88 remote measurement stations (39 hydrologic stations, 49 rainfall stations) in the Dadu River Basin, the rainfall determination model and small rainfall identification model were respectively constructed according to the features of measurement points for rainfall correlation at different remote measurement stations. The experimental results showed that the logistic regression model performed well during the inspection period, with the average accuracy of positive and negative samples of 0.92, the average recall rate of 0.91, and the average F1_Score of 0.91; while in the small rainfall identification model, the linear regression R² was 0.927, and the accuracy rate was 0.60. The model provided a preliminary method for identifying abnormal rainfall, which can reduce artificial misjudgment and greatly improve the identification efficiency of abnormal rainfall.

In order to solve the contradiction between the accuracy of weak signal acquisition and the limited frequency band, a weak signal acquisition system based on adaptive filter is designed in this paper. According to the change of the signal frequency, the proposed adaptive filter based on the technology of hardware adjustable narrowband filtering decides the frequency of the input signal, realizes the narrowband filtering and the large dynamic variation of specific frequency signal. And the high precision acquisition of weak signal can be obtained through the processing of filtering technologies under complex noise conditions, which includes adaptive frequency search, digital low-pass filtering, adaptive adjustment, etc. Finally, the test experiment of the designed filter has been implemented, and the experimental results show that the adaptive filter of the weak signal acquisition system can meet the design requirements and show high performance of data acquisition for weak signal.

A diverse image colorization method based on VAE-GAN and mixture density network is proposed by this paper. The VAE-GAN is used to learn the color representation of images, while the mixture density network is to model multimodal distribution of color with learned representation, so that we can draw samples from the distribution and decode them with VAE-GAN to achieve diverse colorization. In comparison with vanilla VAE, the VAE-GAN is trained with adversarial loss and perceptual loss on the discriminator, which ensures better color representation learned by the model and more consistent color result.

The high grades in postgraduates entrance examination are the stepping stone to improve the professional competition in China, which makes the critical influence factors to the initial examination of postgraduates as a hot topic. In order to measure the correlation between the academic performance of undergraduate students and the probability of successful postgraduate entrance examination, and find some effective ways to improve the success rate in the postgraduate entrance examination, we design regression discontinuity model to search the causal correlation between the preliminary entrance exam and CET-6 scores of candidates. Global parameter estimation and local non-parameter estimation were estimated respectively to improve the accuracy of model fitting. In addition, the validity and robustness of the experiment were tested by sensitivity analysis. The results show that the students who get better grades in college are more likely to have a positive effect in the preliminary entrance examination, and make the success rate of postgraduate entrance examination higher.

Colorization of gray image is a challenging research topic which is often discussed. It can be used to restore photos and assist artistic work. There are many approaches to colorize. From the beginning of the use of artificial colorization, and then the method of automatic colorization appeared. Most automatic colorization methods are based on deep learning. Most previous studies have focused on 2D image, the colorization of 3D objects is rarely discussed. In this paper, we introduce the colorization of 3D objects based on the geometric generation model. Unlike 2D images, 3D objects contain different information. Therefore, there are different problems in colorization. Colorization of 2D image can be regarded as the mapping of grayscale texture and color information. This mapping can be simulated by neural network, such as Variational Autoencoder and Generative Adversarial Networks. For 3D objects, we can use similar method for automatic colorization. In this paper, we introduce a generation model for colorization of 3D objects.

Long sequence time-series forecasting has become a very important task in industrial data analysis. Recently, the Transformer model has been widely used in sequence processing tasks. However, because industrial time series data are generally long and mixed with abnormal data, conventional Transformer model may extract irrelevant information in the context, resulting in poor forecasting. In this paper, we present Transformer with a Sparse Attention Mechanism (SAM) which can ensure local context be better integrated into attention mechanism. Inspired by the gating mechanism of LSTM, the most interesting part of sequence information are retained and the rest of the unimportant information are filtered. More attention can be focused on the factors that contribute most to the forecasting value of the sequence through this method. This method can efficiently capture long-range dependency between output and input. Furthermore, we leverage STL (Seasonal and Trend decomposition using Loess) model and IQR (Interquartile Range) method to address the outlier data. By applying this model to real-world datasets, our method achieves significant performance improvements over other methods.

The era of big data in education has come, data-driven intelligent decision-making has become the development trend of the era of big data. The requirements for the subdivision of education are constantly strengthened, and students prefer to get accurately connected educational courses and practical training. Colleges and universities need to use the existing teaching resources and teacher resources to subdivide the needs of students, and further establish a curriculum and education model in line with the needs of students and personalized characteristics and advantages. Nowadays, precision and personalization have become the key words of education and teaching in the era of big data. This article uses python to search, and uses CiteSpace to define the concept of precision education and analyze and summarize the related literature. Then the article analyzes the application of education big data in precision education, and finally build the overall structure of the precision education big data analysis platform.

Gait recognition based on inertial sensor data develops rapidly. Recently, more and more studies are based on the data collected in-the-wild, where the data quality is limited. Thus, the requirement for data preprocessing is higher. In this work, an adaptive preprocessing algorithm, MAOMP, is proposed to extract the effective components in the gait data. Different from traditional denoising methods, MAOMP recovers valid data from scratch using sinusoidal bases adaptively by projecting signals and bases into the Hilbert space. It can remove invalid data to smooth the signals but highlight the essential extremums at the same time. Finally, MAOMP is evaluated on four publicly available datasets of different grades and three different neural networks. The quantization of SNR shows the data recovered by MAOMP is at a higher level. Compared to the two commonly used preprocessing methods, the performance of MAOMP can be more pronounced as the quality of the datasets decreases. The improvements of the recognition performance are more apparent in the ConvLSTM network compared to the CNN with the data recovered by MAOMP.

The performance of the classifier is weak when the number of the ship radiated noise samples is insufficient. Aiming at above problem, this paper proposes a classification method of ship radiated noise based on simulation signal of variational auto-encoder (VAE). First, build a VAE model, input the real ship radiated noise signals into the model to generate a large number of VAE simulation signals. Then, extract the typical features of simulation signals, and use these features to pretrain a convolutional neural network (CNN) classification model. Finally extract the typical features of the real signals to be predicted, and use the pretrained CNN to complete the classification. Experimental results show that the classification accuracy of the pretrained CNN model is 6% to 12% higher than that of the non-pretrained CNN model.

DNA origami is widely used in DNA computing by its programmability and nano-addressability. This study presents the solution for the satisfiability (SAT) problem utilizing DNA strand displacement and origami DNA: firstly, map all the solutions of the SAT problem to the origami base; secondly, add the initiator chain to make it fully react; finally, determine the feasible solution by whether there is fluorescence on the DNA origami base, so as to search for the final solution. And through a specific case, we verify the feasibility of the model.

Infrared temperature measurement technology has been widely used in industry and agriculture to predict maintenance of electrical equipments due to its advantages of noncontact, harm-free, quick measurement. However, many factors cause errors in infrared thermal measurements, especially the existence of the angle of view causes serious temperature measurement error. Therefore, in this study, based on experimental data, one model related to angle of view and temperature measurement error are established by multiple regression method. Meanwhile, the optimal values of measuring temperature, angle of view is solved by genetic optimization algorithm, which will effectively reduce the measurement error even in the presence of angle of view. Finally, the results of multivariate regression model can effectively compensate for the measurement error, and the correctness of genetic algorithm are verified by experiments. The method proposed in this paper will effectively improve the measurement accuracy, and provide new research method to predict measurement errors for the actual operator of infrared thermal imager.

Bootstrap plays an important role in change point analysis for it is a data driving method and can avoid estimate some redundant parameters. In this paper, we applied three well known bootstrap methods, the sieve AR bootstrap, the fractional differencing sieve bootstrap and the fractional differencing block bootstrap to test the mean change point in the stationary long memory time series. We use the self normalized ratio statistic as the test statistic and approximate its critical values via these three bootstrap methods. We evaluate the empirical size and power performances of three bootstrap methods. Simulations show that the sieve AR bootstrap undergoes serious size distortions when the long memory parameter nears to 0.5, and the fractional differencing block bootstrap always too conservative compared to the other two bootstrap. The fractional differencing sieve bootstrap, in general, has the best finite sample performance. Finally, we illustrated the method via a set annual discharge data in the Nile River and a set of temperature data in the northern hemisphere.

Verification and validation (V&V) are indispensable tools to evaluate the accuracy and reliability in computational simulations. This paper highlights guidelines, related works, as well as recent progresses for evaluating the credibility of model and simulation in computational fluid dynamics (CFD). Herein, detailed definitions of V&V that aims to help researchers to distinguish between the term verification and the term validation, define the conceptual sources of error and uncertainty as well as understanding the difference between them. Then, we discuss related works to V&V for CFD simulations, the most useful methodologies and algorithms have been introduced, problems have been faced as well as solutions has been proposed. Finally, an outlook and perspectives for future challenges in V&V for CFD Simulation have been proposed. For verification assessment, in order to assure the accuracy, the computational solution is compared with an analytical solution or a highly accurate solution. The main strategy of validation is to reflect the accuracy between the computational results and the experimental data, as well as the quantification and estimation of both error and uncertainty. It is hoped from this review to help in research, development as well as the use of CFD simulations by establishing robust methodologies and terminologies for V&V.

Real-time road navigation information update in vehicle networking plays an extremely important role in safe driving and alleviating traffic congestion, but how to protect user privacy from leakage is a major challenge when updating real-time navigation information. In order to solve this problem effectively, a certificateless aggregate signature scheme with real-time navigation information update function is proposed, which is suitable for vehicle networking. In the scheme, when the navigation company needs to access the data, the fog node aggregates the signature message of the vehicle broadcast and uploads it to the trusted center, and then feeds it back to the navigation company after batch verification by the trusted center. The trusted center generates temporary pseudonyms for vehicle users, realizes the anonymity of user identity, and meets the requirements of conditional privacy protection. The aggregate signature technology is used to reduce the computing and communication overhead. Finally, based on the discrete logarithm difficulty problem in the elliptic curve, it is proved that the scheme satisfies the existence and can not be forged under the adaptive selection message attack. The results of numerical analysis show that the scheme has some advantages in terms of computational cost.

Cyber performance is an evolving issue for enterprises and individuals. However, Internet applications with many concurrent accessing are obsessed with the tradeoff between the scale of users and the efficiency for each user. Compared with multi-tenancy catch technology, it is almost impossible to meet the requirements of enormous accessing when improving the hardware performance. Different from the classical multi-tenancy cache system focused on the size of cache partition, we propose a multi-tenancy cache model with adaptive admission strategy. The model evaluates the hit efficiency of cache objects in cache space, and expels cache objects with low hit rate through an effective algorithm. The experimental results show that the proposed method improve the response efficiency, especially in the high concurrent accessing scene for Internet applications.

Based on the current technology of network electromagnetic space situation awareness service by the limitations of traditional system framework and fixed working mode cannot adapt to the harsh environment within strong antagonistic conditions, limited tactical edge computing resources and intermittent network communication, the author focuses on the maintenance of traction network electromagnetic space security demand, based on the network electromagnetic situation global multi-dimensional cognition, combines with artificial intelligence machine learning and big data cloud computing of cutting edge technology for fusion development, aims at intelligent service of network electromagnetic space situation, puts forward to the network electromagnetic space situation intelligent service within edge and cloud computing synergy, builds cloud-edge-terminal regional network electromagnetic space situation cloud information service technology architecture to provide technical support for realizing the goal of network electromagnetic space security maintenance including global awareness coverage, prediction analysis, effective response and deterrence of the regional network electromagnetic space environment.

Aiming at the problems of the existing mining algorithm processing data scale, the time complexity of frequent item set storage, and the limitation of the generated association rules, the CM_HI_pApriori (parallel Apriori algorithm of the compressed matrix based on HashMap and Interest) algorithm is proposed. First of all, the parallel data partition method is used to solve the problem of low serialization efficiency. Secondly, HashMap is used to store and query frequent item sets, which makes the time complexity of storing frequent item sets change from linear growth to constant growth. Finally, the interest model is used to solve the problem that association rules do not conform to the actual situation. Experiments show that the CM_HI_pApriori algorithm not only has high scalability and stability in processing large-scale data but also has high accuracy in generating association rules.

In order to enhance detail information of the image effectively, this paper proposed a novel improvement method for remote sensing image based on NSST. First, adaptive histogram equalization is adopted to enhance the overall contrast; Second, the remote sensing image is broken down into low-frequency coefficients and high-frequency coefficients by NSST; For the purpose of preventing the over-improvement, the low-frequency sub-band will not be processed, adaptive threshold denoising and linear improvement is used for the high-frequency coefficients. At last, a Laplacian filter is used for improving the details and edges of the reconstructed image. Experimental effects show that the technique has achieved obvious results in remote sensing image improvement.

Path planning of mobile robots is an important research content in mobile robot design. In view of the characteristics of classic ant colony algorithm, such as long distance, low efficiency and easy to fail into dead zone, a new pheromone updating algorithm and global/local pheromone distribution models are proposed. The grid method is used to construct two-dimensional plane space with obstacles and the simulation experiment is carried out. The simulation results show that under the improved algorithm, the robot path planning efficiency is high and the path distance is short, which verifies the validity of the model.

As the traffic sign target detection has the defects of small size, low resolution and unobvious features, an improved YOLOv3 network model is proposed. Firstly, using a K-means algorithm TT100K traffic sign data set clustering analysis and redefine the initial candidate frame of the network. Secondly, this paper improves the FPN structure in the original network, retain the large-scale prediction of 52×52 in the original network, and then use the feature map output from the second down-sampling in the YOLOv3 network to establish a larger-scale prediction of 108×108. For the purpose of solving the size of image and distortion problems, the pyramid pooling operation with fixed block sizes of 5, 9, 13 is used before the detection layer, and then the output features are merged with the original feature map, so as to achieve the same size for inputs of different sizes output. We use the improved YOLOv3 network model with its original model and other small target detection algorithms to conduct comparative experiments on the TT100K data set. The results show that the improved YOLOv3 network model can detect traffic signs more effectively, with better detection accuracy and real-time performance.

The F10.7 flux of the sun is an important parameter that characterizes the level of solar activity. However, due to the long-term periodicity and short-term randomness of solar activity, it is difficult to obtain accurate prediction results for F10.7 using statistical methods. The Prophet algorithm is based on time series decomposition and machine learning fitting. It can deal with the situation where there are some outliers in the time series, and it can also deal with the problem of partial missing values. F10.7 is a typical time series data, composed of two parts: time and observations, and has a history of nearly one hundred years of observation. It is inevitable that there will be some outliers and missing values in the observation process. Prophet's data processing characteristics make it suitable for the requirements of the solar F10.7 observation data. Through reasonable selection of change points, it can realize the forecast of the future and the forecast of the seasonal trend, and finally realize the model fitting. The experimental results show that using the Prophet model to predict the consistency of F10.7 data and real data can reach more than 90%.

In order to quickly identify the fire hazards and formulate the corresponding measures inside the transformer substation. A fire detection algorithm of transformer substation based on block marking is studied. On the basis of the transformer substation monitoring platform, the image of primary equipment in the transformer substation is collected through the video monitoring software. According to the flame characteristics, the affected parts are identified and located by block marking. The fire detection algorithm is realized by using OpenCV. Through the fire identification software, the operation of the equipment can be monitored inside the substation in real time, and the fire hazards can be detected in time. The experimental results show that: through the image preprocessing and block marking, comparing the characteristics of the flame, the monitoring and alarm analysis of the primary equipment area in the transformer substation can be realized, and the rapid response to the fire can be realized accurately.

Traditional fundamental frequency extraction algorithms mainly include cepstrum method, short-time autocorrelation method. Some end-point extraction methods can't correctly select the part of effective value due to the particularity of dialect audio and the noise influence. And, some fundamental frequency extraction methods cannot get the data consistent with the trend of tuning values. Against the above, this paper proposes an improved double threshold method for the extraction of the basic frequency according to the characteristics of the vowels of the initials of the one-word tone, which extracts the continuous vowels of the one-word tone, and looks for the basic frequency extraction algorithm that meets the classification of the tones of one-word. Finally, the effective basic frequency data is obtained. This method is more effective than other methods in speech tone recognition. It is of great significance to the study of speech tone, especially the study of dialect tone.

Aiming at the problem of huge amount of information data for solving the shortest path of urban traffic network, based on the analysis of the characteristics and defects of genetic algorithm, a parallel genetic algorithm based on MapReduce is proposed to solve the shortest path of urban traffic network. By designing the adaptive function, the algorithm can more effectively solve the shortest path problem and effectively avoid local optimization, improve the running speed and convergence of the algorithm. The experimental results show that the parallel genetic algorithm based on MapReduce is an effective shortest path solution method, which is efficient, effective and feasible.

This paper compares the mode based lithium-ion battery state of charge (SOC) estimations using offline and online parameters under varying temperature. An innovative offline identification method based on genetic algorithm (GA) is used for off-line identification of battery model parameters. The common extended Kalman filter (EKF) and the joint extended Kalman filter (JEKF) are implemented as the algorithms to implement SOC estimation with offline and online parameters. The SOC estimations by JEKF using online parameters and by EKF using offline parameters from mismatched temperature are compared. The results are as follows. When battery temperature is inaccurate, the inaccurate temperature can result in inaccurate offline parameters parameters, which will further increase the SOC estimation errors by EKF using offline parameters. In contrast, SOC estimation accuracy by JEKF are still accurate when no temperature information is provided, because the parameters are online updated by JEKF.

With the continuous development of computer technology, simulation technology has been more widely used. Through the ANSYS-Workbench software to simulate the solder ball shear experiment, the strain stress and the overall deformation cloud map were obtained, and the distribution law of the solder ball and IMC stress and strain in the shear test was found. There is no significant difference compared with the actual experimental results. The reliability is high, which provides a certain reference for the development and experiment of SAC solder joints, and accelerates the development process.

In aircraft manufacturing and other assembly scenarios that require high product assembly accuracy, the true topography of the mating surface of the part is one of the important factors that affect the aircraft's aerodynamic shape, step difference, and other key characteristics. However, the existing Computer-Aided Tolerancing (CAT) system is difficult to express the true topography of the part surface and the mating simulation. To solve this problem, an analysis method that reflects the true contact state of the part is proposed—established a realistic assembly model containing the geometric information and the assembly force information of the part, designed the matching surface area segmentation and contact point search algorithm, and the contact state determination algorithm considering the assembly force. This paper takes the analysis of the contact state of step surface parts subjected to uniform assembly force as an example and compares the analysis results of commercial software packages and the difference surface method to verify the feasibility of this method.

With the vigorous development of the oil industry, the number of tank trucks has increased significantly in recent years. Accidents such as explosion and leakage of oil tank truck during transportation are common. Therefore, it is particularly important to study the stress analysis, strength optimization and safety reliability of oil tank truck. This paper establishes the three-dimensional parametric model of the tank car by the solid works software and changes the thickness and height of the wave preventer, and analyze the influence of the thickness and height of the wave preventer on the tank car under the starting and braking conditions by the ANSYS software. The deformation diagram, strain nephogram and stress nephogram under different working conditions were compared and analyzed to optimize the dimension parameters of the wave preventer. The results show that the optimal thickness and height of the wave preventer are10-15mm and 990-110mm. It can not only meet the oil transportation safety of tank truck, but also save the manufacturing cost, which provides important reference value for the design of tank truck.

Because of the change in the diameter of the elliptic current, the magnetic field distribution of its arbitrary spatial position cannot be transformed into an elliptic integral problem. It is very difficult or even impossible to solve this problem by analytical method. Intensity distributions of the magnetic field excited by an elliptic electric current are studied. The magnetic field intensity of the central axis is discussed using analytical and numerical simulation methods. The results obtained by the two methods are in good agreement. Application of numerical simulation method in rectangular coordinate system, spatial distributions of the magnetic field of an elliptic electric current is displayed intuitively, the effect of flattening of ellipse on the spatial magnetic field distribution is analyzed, and the characteristics of spatial magnetic field distribution of elliptic electric current are revealed further.

With the development of aging society, a great number of stroke patients are rapidly increasing. To meet the rehabilitation training needs of stroke patients, this paper proposed a simple, reliable and universally applicable lower limb rehabilitation training robot based on human lower limb movement characteristics. The robot is mainly composed of trolley, electric lifting column, screw adjustment component and left and right training mechanism. The kinematic model was developed by the D-H parameters method. Based on the rehabilitation mechanism, two modes of active training and passive training were designed. The robot can replace the work of therapists, significantly reduce the workload of therapists, and effectively alleviate the social situation of therapist shortage.

Currently, according to statistical data showed by world health organization, the aging problem has been becoming an extraordinary serious social challenge. China has become one of the countries with the largest elderly population and the fastest growth in the word. Compared with the traditional one-to one between therapist and patient, this paper proposed an intelligent bridge-style movement rehabilitation robot. The rehabilitation robot assisted patients to complete the bridge-style movement via auxiliary mechanism. Then, applying muscle parameters and skeletal muscle carried out the simulation analysis of the auxiliary mechanism. The results proved that the auxiliary mechanism performs well in the process of auxiliary bridge-style movement.