Aiming at the low detection precision of strip steel surface defects in actual scenarios， which is prone to missed detection and false detection， a YOLOv5-CFD model consisted of CSPDarknet53， Frequency channel attention Network （FcaNet） and Decoupled head was constructed to detect strip steel defects more accurately. Firstly， Fuzzy C-Means （FCM） algorithm was used to cluster anchor boxes in NEU-DET hot-rolling strip steel surface defect detection dataset published by Northeastern University to optimize the matching degree between the prior box and the ground-truth box. Secondly， in order to extract the rich detailed information of the target area， the frequency domain channel attention module FcaNet （Frequency channel attention Network） was added to the original YOLOv5 algorithm. Finally， the decoupled head was used to separate the classification and regression tasks. Experimental results on NEU-DET dataset show that with introducing a small number of parameters to the original YOLOv5 algorithm， the improved YOLOv5 algorithm has the detection precision increased by 4.2 percentage points， the detection mean Average Precision （mAP） of 85.5%； and the detection speed reaches 27.71 Frames Per Second （FPS）， which is not much different from the original YOLOv5 so that YOLOv5-CFD can meet the real-time detection requirements.

Under increasingly complex network environment, traditional intrusion detection methods have high false alarm rate, low detection efficiency and the contradiction between accuracy and interpretability in the optimization process. Therefore, an Improved Rough Set Attribute Reduction and optimized K-means Clustering Approach for Network Intrusion Detection (IRSAR-KCANID) was proposed. Firstly, the dataset was preprocessed based on the attribute reduction of fuzzy rough set in order to optimize the anomalous intrusion detection features. Then, the threshold of intrusion range was estimated by improved K-means clustering algorithm, and the network features were classified. After that, according to the linear canonical correlation used for feature optimization, the feature association impact scale was explored from the selected optimal features in order to form the table of feature association impact scale, and the detection of anomalous network intrusion was completed. The experimental results show that the minimum measured feature association impact scale table after feature optimization clustering can minimize the complexity of intrusion detection process and shorten the completion time on the premise of guaranteeing maximum prediction accuracy.

Although the existing autoregressive translation models based on recurrent neural network, convolutional neural network or Transformer have good translation performance, they have the problem of low translation speed due to low decoding parallelism. Therefore, a non-autoregressive model based learning rate optimization strategy was proposed. On the basis of the non-autoregressive sequence model based on iterative optimization, the learning rate adjustment method was changed, which means that warm up was replaced with liner annealing. Firstly, liner annealing was evaluated to be better than warm up; then liner annealing was applied to the non-autoregressive sequence model in order to obtain the optimal balance between translation quality and decoding speed; finally a comparison between this method and the method of autoregressive model was carried out. Experimental results show that compared with the autoregressive model Transformer, when the decoding speed is increased by 2.74 times, this method has the BiLingual Evaluation Understudy (BLEU) score value of translation quality of 41.31, which reached 95.34% of that of the Transformer. It can be seen that the non-autoregressive sequence model of liner annealing can effectively improve the decoding speed under the condition of reducing a little translation quality, which is suitable for the platforms with urgent need for translation speed.

Aiming at the shortages in trust analysis of existing social recommendation algorithms, a social recommendation algorithm based on multi-dimensional trust and collective matrix factorization was proposed with full use of user trust relationship mined from social auxiliary information. Firstly, the dynamic and static local trust relationships were extracted respectively from social interaction behaviors and social circle features of the user, and the global trust relationship was extracted from the structural features of trust network. Then, a social recommendation algorithm was presented by collective factorizing the enhanced following relationship matrix and the social trust relationship matrix, and a stochastic gradient descent method was utilized to solve the algorithm. The experimental results on the Sina microblog dataset indicate that the proposed algorithm outperforms some popular social recommendation algorithms such as socialMF, LOCABAL, contextMF and TBSVD (Trust Based Singular Value Decomposition), in terms of recommendation accuracy and Top- K performance.

For the complex and autonomous landing scene is difficult to be recognized in wild flight environment for low-small-slow Unmanned Aerial Vehicles (UAV), a novel field scene recognition algorithm based on the combination of local pyramid feature and Convolutional Neural Network (CNN) learning feature was proposed. Firstly, the scene was divided into small scenes of 4×4 and 8×8 blocks. The Histogram of Oriented Gradient (HOG) algorithm was used to extract the scene features of all the blocks. All the features were connected end to end to get the feature vector with the characteristics of spatial pyramid. Secondly, a depth CNN aiming at the classification of scenes was designed. The method of tuning training was adopted to obtain CNN model and extract the characteristics of deep network learning. Finally, the two features were connected to get the final scene feature and the Support Vector Machine (SVM) classifier was used for classification. Compared with other traditional manual feature methods, the proposed algorithm can improve the recognition accuracy by more than 4 percentage points in data sets such as Sports-8, Scene-15, Indoor-67 and a self-built one. The experimental results show that the proposed algorithm can effectively improve the recognition accuracy of the landing scene.

The distribution of color image pixel value is difficult to describe in hierarchical Markov Random Field (MRF) segmentation algorithm, therefore, a hierarchical MRF segmentation algorithm based on RGB color statistic distribution was proposed to solve this problem. The key parameters of the MRF model were set up, and the related formulas were deduced. With the RGB color statistic distribution model, the hierarchical MRF energy function was rewritten, and the k-means algorithm was used as presegmentation method to realize unsupervised segmentation. The proposed algorithm has fewer color distribution parameters and lower computational cost in comparison with traditional MRF segmentation model, which describes color distribution more accurately; and it can describe different targets and background very well without being restricted by target and background color distribution and target spatial distribution. Experimental results prove the effectiveness of the proposed algorithm, which is superior to the MRF algorithm and Fuzzy C-Means (FCM) algorithm in computing speed and segmentation accuracy.

To solve the storage and network transmission problem of the three-Dimensional (3D) mesh model, a new 3D model compression algorithm was proposed. Based on the slicing for 3D mesh, the proposed algorithm was composed of the following three steps: slice vertex calculation, slice boundary sampling and encoding for the image obtained by slicing. For a given 3D mesh model, the bounding box of the model was firstly calculated; then the model was sliced along the longest direction of the bounding box. In the procedure of slicing, the intersection point of the slice with the edge of the mesh was calculated, and as a result, all the intersection points in the same slice constituted a polygon. Then the boundary of the polygon was uniformly resampled so that each layer of the slice had the same number of vertices. After resampling of the polygon boundary, the coordinates of vertices in each slice were converted into the polar form. In this way, all ρ-coordinates and θ-coordinates of the vertices in each slice could constitute one image respectively, and the original 3D model could be represented by these two images. The new representation method has two obvious advantages: first, the dimension of the data is reduced, thus the amount of the data is effectively reduced; second, the data in these two images have great data correlation, and as a result, the entropy of the data is further reduced. Based on these two advantages, the proposed algorithm compressed these two images by difference coding technique and arithmetic coding technique, and then the compressed files were obtained. Compared with Incremental Parametric Refinement (IPR) method, the coding efficiency of the proposed algorithm was increased by 23% under the same quality of the decoded model. The experimental results show that the proposed algorithm can obtain good compression efficiency, and effectively reduce the data amount in the application of 3D model storage and transmission.

To solve the problems of Web service composition and verification, a formal model was proposed based on the framework of category theory. Process Algebra was introduced into the framework to describe the external behavior of service component, establishing a formal semantic model for the architecture of Web service system. The service network was described with category diagrams, in which Web services were used as categorical objects, and the interactive and composition relationships between services were used as morphisms. On the basis of the formal definitions of service interface, Web service and service composition, a further analysis and discussion about the semantics of service composition and interaction was undertaken. The concepts on Web service substitutability and service request satisfiability were formally defined. The application research shows that the proposed framework enhances semantic description capabilities of Web service architecture.

Spherical Voronoi diagram generating algorithm based on distance computation and comparison of Quaternary Triangular Mesh (QTM) has a higher precision relative to dilation algorithm. However, massive distance computation and comparison lead to low efficiency. To improve efficiency, Graphic Processing Unit (GPU) parallel computation was used to implement the algorithm. Then, the algorithm was optimized with respect to the access to GPU shared memory, constant memory and register. At last, an experimental system was developed by using C++ and Compute Unified Device Architecture (CUDA) to compare the efficiency before and after the optimization. The experimental results show that efficiency can be improved to a great extent by using different GPU memories reasonably. In addition, a higher speed-up ratio can be acquired when the data scale is larger.

Concerning the problems of designing and modeling the process of Point-Of-Care Testing (POCT) system, a method for concurrence system modeling and analyzing based on Time-Petri Net (TPN) was proposed which built more accurate information model for the process designing of POCT system. The activity holding duration was introduced into classical TPN, and the TPN modeling method for POCT control process was proposed. The scheduling simulator embedded in Petri net model was also designed for assisting the analysis, and optimization of the POCT control process. The simulation results show that the proposed modeling method for TPN can satisfy the practical requirement of process modeling of the parallel multi-class POCT control system in the fields such as reachable nodes and running time and provide powerful tool for process simulation and analysis. Furthermore, the proposed TPN can assist the system designer for the optimization of POCT system.

The objective of Blind Source Separation (BSS) is to restore the unobservable source signals from their mixtures without knowing the prior knowledge of the mixing process. It is considered that the potential source signals are spatially uncorrelated but temporally correlated, i.e. they have non-vanishing temporal structure. A second-order statistics based BSS method was proposed for such sources. The robust prewhitening was firstly performed on the observed mixing signals, where the dimension of the sources was estimated based on the Minimum Description Length (MDL) criterion. Then, the blind separation was realized by implementing the Singular Value Decomposition (SVD) on the time-delayed covariance matrix of the whitened signals. The simulation on separation of a group of speech signals proves the effectiveness of the algorithm, and the performance of the algorithm was measured by Signal-to-Interference Ratio (SIR) and Performance Index (PI).

According to the influence of earliness and reworking penalties, the production order acceptance problem of hot-rolled bar was studied. A mathematical model with the objective of maximize gross profit of order was proposed. A hybrid algorithm with improved NEH (Nawaz-Enscore-Ham) algorithm and Modified Harmony Search (MHS) algorithm was proposed for the model. With the consideration of the constraints in the model, an initial solution was generated by the improved NEH algorithm and further optimized by MHS algorithm. Furthermore, the idea of Teaching-Learning-Based Optimization (TLBO) was introduced to the process of selection and updating for harmony vector to take control of the acceptance of new solutions. Meanwhile, in order to balance the breadth and depth of this algorithm's searching ability, the parameters were adjusted dynamically to improve the global optimization ability. The simulation experiments with practical production data show that the proposed algorithm can effectively improve total profit and acceptance rate, and validate the feasibility and effectiveness of the model and algorithm.

To deal with the problems of poor exploration capability and slow convergence speed in Biogeography-Based Optimization (BBO) algorithm, a hybrid quasi-oppositional learning based BBO algorithm named HQBBO was proposed. Firstly, the definition of heuristic hybrid quasi-oppositional point was given and its advantage in searching efficiency was proven theoretically. Then, the hybrid quasi-oppositional learning operator was brought forward to enhance the exploration capability and accelerate convergence speed. Meanwhile, the dynamic scaling strategy of searching domain and the elitism preservation strategy were utilized to boost optimization efficiency further. Simulation results on eight benchmark functions illustrate that the proposed algorithm outperforms the basic BBO algorithm and the oppositional BBO (OBBO) algorithm in terms of convergence accuracy and speed, which verifies the effectivity of hybird quasi-oppositional learning operator for improving the convergence speed and global exploring ability.

To resolve the problems of system composition and semantic verification in the construction process of distributed computing model, a semantic description framework for Agent-based distributed computing system was proposed based on category theory and process algebra. The structural semantics of the system model was described within category diagrams, and the relations between components were formally described by morphisms. On this basis, the semantic properties that should be preserved during the process of system modeling, refinement and migration were further analyzed and discussed. The application research shows that the proposed framework can not only be used for distributed system modeling, but also be used for the correctness analysis of system decomposition and composition.

Concerning the identification of the accurate model parameters of biological fermentation process, a parameters estimation method for non-structural dynamical model of penicillin fermentation using the Gravitational Search Algorithm (GSA) was proposed. Based on the rule of fermentation mechanism, the appropriate state equations of non-structural dynamical model were chosen; and through virtue of the global searching ability of GSA, the parameters of state equation were estimated and the accurate fermentation model was obtained. The simulation results show that GSA accurately estimated model parameters in penicillin fermentation process, the accuracy of the obtained model can meet the requirements of state estimation and condition control in penicillin fermentation process. Therefore, GSA can be applied to model parameters estimation effectively.

Hadoop has been widely used in large data parallel processing. The existing tasks assignment strategies are almost oriented to a homogenous environment, but ignore the global cluster state, or not take into account the efficiency of the implementation and the complexity of the algorithm in a heterogeneous environment. To solve these problems, a new tasks assignment algorithm named λ-Flow which was oriented to a heterogeneous environment was proposed. In λ-Flow, the tasks assignment was divided into several rounds. In each round, λ-Flow collected the cluster states and the execution result of the last round dynamically, and assigned tasks in accordance with these states and the result. The comparative experimental result shows that the λ-Flow algorithm performs better in a dynamic changing cluster than the existing algorithms, and reduces the execution time of a job effectively.

This paper gave a solution to emergency task in space. When the emergency task occurred, at first, the existing satellite constellation was used to cover the target locations and the coverage rate was computed. If the coverage performance did not meet the mission requirements, the Non-Dominated Sorting Genetic Algorithm Ⅱ （NSGA-Ⅱ） was used to optimize the layout of satellite constellation, i.e. optimizing the anomaly of each satellite in the constellation. Then the phase modulation maneuver was used to achieve constellation optimization results, i.e. maneuvering the satellites to the specified locations, and calculating maneuver time and energy of each satellite. Finally, an emergency task example and its solution process were provided, and the time and energy of the satellite orbit maneuver were calculated.

Network Time Protocol (NTP) is a simple, economic and efficient way to accomplish time and frequency synchronization of multiple nodes, while relevant study on the performance evaluation is hard to find in literature, which makes it a question whether to use NTP in application. Concerning this problem, the local network NTP performance and impact of system / network load were measured and analyzed on Windows platform. By comparing time value obtained from IRIG-B time code reader and GetLocalTime Windows API, frequency skew of computer clock signal was approximated. The skew value was close to the value calculated by NTP.

Many different communication radius of the communication nodes that may cause an unstable network performance can be easily found in Epidemic Routing (ER) network. A network model that combines network coding and epidemic routing can solve this problem. Compared with the traditional epidemic routing, the Network Coding Based Epidemic Routing (NCER) can transmit packets with network coding. In order to compare the performances of the ER and NCER, a probability model of the transmission delay of the network was built. The comparative results between the two protocols with the probability model above show that NCER can be more efficient and stable than ER. The correctness of this probability model has been proved in the simulation. Finally, according to the model evaluation results, a scheme has been given to reduce the network transmission delay.

Network anomaly detection is an important research topic in the field of intrusion detection. However, it is inefficient in practice because the detection rate and false alarm rate restrain each other. Based on the anisotropic centroidal Voronoi diagram, a new algorithm of network anomaly detection was proposed. In this new algorithm, the anisotropic centroidal Voronoi diagram was used in the clustering of data set at first, then the point density for each data point was computed out, which was used to determine whether the data point was normal or not. The laboratory tests on KDD Cup 1999 data sets show that the new algorithm has a higher detection rate and a lower false alarm rate.

To eliminate or reduce the image degradation, the image restoration techniques are often used. In this paper, a new image restoration way was obtained for blind image. Firstly, the Point Spread Function (PSF) of blurred image was estimated by cepstrum. Secondly, the blurred image was restored by the improved Kalman filter. The cepstrum was the way that the PSF of blurred image could be obtained through analyzing the relationship between two parts. One reflected original image, and the other reflected blurred system. The improved Kalman filter took account of the model error of system during the process of estimation. Some digital simulation experiments were done through Matlab. The results using the improved Kalman filter algorithm based on estimated PSF indicate that the proposed method effectively eliminates the impact caused by inaccurate PSF and it has better effects than Kalman filter.

A variable size block matching bi-directional motion estimation and compensation scheme was introduced.The optimal segmentation of the encoding frame was achieved by the entropy criterion.Assuming the backward frame(reference frame) was encoded by the variable size block matching backward motion estimation and compensation,then the current frame was encoded by the bi-directional motion estimation and compensation.For this frame,backward motion estimation and compensation used the segmentation information of the backward frame firstly.Then,from the backward frame to the current frame,forward motion estimation and compensation was performed on the smaller blocks.The overlapped compensation pixels were the average filter.Because it used variable sizes block matching and utilized the well performance of forward motion estimation,the prediction quality was further improved.Simulation results show that the average peak signal-noise ratio is increased with various video sequences.

According to the data characteristics in sensor networks and the good performances of wavelet transforming in data stream compression, a novel mixed -entropy data compression algorithm based on interval wavelet transforming was proposed for sensor network. Theoretical analysis and simulation results show that, the new method can compress the data stream for sensor networks effectively, and reduce the energy costs of nodes in data transferring. So, it can prolong the lifetime of the whole networks to a greater degree combined with those traditional DC (Data Centric) routing algorithms such as DD (Directed Diffusion) protocol.