International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems

Pop-ups are a recently popular way of human-computer interaction, allowing viewers to actively participate in the discussion of a film video, but pop-ups can be deleted due to exceeding the size limit of the pop-up pool, and pop-up data is difficult to obtain in its entirety. To this end, this paper proposes a local convolutional neural network (CNN) based pop-up text sub-recognition algorithm. Finally, the fuzzed data is reconstructed into a data look-up table and the contents of the data look-up table are formatted for output. The results of this text classification and recognition algorithm are compared with those of four commonly used domestic and international search engines by subjective evaluation method, and the proposed text classification and recognition algorithm is found to be advanced.

Real-time simulation of ring-grid laser images using the conventional method has several drawbacks, including weak edge fitting, poor clarity, low image recognition rates, uneven brightness distribution, and significant ring-grid laser image interference from the surrounding environment. In this research, we develop the diagonalizable and readily invertible Hessian matrix, offer an enhanced Newton projection iteration algorithm, and use the degenerate gradient magnitude as the constraint set. We also develop a real-time re-illumination model appropriate for modeling ring-grid laser maps and suggest a method for simulating ring-grid laser maps based on it. The simulations demonstrate that the suggested method successfully matches ring-grid laser pictures in real time with little relative error and deviation, a high signal-to-noise ratio, and pleasing aesthetic effects.

To effectively promote the efficient dissemination of sci-tech journals and improve the influence of sci-tech journals, a precise push method of sci-tech journals based on knowledge graph reasoning is proposed, using knowledge graph to build network to realize push reasoning of scientific and Technological Journals. Based on the one-way author relationship network diagram and the two-way keyword network diagram, the PageRank algorithm is used to calculate the weight of network vertices, quantitatively and accurately identify the research direction of push customers and predict customers, and realize the accurate push management of content, so as to obtain the basis for the construction of network diagram. Based on the previous article in the Journal of the information society of science and technology of China, this paper constructs a knowledge reasoning diagram, mines 52 co-author cases and 60 related papers as experimental data, and verifies the progressiveness method through customer value. According to the network diagram, the progressiveness of this method is verified. The experimental results show that the knowledge map network reasoning method proposed in this paper has high accuracy in identifying the push objects of scientific and technological journals and accurately predicting the research direction. The matching degree between the number of customers and the pushed content. Using knowledge map network can ensure the timeliness and accuracy of pushing scientific and technological journals.

The spatial land change of villages and towns can be analyzed through external spatio-temporal characteristics, which can realize the evaluation of historical development and assist the implementation of planning and design norms and the implementation of problem rectification. To explore the specific process and the overall pattern of land use change, it is necessary to analyze and predict all kinds of data in a unified way to grasp the law of development and the direction of future development. This paper takes the spatial development and planning of villages and towns as the research object, selects the remote sensing image data of regional historical development as the driving factor, combines with Gaussian function fitting to calculate the training samples, and constructs the regional spatial environment analysis system. According to the processing model of external characteristics of time series, the spatial distribution and evolution of Risk-Screening Environmental Indicators (RSEI) and its components are analyzed. The multi-dimensional remote sensing data is innovatively fused into the time series processing model. Besides, the curve which can characterize the spatial change law and is not completely symmetrical is obtained through the least square method fitting so that the data with unequal time series intervals can be flexibly processed. Finally, through the experimental analysis, the traditional Asymmetric Gaussian model (A-G) and Savitzky-Golay (S-G) are taken to compare and analyze the effect of various factors on the processing of time series characteristics. Tests show that the proposed algorithm is superior to the comparison method in noise removal of remote sensing data. The data fitting coefficient is 0.02, and the root mean square measurement error is more than 7% higher than the comparison method. It can predict the trend of land use change distribution, play a key role in assisting decision-making for different land use transformations and utilization in villages and towns, and make a scientific evaluation for the impact of social and economic development potential.

With the rapid development of intelligent vehicle technology, traditional vehicle design and manufacturing methods are facing many challenges. Traditional vehicle design and manufacturing methods have problems such as high cost, long development cycle and lack of personalized design. This paper took advantage of the high degree of flexibility, rapid customization and cost savings of 3D printing technology, which can provide entirely new solutions for intelligent vehicle design. This paper mainly analyzed the application of three-dimensional (3D) printing and manufacturing technology in intelligent vehicle design. In intelligent vehicle design, 3D printing can carry out 3D printed body structure design to make parts lightweight. The displacement sensor in the intelligent sensor can measure the displacement, deformation and vibration of the body components, and detect whether the body structure exceeds the safety range, which can provide safety feedback to the owner. 3D printing technology has been involved in automotive dashboard, interior design and customization. In this paper, the effects of vehicles designed under traditional methods and vehicles designed under 3D printing manufacturing technology were compared. The research and development cycle of class A, Class B, Class C and Class D in brand 1 under traditional manufacturing was 3.85 years, 3.13 years, 4.27 years and 4.32 years respectively. The R&D (research and development) cycle of grade A, B, C and D in brand 1 under 3D printing technology was 2.43 years, 2.05 years, 1.64 years and 1.73 years respectively. Under the traditional method, the average R&D cost of different levels of vehicles in brand 1, brand 2 and brand 3 was 191,000 yuan, 20900 yuan and 227,200 yuan respectively. Under 3D printing technology, the average R&D cost of different levels of vehicles in brand 1, brand 2 and brand 3 was 132,600 yuan, 145,300 yuan and 157,500 yuan respectively. The results of this paper emphasize that the application of 3D printing manufacturing technology shortens the development cycle of intelligent vehicles, reduces the development cost, and can be customized, which can be more in line with the needs of users.

Traditional digital image processing techniques face problems such as complex feature extraction and weak robustness when dealing with surface defects of multiple categories of electronic components. Deep learning is widely used in industrial defect detection. However, the performance of electronic component defect detection at the pixel segmentation level needs to be improved. For pixel-level defect detection, this paper constructs a defect detection model (ECSDDNet) for electronic component surface defects in computer vision engineering management system. To improve the segmentation accuracy and detection effect, three stages of experiments are conducted to address mis-segmentation problems and the shortcomings of the Unet network structure. Firstly, a classification network that can perform weight transfer is used to replace the encoding structure in the Unet network. Secondly, a simplified version of the feature fusion is proposed and added to the skip connection of the Unet network. Finally, label smoothing is used to optimize the loss and improve the generalization of the network. After the optimization experiment, some noisy contours and small defect contours that are mis-segmented are removed. Experimental results show that ECSDDNet has good segmentation effects on electronic component surface defects and can meet the segmentation and detection needs of electronic component surface defects.

With the rapid development of information technology, enterprise digitalization, and Internet + application, the digitalization and improvement of the used car trading business and its management process improvement are facing many challenges. This paper attempts to use the CMMI process improvement model to analyze the factors affecting the digital construction, business process optimization, and digital management process improvement of the used car market through research on the digital management process and the related theories of CMMI, such as talent training, application of information technology, immature business development, irregular business process, and industry policy restriction. The FAHP analysis method is used to establish a comparison matrix to explore the construction of an evaluation system model for the improvement of the digital management process of the used car trading market based on the CMMI model.

Highly dynamic Internet of Vehicles spectrum sharing can share spectrum owned by vehicle-to-infrastructure links through multiple workshop links to achieve efficient resource allocation. Aiming at the problem that the rapid variations in channel states in highly dynamic vehicular environments can make it challenging for base stations to gather and manage information about instantaneous channel states, we present a multi-agent deep reinforcement learning-based V2X spectrum access algorithm. The algorithm is designed to optimize the throughput of V2I user under V2V user delay and reliability constraints, and uses the experience gained from interacting with the communication environment to update the Q network to improve spectrum and power allocation strategies. Implicit collaborative agents are trained through an improved DQN model combined with dueling network architecture and long short-term memory network layers and public rewards. With lagged Q-learning and concurrent experience replay trajectories, the training process was stabilized and the non-stationarity problem caused by concurrent learning of multiple agents was resolved. Simulation results demonstrate that our presented algorithm achieves a mean successful payload delivery rate of 95.89%, which is 16.48% greater than that of the randomized baseline algorithm. Our algorithm obtains approximately the optimal value and shows performance close to the centralized brute force algorithm, which provides a better strategy for further minimizing the signaling overhead of the Internet of Vehicles communication system.

At present, as a research hotspot for time series data (TSD), the deep clustering analysis of TSD has huge research value and practical significance. However, there still exist the following three problems: (1) For deep clustering based on joint optimization, the inevitably mutual interference existing between deep feature representation learning progress and clustering progress leads to difficult model training especially in the initial stage, the possible feature space distortion, inaccurate and weak feature representation; (2) Existing deep clustering methods are difficult to intuitively define the similarity of time series and rely heavily on complex feature extraction networks and clustering algorithms. (3) Multidimensional time series have the characteristics of high dimensions, complex relationships between dimensions, and variable data forms, thus generating a huge feature space. It is difficult for existing methods to select discriminative features, resulting in generally low accuracy of methods. Accordingly, to address the above three problems, we proposed a novel general two-stage multi-dimensional spatial features based multi-view deep clustering method 1DCAE-TSSAMC (One-dimensional deep convolutional auto-encoder based two-stage stepwise amplification multi-clustering). We conducted verification and analysis based on real-world important multi-scenario, and compared with many other benchmarks ranging from the most classic approaches such as K-means and Hierarchical to the state-of-the-art approaches based on deep learning such as Deep Temporal Clustering (DTC) and Temporal Clustering Network (TCN). Experimental results show that the new method outperforms the other benchmarks, and provides more accurate, richer, and more reliable analysis results, more importantly, with significant improvement in accuracy and spatial linear separability.

Passive RFID has the advantages of rapid identification of multi-target objects and low implementation cost. It is the most critical technology in the Industrial Internet of Things information-gathering layer and is extensively applied in various industries, such as smart production, asset management, and monitoring. The signal collision caused by the communication between the reader/writer and tags sharing the same wireless channel has caused a series of problems, such as the reduction of the identification efficiency of the reader/writer and the increment of the missed reading rate, thus restricting the further development of RFID. At present, many hybrid anti-collision algorithms integrate the advantages of Aloha and TS algorithms to optimize RFID system performance, but these solutions also suffer from performance bottlenecks. In order to break through such performance bottleneck, based on the ISE-BS algorithm, we combined the sub-frame observation mechanism and the Q value adjustment strategy and proposed two hybrid anti-collision algorithms. The experimental results show that the two algorithms proposed in this paper have obvious advantages in system throughput, time efficiency and other metrics, surpassing existing UHF RFID anti-collision algorithms.

Today’s society has entered the pace of information development era. All kinds of information are digitized, while the loss of information data also directly affects the normal operation of the application system and has become the biggest obstacle to the development of information technology. The existing missing filling methods do not take into account the time-series information of the data set. Based on the neural network method for filling missing data in time series, an end-to-end missing data filling method for time series based on the residual of regression equation is proposed. Under the assumption that the activation function and the noise diffusion coefficient function are satisfied, the exponential stability of the complex-valued stochastic inertial neural network systems with additive time-varying delays is studied. The bipartite graph model of the time series data missing value filling method of the countermeasure network and the filling cyclic neural unit are generated. According to the end-to-end time series data missing value filling method of the self encoder, the corresponding low-dimensional feature vector can be automatically generated for each time series data. It avoids overfitting by using nonlinear methods directly in low-dimensional space. Experiments show that the method proposed in this paper uses Monte Carlo to fill data missing at different proportions. In data gap filling process, the coefficient variance is less than 0.05, which enhances the rationality of filling data. It is of great research value and practical significance to reasonably fill in the missing values of time series data. The research can accelerate the improvement of big data system, improve the level and effectiveness of database management, and is an important means of computing capacity of existing data centers.

Video action recognition is a technique for automatically determining the category of a video action. It is necessary to design an efficient video action recognition algorithm to predict video labels. This work proposes a video action recognition model based on dual-stream information fusion with attention mechanisms (DSIFAM), which consists of three different sub-modules. First, this proposes an improved keyframe extraction method (IKFE). Based on K-means clustering, this uses convolutional features to calculate the similarity between video frames instead of pixel points. After obtaining preliminary clustering results, the method performs secondary optimization to obtain more representative keyframes. Second, this proposes a video action recognition model based on dual-stream information fusion (DSIF). The method introduces ConvLSTM in the spatial stream and uses P3D instead of the original convolutional network in the temporal stream, which can better extract spatial-temporal information and improve the classification performance. Third, this designs a multi-scale attention mechanism (MSAM) to enhance the feature extraction stage and obtain higher quality classification features. The resulting features are more prominent and have stronger representation capabilities. Finally, this work conducts systematic experiments on different datasets, the results verify the superiority of DSIFAM for video action recognition.