Focused crawlers are targeted to search the internet for web pages on specific topics. Its main task is to collect preprocessed and topic related web pages and ignore irrelevant web pages. Traditional focused crawlers have limited success in achieving multi-text categorization of web pages. Due to the large amount of unstructured data present in web pages, the correct classification of web pages based on a given topic is the main practical challenge for focused crawlers.The main objective of this work is to design an improved focused crawling approach using web page classification. In this paper, a text classification model based on the combination of GloVe word vector model and TF-IDF weighting technique is proposed to improve the accuracy of web page classification. The GloVe-based text classification model is further utilized to guide focused crawlers to classify web pages.The proposed GloVe and TF-IDF text categorization models are validated on 10 different datasets and the results are compared with traditional machine learning algorithms as well as different methods based on Naive Bayes, Bag-of-Words and Word2Vec. According to the experimental results, the proposed text classification model is 7-12% better than traditional machine learning algorithms.

In order to solve the problems of traditional traffic accident scene investigation, such as taking a long time, evidence easily lost and difficult to save in case of bad weather, low survey accuracy, and field measurement data, DJI Mavic 3E UAV is used to convert the collected data into digital two-dimensional ortho image and three-dimensional model by using DJI Intelligent map software, such as mid-way point flight, map construction aerial photography and oblique shooting. One-stop help traffic accident investigation comprehensively improve the efficiency of scene investigation, standard forensics, improve the accuracy of accident scene investigation, in order to quickly restore traffic order, ease the demand for police, and improve the identifiability, safety and timeliness of traffic accident scene investigation.

By improving the standard U-Net architecture, this paper proposes a novel semantic segmentation model, which incorporates multiple attention mechanisms to enhance the model’s capacity to capture multi-scale features. Specifically, we introduce the Efficient Multi-Scale Attention Module with CrossSpatial Learning (EMA), Spatial and Channel Squeeze and Excitation (SCSE), and Squeeze-andExcitation (SE) mechanisms into the standard U-Net network. These modules assist the network in learning significant information from feature maps at multiple scales while suppressing interference from irrelevant background. Experimental results demonstrate that incorporating attention mechanisms effectively enhances the prediction accuracy of the standard U-Net network for lane line semantic segmentation. The new model outperforms the standard U-Net model on our custom dataset, with particularly significant improvements in lane detection accuracy in scenarios with certain interference.

This research proposes a new optimization technique for reinforcement concrete filled structural tubular columns using genetic algorithms and unified strength theory. A complete theoretical model to determine the axial bearing capacity of reinforced CFST columns incorporating modified confinement coefficients and enhanced steel section properties was developed. The optimization procedure deals with performance of the structure, materials usage and construction convenience as the optimization goals. Experimental validation for ultimate bearing capacity of five full scale specimens was carried out and the deviation was found out to be 5.2% which was found to be predicted by the theoretical model accurately. Internal stiffeners are likely to increase axial capacity by about 15.7%-23.4% over traditional CFST columns. The relationship between stiffener parameters and performance of the structure was found to be critical with optimal height to thickness of the stiffener to be in the range of 30 to 45 and space to diameter ratio no greater than 0.5. The problem sets out such mathematics as is nowadays simply necessary for the modern construction world to have at their disposal, as well as reasons for designing reinforced CFST columns.

This paper presents the design and implementation of the non-electric contacting power supply system for the electronic scale, which mainly focuses on improving power transfer and measurement accuracy. The whole system architecture includes electromagnetic coupling, an advanced algorithm of control, and safety. Simulation results have shown that, under standard conditions, it is possible to reach a high power transfer efficiency higher than 90% while keeping voltage regulations within ±1% and limiting current ripple to below ±1.8%. Therefore, this provides a measurement resolution of ±0.1g for the system while granting stable performance for a variety of conditions in both coupling and load. Protection mechanisms set within the system ensure reliable operation; fault detection time is less than 10μs. The proposed method represents a relatively good guide for non-contact power supply towards precision measurement, thus solving the challenge of WPT in an electronic scale system.

Under the background of globalization and knowledge economy, the importance of innovation and entrepreneurship education for college students is becoming more and more prominent. This paper combines fuzzy logic and decision tree algorithm to construct a cultural confidence recognition model of innovation and entrepreneurship education. Feature selection and classification are carried out on the salient features of the collected data information on innovation and entrepreneurship education. First, eight types of statistical features, such as the degree of integration of excellent traditional culture, the degree of value leadership and moral cultivation, the innovative power of grounded cultural knowledge, and the effect of social responsibility cultivation, are extracted as inputs to the C4.5 algorithm, and a decision tree is constructed for feature selection. Then, according to the constructed decision tree, the affiliation function and IF-THEN rule of the fuzzy inference model are designed. Finally, the designed fuzzy inference model is used to classify the degree of cultural confidence. The method achieves 100% accuracy in recognizing the lack of cultural self-confidence in innovation and entrepreneurship education, and more than 90% in recognizing the overall effect of general cultural self-confidence and rich cultural self-confidence. The experimental results show that the combination of decision tree and fuzzy inference modeling is feasible for the detection and classification of college students’ innovation and entrepreneurship education, and has strong practical application value.

The development of blockchain technology in modern business and finance is of great importance. The study delves into the blockchain-based shareholder voting system and the role of blockchain on corporate governance. On this basis, relevant research hypotheses are formulated. After completing the definition of research variables, the research model is constructed to empirically investigate the impact of blockchain-based shareholder voting system on corporate governance. The research hypotheses are tested through regression analysis and the robustness test is utilized to ensure the reliability of the research findings. The minimum value of blockchain-based shareholder voting and corporate governance level are both 0, the maximum value is 4.954, 0.624, and the average value is 0.821, 0.089, respectively. There is variability in shareholder voting and corporate governance level across companies. Before and after the control variables, the coefficients of blockchain-based shareholder voting system are 0.225 and 0.247 respectively, and both are significantly positive at 1% level. Blockchain-based shareholder voting system can improve corporate governance.

This paper constructs the key business index system of electric power system consisting of electric power supply, electric power transmission, electric power distribution, electric power equipment and electric power system management. By evaluating the validity optimization, reliability optimization, and redundant indicator removal based on the neural network analysis method of the indicator system, a new power system key business indicator system is formed, and the weights of the optimized indicators are calculated. The power system key business indicator control program is designed based on the weight parameters, and a new power system key business indicator control platform is developed. Extract power data using the weighted FCM clustering algorithm, and classify user power data on the cloud platform. Resource utilization and performance response analysis are performed on the power system key business index control platform. The power system key business index control platform designed by index weights developed in this paper is able to meet the transaction demand under different concurrent user numbers, and always maintains a memory utilization rate within 10, with good operating conditions.

In the context of the digital economy driven by the Internet of Everything, the dissemination of cultural heritage is facing the challenge of transitioning from traditional to digital media. The study develops an introduction to the visual SLAM system, models the binocular camera configuration and the indoor and outdoor dense 3D reconstruction process, and designs a complete set of algorithms based on the calibration of the actual binocular camera, image correction, binocular stereo matching algorithm (SELAS), and real-time dense point cloud 3D reconstruction. Based on the real laboratory scene, the original ELAS algorithm is compared with the improved method for experiments, and the results show that the mean value of the deviation of the optimized S-ELAS algorithm is -0.046m, and the algorithm accuracy is remarkable. Then a virtual cultural relics museum based on the combination of visual SELAM system and VR technology is designed to realize close interaction with S-ELAS stereo matching algorithm. In order to test the performance of the designed cultural relics museum system, the users are firstly acclimatized, and then the screened users are tested to experience the virtual museum system, and the MOS scores are made after the test. The MOS scores show that the virtual cultural relics museum system has better interactivity and experience.

In order to alleviate the problems of short supply of parking spaces and traffic congestion, intelligent driving solutions have emerged. Automatic parking has now become the first application scenario for driverless driving due to the more fixed scenario and lower traveling speed. In this study, the traditional A* algorithm is improved using the cost function, and the hybrid algorithm of parking space path search and planning is designed by combining the improved A* algorithm with the Reeds-Shepp curve, and then combined with the collision constraints to improve the algorithm’s path planning performance. The results of simulation experiments and in-loop test experiments show that the maximum lateral error and heading error are low in parallel and perpendicular parking scenarios, and it is found that the average lateral error during the whole parking process is only 0.177m in the in loop test, which is a good tracking effect for vehicles. The path search and planning algorithm designed in this paper can better realize the autonomous parking function and has high tracking accuracy and stability in the simulation scenario.