Computer Science and Technology is an international, peer-reviewed open access journal dedicated to advancing research the field of computer science and technology. The journal provides a rapid publication process to ensure wide dissemination of high-quality articles to scientists, professionals, and interested individuals worldwide. Our goal is to serve as an efficient, reliable, and trusted platform for scholars and readers, publishing cutting-edge research in the field.
Abstract: To address the issue of decreased vehicle detection accuracy in rainy conditions due to reduced visibility, raindrop obstruction, and insufficient contrast, this paper proposes a rain-based vehicle detection method (R-RTDETR) based on an improved RT-DETR-R18 model. The original backbone network is replaced with a combined Conv and C2f-wConv network, and weighted convolutions are used to adaptively adjust features, enhancing the representation of key region information. In the encoder, a Global-Local Spatial Attention (GLSA) module is introduced to achieve efficient fusion of global semantics and local details. In the decoder, a P2 shallow feature branch is introduced to optimize the multi-scale fusion path and strengthen vehicle perception capabilities. Experimental results show that R-RTDETR improves mAP by 1.52% and recall by 1.25% on the test set compared to the baseline RT-DETR-R18; on the ACDC rainy dataset, mAP@0.5 is improved by 2.33%, demonstrating improved stability and robustness of rain-based vehicle detection while maintaining computational efficiency, validating its application potential in intelligent transportation scenarios.Abstract: To address the issue of decreased vehicle detection accuracy in rainy conditions due to reduced visibility, raindrop obstruction, and insufficient contrast, this paper proposes a rain-based vehicle detection method (R-RTDETR) based on an improved RT-DETR-R18 model. The original backbone network is replaced with a combined Conv and C2f-wConv network...Learn More
Abstract: Design of control systems needs a mathematics model of controlled plant in general, however the model is not exactly obtained and the model is uncertain, these limits led to the poor control system performance. The paper aim at when a motion (/reaction) speed of controlled plant process (/body) is greatly slower than speed of light, its motion process can described by Newton's Laws of Motion, and these laws can be applied to the design and analysis the control systems quality. This paper designed 3 control systems: PID control system with improved Smith predictor in time-invariance systems, PIDC control system with compensator (in time-variance uncertain systems) and CSMFDNLM (Control System of Uncertain System with Model-Free Design Based on Newton's Laws of Motion). A control system for uncertain plan with Model-Free is designed based on Newton's laws of motion in the paper. An OUAM filter based on Uniform Acceleration Movement is proposed by applying Kalman filter theory with constructing three state variables of the controlled system, i.e., position, speed and acceleration. The simulation and application results of PIDC control system with compensator and CSMFDNLM showed good control quality and robust performance for uncertain systems.Abstract: Design of control systems needs a mathematics model of controlled plant in general, however the model is not exactly obtained and the model is uncertain, these limits led to the poor control system performance. The paper aim at when a motion (/reaction) speed of controlled plant process (/body) is greatly slower than speed of light, its motion proc...Learn More
Abstract: For the problems of limited feature representation, unstable parameter learning, and insufficient reliability of classification decisions in online handwritten signature verification under small-sample conditions, a verification method integrating lightweight representation, multi-scale feature modeling, and probabilistic discrimination is proposed. Aiming at noise interference, feature redundancy, and insufficient utilization of dynamic information in online signature sequences, the method first performs smoothing on the original signature sequences, and then combines feature importance evaluation with Principal Component Analysis to conduct feature selection and dimensionality reduction, thereby constructing an input representation that contains both global statistical attributes and local dynamic variation information. In the feature extraction stage, Ghost feature mapping is adopted for the initial representation of input information, and the InceptionNext-TF module together with the DASE module is used for multi-scale deep feature modeling to characterize variation patterns of signature samples at different scales. In the classification stage, a variational Bayesian fully connected layer is introduced to model the output weights in a distributional manner, and a Bayesian optimization-based adaptive parameter search strategy is further employed to adjust the relevant key hyperparameters. Experimental results on the public MCYT-100 and SVC-2004 Task2 datasets show that, under the 10-shot setting, the Equal Error Rates are 1.46% and 3.05%, respectively.Abstract: For the problems of limited feature representation, unstable parameter learning, and insufficient reliability of classification decisions in online handwritten signature verification under small-sample conditions, a verification method integrating lightweight representation, multi-scale feature modeling, and probabilistic discrimination is proposed...Learn More
Abstract: Wire arc additive manufacturing (WAAM), owing to its advantages of high deposition efficiency, low cost, and low limitations on part dimensions, exhibits significant potential in the rapid fabrication of large and complex metal components. However, issues such as difficulties in geometrical accuracy control, internal micro-defects (e.g., porosity, cracks), and anisotropy of mechanical properties caused by the cyclic thermal history during layer-by-layer deposition severely restrict the industrial application and reliability of this technology. In response to the strongly coupled, nonlinear, and randomly disturbed characteristics of WAAM process parameters, intelligent online monitoring has become a key approach to improving manufacturing quality. This paper first reviews the diverse sensing schemes for WAAM process information monitoring, covering vision-based (visible light, laser, infrared, X-ray), acoustic emission, and arc spectroscopy methods, and comparatively analyzes the performance boundaries of each technique in capturing melt pool morphology and internal quality features. Subsequently, the research progress of machine learning (ML) in signal feature extraction and state classification is discussed, with emphasis on the robustness of deep neural networks and support vector machines in defect identification and process prediction. On this basis, intelligent control strategies based on data-driven models and reinforcement learning algorithms are summarized, and their application value in achieving adaptive control of key parameters such as droplet transfer and wire feed speed is explored. Finally, in view of current bottlenecks including difficulties in multi-source data fusion, scarcity of high-quality sample datasets, and weaknesses in multivariable cooperative control, the development trends of intelligent WAAM manufacturing are prospected.Abstract: Wire arc additive manufacturing (WAAM), owing to its advantages of high deposition efficiency, low cost, and low limitations on part dimensions, exhibits significant potential in the rapid fabrication of large and complex metal components. However, issues such as difficulties in geometrical accuracy control, internal micro-defects (e.g., porosity, ...Learn More
Abstract: With the rising living standards and shifting consumer habits, an increasing number of individuals have embraced online shopping. Concurrently, merchants are deploying intensive promotional campaigns, particularly during annual events such as "618," "Double 11," and live-streaming sales, which often trigger instantaneous "order explosion" surges. To address the efficiency-safety conflict arising from intense spatial competition within large-scale, high-density human-robot collaborative smart warehouses under such emergency conditions, this paper proposes an Emergency Dynamic Isolation System (EDIS). The system utilizes Ultra Wide Band (UWB) positioning and Long Short-Term Memory (LSTM) network-based trajectory prediction to perceive real-time human and robot dynamics. It then calculates and generates temporary, minimized virtual safety boundaries based on an improved Dynamic Risk Field Model (DRFM). To evaluate its effectiveness, a high-fidelity discrete-event simulation model was developed in Python, simulating the operation of a super-large book warehouse—with a total area of 30,000 square meters, storing 100,000 bins and 20,000 pallets, and equipped with 285 bin-carrying robots, 57 shuttle robots, 20 pallet-pulling robots, 15 autonomous three-directional forklifts, and 70 operators—under a 500% surge in order volume. Comparative experiments with traditional static physical isolation and static electronic fencing schemes demonstrate that, while ensuring zero collision risk, EDIS improves the overall order completion rate by 46.8%, reduces the average walking distance of pickers by 31.2%, achieves a peak dynamic space-sharing rate of 78.4% in key aisles, and decreases production interruptions due to avoidance maneuvers by 94.7% during the 6-hour emergency peak period. This study provides a quantitatively validated innovative solution for ultra-large, high-density human-robot collaborative smart warehouse systems to cope with extreme operational fluctuations.Abstract: With the rising living standards and shifting consumer habits, an increasing number of individuals have embraced online shopping. Concurrently, merchants are deploying intensive promotional campaigns, particularly during annual events such as "618," "Double 11," and live-streaming sales, which often trigger instantaneous "order explosion" surges. T...Learn More
