Xuan Liu

With the rapid development of intelligent teaching technologies, automatic classroom behavior detection can provide objective visual evidence for classroom participation analysis and teaching feedback. However, real classroom scenes commonly involve dense student distributions, large scale variations between front and rear rows, frequent occlusion, and subtle visual differences between similar behaviors such as reading and writing. To address these challenges, this paper proposes SPD-Net, a classroom behavior detection network based on YOLO11m. The proposed method introduces three complementary improvements to the baseline: detail-enhanced upsampling for small-scale behavior regions, multi-scale and fine-grained feature enhancement for visually similar behaviors, and feature-conditioned spatial-prior attention for suppressing irrelevant background responses. Experiments are conducted on a three-class subset constructed from SCB-Dataset5, including reading, writing, and hand-raising. SPD-Net achieves a mean average precision (mAP) of 71.6% at mAP@0.5 and 54.7% at mAP@0.5:0.95, outperforming the YOLO11m baseline by 2.3 and 1.8 percentage points, respectively. In addition, the detected behavior instances are summarized into interpretable behavior-distribution information, providing support for classroom review and teaching-feedback analysis.

To address over-extrusion and forming defects at path corners caused by path overlap in additive manufacturing, this paper proposes an angle-adaptive look-ahead compensation algorithm based on a Sech attenuation curve. This method establishes a mapping model between the path angle and the adaptive look-ahead distance of the overlapping area, aiming to eliminate the material accumulation at the corner by precisely identifying the overlapping area and modulating the flow rate. By building a Beckhoff five-axis 3D-printing device and relying on the TwinCAT control platform, the compensation triggering logic based on PLC real-time Euclidean distance calculation was realized, and a slicing software with dynamic bias compensation was also developed. Experiments were conducted on triangular samples with extreme acute angles of 5°, universal acute angles of 85°, and 90° straight angles for printing verification. The results show that this algorithm can effectively suppress the material over-extrusion and accumulation at the path overlap in multiple angles, achieving a smooth transition of the sharp corners in the printed contour. The research confirms that the algorithm proposed in this study, together with the integrated software and hardware system, can ensure the forming accuracy of extreme and conventional geometric features while also guaranteeing the printing efficiency, providing an important reference for ensuring the quality coordination control of the formation process of extreme geometric features in additive manufacturing.

In traditional fixed-direction 3D printing, complex geometric models require support structures for large overhanging regions, leading to difficulties in removal, surface damage, and material waste. This study proposes a model decomposition algorithm for multi-axis 3D printing, aiming to divide the model into sub-regions requiring no or minimal support, which are printed sequentially via intermittent rotation of the print platform using a multi-DOF 3D printing device. Our algorithm employs a heuristic optimization method to minimize overhanging area and decomposition steps, leveraging Pareto multi-objective optimization to compute the Pareto front and adopting a beam search strategy to generate decomposition sequences. Decomposition and printing experiments on complex models demonstrate the method’s effectiveness in significantly reducing computational time, support requirements, and material costs while improving printing efficiency, enabling effective support-free multi-axis 3D printing.

When printing overhanging parts with traditional additive manufacturing (AM) equipment, it is necessary to add support structures under the overhanging structure. The process of printing support structures not only wastes materials, but also increases the manufacturing time. Therefore, in order to reduce or eliminate the need for support structures when printing parts with overhanging structures, such as propellers, a five-axis support-free printing method for overhanging parts is proposed for one of the most commonly used processes involving AM technology: fused deposition modeling (FDM). By offsetting the surface of the basic part, the offset surface is intersected with the model to be printed to obtain the spatial surface-layered curve. The contour offset method for the spatial curve is used to obtain the printing path, and continuous path planning is performed on it. While the presented method is targeted specifically at this ideal overhanging part, physical experiments on five-axis FDM equipment are performed. Compared with the traditional three-axis AM method, the time taken to print parts using this support-free five-axis AM method is shortened by 13.76–26.93%, and the printing material required is reduced by 17.24–29.29%. The experimental results show that this method realizes support-free printing of overhanging parts with the five-axis AM equipment, which not only saves materials and time consumed during part of the printing, but also improves the surface quality of the parts.

The research on additive manufacturing (AM) path planning mainly focuses on the traditional three-axis AM path planning and five-degree-of-freedom (DOF) AM path planning, while there is less research on six-DOF AM path planning. In the traditional AM path planning algorithm, the filling path is discontinuous and there is long straight-line printing in a certain direction, which can easily lead to warpage deformation. Therefore, in this work, the six-DOF manipulator is taken as the main object to build an AM platform, and the mechanism of AM path planning of the manipulator is studied. The path planning algorithm combining the contour offset filling method and Hilbert curve filling is optimized by using a cubic uniform B-spline curve, and an AM path planning algorithm suitable for a six-DOF manipulator is obtained. A continuous printing path can be generated by this algorithm. It reduces the existence of long straight-line printing in a certain direction, thereby reducing the warpage deformation of the model and improving the molding quality of the model. The traditional three-axis AM device and the six-DOF AM platform were used to print two kinds of models. By comparing the printing time, the six-DOF AM platform was 43.70% and 37.94% shorter than the traditional three-axis AM device. The same model was printed on a six-DOF AM platform by using the parallel scanning filling method, the path planning algorithm combining contour offset and Hilbert curve, and the method proposed in this paper. Through experimental verification, the average warpage deformation of the model printed by the method proposed in this paper was reduced by 37.81% and 13.79%, respectively, compared with the other two methods.

When using traditional 3D printing equipment to manufacture overhang models, it is often necessary to generate support structures to assist in the printing of parts. The post-processing operation of removing the support structures after printing is time-consuming and wastes material. In order to solve the above problems, a support-free five-degree-of-freedom additive manufacturing (SFAM) method is proposed. Through the homogeneous coordinate transformation matrix, the forward and inverse kinematics equations of the five-degree-of-freedom additive manufacturing device (FAMD) are established, and the joint variables of each axis are solved to realize the five-axis linkage of the additive manufacturing (AM) device. In this research work, initially, the layered curve is obtained through the structural lines of the overhang model, and a continuous path planning of the infill area is performed on it, and further, the part printing experiments are conducted on the FAMD. Compared with the traditional three-axis additive manufacturing (TTAM) method, the SFAM method shortens the printing time by 23.58% and saves printing materials by 33.06%. The experimental results show that the SFAM method realizes the support-free printing of overhang models, which not only improves the accuracy of the parts but also the manufacturing efficiency of the parts.

Aiming at the problems of the complex shape, difficult three-dimensional (3D) digital modeling and high manufacturing quality requirements of gas turbine blades (GTB), a method of fitting the blade profile line based on a cubic uniform B-spline interpolation function was proposed. Firstly, surface modeling technology was used to complete the fitting of the blade profile of the GTB, and the 3D model of the GTB was synthesized. Secondly, the processing parameters of the additive manufacturing were set, and the GTB model was printed by fused deposition technology. Then, the rapid investment casting was completed with the printed model as a wax model to obtain the GTB casting. Finally, the blade casting was post-processed and measured, and it was found to meet the requirements of machining accuracy and surface quality.