Because the threshing device of a combine harvester determines the harvesting level and threshing separation performance of a combine harvester, the analysis and study of the threshing device of a combine harvester is key to improving its performance. Based on the threshing device of a half-feed combine harvester, the simulation model of a discrete element threshing device is established in this paper. With the threshing drum rotation speed, feed volume, and concave sieve vibration frequency as the variable factors, the BP neural network model and linear regression equation model established for the loss rate and impurity content for two kinds of threshing performance indicators, respectively, and through the discrete element threshing performance test, two kinds of methods of threshing performance prediction are analyzed. The results show that the neural network and linear regression can be used for the threshing performance indicators, however, the BP neural network prediction effect has a better prediction precision, better reliability, and the trained neural network can be used in the general case of the threshing performance indicators. This provides a new idea for improving the threshing performance of a combine harvester.

Micro-defects, such as ink spots, scratches, and sintering formed during the screen printing process of photovoltaic cells, significantly impair module performance. Traditional machine vision methods exhibit limited detection efficiency and high false-positive and missed-detection rates, while existing deep learning algorithms struggle to achieve accurate and adaptive detection of small-target defects and background similar defects in complex industrial environments. This study proposes an enhanced defect detection methodology based on an improved YOLOv8 algorithm. A multi-focus image acquisition platform using primary and auxiliary CCDs was independently developed, integrating a high-frame-rate industrial camera and a high-resolution electron microscope, with an LED ring light employed to suppress reflections, thereby establishing a high-quality dataset covering three defect categories. The algorithm was optimized through multiple dimensions: the RepNCSPELAN4 module was incorporated into the backbone network to improve multi-scale feature fusion, and a novel wavelet transform-based WaveConv module was designed to replace traditional downsampling, thereby better preserving defect edges and texture details. The neck network integrates a lightweight shuffle attention mechanism and a new detail enhancement module to strengthen critical features while controlling model complexity. Additionally, a dedicated auxiliary detection head was added for spotting tiny ink dots. Experimental results demonstrate a marked improvement in performance: on the custom dataset, the improved model achieves a stable mean average precision of approximately 92%. Specifically, ink spot detection reached a precision of 84.9% and recall of 77.7%, effectively reducing missed small-target defects; sintering defect detection attained 98.9% precision and 100% recall, addressing previous misclassifications due to background similarity; and scratch detection precision improved to 92.2%. Visual comparisons confirm that the enhanced model effectively overcomes the limitations of the original approach. By constructing a specialized dataset and implementing targeted, coordinated optimizations to the YOLOv8 architecture, this study significantly enhances the accuracy and robustness of screen-printing defect detection in photovoltaic cells, providing an effective solution for real-time online quality inspection in smart manufacturing lines.

This study addresses key challenges in high-precision industrial motion control, including dynamic load disturbances, nonlinear parameter coupling, and degradation in synchronization accuracy. A dual-motor cross-coupled synchronous control strategy is proposed, integrating Hertzian contact theory with an adaptive fuzzy PI control algorithm. First, a precise pressure measurement model for the printing contact zone is established based on Hertzian contact theory. The model quantitatively characterizes the relationship between structural parameters and pressure distribution. Key parameters include cylinder radius and plate thickness. This provides a theoretical foundation for precise regulation. Subsequently, a fuzzy PI controller with parameter self-tuning capability is incorporated into the motor speed loop, enabling real-time adjustment of control parameters to effectively compensate for system nonlinearities and time-varying disturbances. Furthermore, a cross-coupled synchronization architecture is designed to enable bidirectional compensation between the two motors, significantly improving synchronization accuracy under complex operating conditions. Simulations were performed in MATLAB/Simulink. The tests covered typical operational scenarios, including load start-up, single-motor disturbance, and multi-disturbance conditions. The results demonstrate that the proposed system achieves high performance: dual-motor speed synchronization accuracy reaches 99.5%; the response time for disturbance compensation is within 0.3 s; and printing-pressure fluctuation is confined to ±0.8%. This performance represents a 62.5% improvement in stability over conventional single-motor control systems. This research not only resolves the long-standing issue of pressure non-uniformity in flexographic printing but also provides a generalizable framework for multi-motor synchronous control in precision manufacturing. The findings offer substantial academic insight and practical value for advancing intelligent industrial measurement and control technologies.