Potato combine harvesters often face the challenge of balancing efficient potato–soil separation with minimizing tuber mechanical damage, which significantly affects harvest quality and economic returns. To address this issue, a dual-vibration potato–soil separation conveyor was designed based on agronomic planting parameters and soil physical characteristics. A high-fidelity DEM-MBD coupling simulation model was developed to analyze soil clod breakage behavior and potato collision-induced jumping dynamics, and to identify key operational factors influencing separation performance. The porosity was verified using computer vision combined with CT technology to ensure the model’s fidelity. Single-factor simulations and a central composite design (CCD) response surface experiment were conducted using potato damage rate and soil removal efficiency as evaluation indices. The results showed that the inclination angle α, conveying line speed V_f, and vibration frequency f were the dominant factors affecting separation efficiency and tuber integrity. Multi-objective optimization determined optimal operating parameters of α = 18.51°, V_f = 1.995 km·h⁻¹, and f = 6.22 Hz, under which soil removal efficiency reached 98.43% and the minimum damage rate was 1.60%. Field experiments using a 4U-1000 combine harvester verified the simulation results, with an average soil removal efficiency of 97.8% and an average damage rate of 1.62%. These findings confirm the accuracy of the DEM-MBD simulation model and provide theoretical guidance for optimizing separation devices in large-scale potato harvesting equipment.

Potato is one of the most important food crops worldwide, playing a critical role in global food security and agricultural production. The broken and impurity rates are important indicators for evaluating the harvesting quality of potato combine harvesting operations. To address the difficulty of achieving continuous and online detection using traditional methods, this study investigates an online monitoring approach for potato combine harvesting based on machine vision. Considering the characteristics of large material volume, severe overlap, and similar appearance features under field operating conditions, an online monitoring device suitable for potato combine harvesters was designed, along with a corresponding image acquisition and processing workflow. For the online monitoring device, an improved You Only Look Once version 11 (YOLOv11) detection model, was proposed to meet the requirements of multi-object detection in complex operating scenarios. The model incorporates Multi-Scale Depthwise Convolution (MSDConv), C2PSA_DCA (with Directional Context Attention, DCA), and Directional Selective Attention (DSA) modules, and introduces the Efficient Intersection over Union (EIoU) loss function to enhance recognition capability for broken potatoes and multiple types of impurity targets. While maintaining lightweight characteristics, the improved model demonstrates favorable detection accuracy. Field experiment results show that when the combine harvester operates at a forward speed of 3 km/h, the relative errors for broken and impurity rates are measured as 3.78% and 3.67%, respectively. Under extreme operating conditions with a speed of 4 km/h, the corresponding average relative errors rise to 8.30% and 8.72%, respectively. Overall, the online detection results exhibit satisfactory consistency with manual measurements, providing effective technical support for real-time monitoring of harvesting quality in potato combine harvesting operations. Future research will focus on expanding multi-scenario datasets under diverse soil and illumination conditions, as well as integrating detection results with adaptive control strategies to further enhance intelligent harvesting performance.

Aiming at the challenges of potato storage scenarios, severe target stacking and occlusion, and large variations in defect characteristics in potato storage environments, a potato defect detection method named MDS-DETR was proposed, and its deployment on a robotic sorting platform was validated. First, potato images under different illumination conditions and stacking states were collected in potato storage warehouses to construct a potato defect dataset containing defects such as black spot, sprouting, dry rot, decay, green skin, and cracking. Subsequently, according to the characteristics of object detection tasks in storage scenarios, an MCF module was designed to strengthen the extraction and integration of contextual features across different spatial scales. An AIFI-DyMona structure, namely Anchor-free Instance Feature Interaction with Dynamic Mona, was constructed to improve the stability of feature representation under different illumination conditions. In addition, the Shape-IoU regression strategy was incorporated to improve the network sensitivity to irregular defect contours and geometric characteristics. Validation experiments demonstrated that the proposed MDS-DETR framework achieved 96.3% mAP@0.5, with only 14.2 M parameters and 42.9 G FLOPs. Compared with several representative object detection algorithms, the proposed method exhibited superior recognition capability and stronger robustness under complicated storage conditions, while also showing better suppression of missed and incorrect detections. To further evaluate its practical applicability, the trained network was integrated into an intelligent potato sorting robot and tested in real warehouse scenarios. Experimental observations indicated that the robotic system consistently maintained a sorting accuracy exceeding 95%, demonstrating the effectiveness and practical deployment potential of the proposed approach for potato storage applications. This study can provide a reference for intelligent detection and automated sorting in potato storage processes.

This study addresses the limitation of single-factor environmental assessment in autonomous sweet potato farming under open-field conditions. An IoT-based sweet potato growth environment monitoring and comprehensive evaluation system was developed by integrating an STM32-based acquisition terminal, multi-sensor data collection, Narrow Band Internet of Things (NB-IoT) transmission, and cloud-based visualization. Five key environmental variables, namely soil temperature, soil moisture, soil available nitrogen, photosynthetically active radiation (PAR), and CO₂, were continuously monitored. To improve the evaluation of heterogeneous and uncertain environmental information, a multi-factor environmental quality assessment method combining fuzzy membership functions and an improved D-S evidence theory was proposed. Field experiments were conducted in Danzhou, Hainan, China, and 600 valid synchronized samples were obtained for analysis. The results showed that most samples were classified as Suitable (63.5%), followed by Normal (30.8%) and Poor (5.7%), with a mean comprehensive environmental score of 0.802. Among the monitored variables, PAR and soil temperature showed relatively high adaptive weights, indicating their important roles in environmental quality discrimination. Furthermore, the comprehensive environmental evaluation result exhibited a significant positive correlation with sweet potato yield (r = 0.6501, p = 2.3724 × 10⁻⁷³), demonstrating good explanatory ability for yield variation. The proposed system provides an effective technical framework for real-time environmental monitoring, quantitative suitability evaluation, and precision management in autonomous sweet potato farming.