In agricultural production, plant diseases and pests are among the major threats to crop yield and quality. Existing agricultural pest and disease identification methods have problems such as small target scales, complex background environments, and unbalanced sample distributions. This paper proposes a lightweight improved target detection model, YOLOv5s-LiteAttn. Based on YOLOv5s, the model introduces GhostConv and Depthwise Conv to reduce the number of parameters and computational complexity, and it combines CBAM and Coordinate Attention mechanisms to enhance the network’s feature representation capability. Experimental results show that, compared with the basic YOLOv5s model, the number of parameters of the improved model is reduced by 22.75%, and the computational load is reduced by 16.77%. At the same time, mAP@0.5–0.95 is increased by 3.3 percentage points, and recall is improved by 1.1 percentage points. In addition, the inference speed increases from 121 FPS to 142 FPS at an input resolution of 640 × 640, further confirming that the proposed model achieves a favorable trade-off between accuracy and efficiency. The average precision of YOLOv5s-LiteAttn is 97.1%, which outperforms the existing mainstream lightweight detection models. Moreover, an independent test set containing 4328 newly collected field images was established to evaluate generalization and practical applicability. Despite a slight performance decrease compared with the validation results, the model maintained an mAP@0.5–0.95 of 95.8%, significantly outperforming the baseline model, thereby confirming its robustness and cross-domain adaptability. These results confirm that the model has high precision and is lightweight, making it effective for the detection of agricultural diseases and pests.

Aiming at the problems of uneven radiation distribution and difficult regulation in double-film multi-span greenhouses in the Gobi Desert, a spatiotemporal simulation model of the radiation environment based on the coupling of Rhino–Grasshopper and Radiance was constructed in this study. Parametric simulation and multi-objective optimization were adopted to significantly improve the solar radiation capture and distribution uniformity inside the greenhouse, providing a scientific basis for greenhouse design in the Gobi area. The results show that the model has high accuracy (R² > 0.98), and the radiation inside the greenhouse presents a distribution pattern of “higher in the northeast, lower in the southwest, higher in the upper layer and lower in the lower layer”. The optimal orientation is 1° west of south, and the optimal configuration is 8 m span, 5 m eave height, and 30° roof slope. This study can provide quantitative support for the structural design, planting layout and energy-saving regulation of double-film multi-span greenhouses in arid desert areas, and has important practical value for promoting the efficient and sustainable development of facility agriculture in the Gobi Desert.

High summer temperatures, intense solar radiation, and restricted ventilation in arid desert regions often cause heat accumulation and spatially non-uniform temperature distributions in greenhouses, constraining sustainable protected agriculture. This study investigated a multi-span plastic greenhouse in Xinjiang, China, using a field-validated three-dimensional computational fluid dynamics (CFD) model. Nine combinations of side- and roof-vent openings were evaluated under sunny and cloudy conditions to compare crop-zone temperature and airflow characteristics. The results showed that side-vent opening primarily controlled outdoor-air inflow, airflow penetration, and crop-zone cooling, whereas roof-vent opening mainly affected upper-level heat removal and airflow continuity. Their performance therefore depended on coordinated opening ratios. Under sunny conditions, high-temperature zones above 45 °C developed near the roof and arch shoulders. Enlarging the side vents and appropriately matching the roof-vent opening effectively reduced heat accumulation. In the integrated crop-zone assessment, S3T3 achieved the lowest mean crop-zone temperature of 37.32 °C, 7.35% lower than that under S1T1. Its temperature standard deviation was σ = 0.012, and its cooling demand index was Y_c = 0.406, approximately 74.5% lower than that of S1T1. Optimizing side- and roof-vent combinations can therefore improve the greenhouse thermal environment without energy-intensive mechanical cooling, providing a basis for energy-efficient operation and sustainable protected crop production in extreme climates.