Complex fluid–solid systems generally exist in process engineering. The cognition of complex flow systems depends on numerical and experimental methods. The computational fluid dynamics–discrete phase method simulation based on coarsening technology has potential application prospects in industrial-scale equipment. This review outlines the computational fluid dynamics–discrete phase method and its application in several typical types of process engineering. In the process research, more attention is paid to the dense condition and multiphase flow. Furthermore, the CFD-DPM and its extension method for comprehensive hydrodynamics modeling are introduced. Subsequently, the current challenges and future trends of the computational fluid dynamics–discrete phase method are proposed.

The Chinese mitten crab (Eriocheir sinensis) is a high-value seafood. Efficient quality-grading methods are needed to meet rapid increases in demand. The current grading system for crabs primarily relies on manual observations and weights; it is thus inefficient, requires large amounts of labor, is costly, and no longer meets the requirements for the market. Here, we employed computer vision techniques combined with deep learning modeling to efficiently quantify key physiological traits, such as sex identification, carapace dimensions (length and width), and fatness assessment for quality classification. To this end, a YOLOv5-seg integrated with an SE attention model was developed using 2282 RGB images and manual measurements of the physiological traits of 300 crabs. The RGB dataset was further augmented by rotating and resizing. The results revealed that the accuracy of sex recognition was 100%, and the mAP for carapace segmentation was 0.995, which was superior to YOLOv8-seg and other variants. In addition, we proposed an improved conditional factor K to evaluate the fatness of crabs and classify their quality based on fatness. The consistency between the grading method proposed in this article and manual grading was 100%. This study could aid in developing precise and non-destructive grading systems for the aquaculture and food industries.

Image datasets with imbalanced sampling, masking, missing and noise brought challenges to the development of an intelligent agricultural monitoring system. To tackle these issues, this paper proposes a cloud-based, multi-model integrated intelligent monitoring vision platform for agricultural greenhouse crops (named the CloudCropFuture platform), complete with algorithmic APIs, facilitating streamlined data-driven decision-making. For the CloudCropFuture platform, we first propose an image augmentation technology that employs an improved diffusion model to rectify deficiencies in image data, thereby enhancing the accuracy of agricultural image analysis. Experimental results demonstrate that on datasets enhanced by this method, the average precision of multiple YOLO models is improved by 5.6%. Then, a multi-level growth monitoring platform is introduced, integrating enhanced YOLOv11-based image models for more accurate and efficient crop observation. Furthermore, an intelligent model base comprising multiple integrated detection methods is established for assessing agricultural pests, maturity, and quality, leveraging the enhanced performance of vision models. CloudCropFuture offers a holistic solution for intelligent monitoring in agricultural greenhouses throughout the entire crop growth cycle. Through model verification and application across various greenhouse crops, this work has demonstrated the ability of the intelligent platform to provide reliable and stable monitoring performance. This research paves the way for the future development of agricultural technologies that can adapt to the dynamic and challenging conditions of modern farming practices.

Reliable bearing fault diagnosis is important for the stable operation of rotating machinery under complex and varying conditions. Existing graph-based diagnostic methods commonly rely on either predefined fixed topologies or fully data-driven graph learning, making it difficult to balance structural reliability with state-dependent adaptability. To address this issue, this paper proposes a Prior-Guided Heterogeneous Dynamic Graph Convolutional Network (PHDGCN). PHDGCN constructs two relation-specific prior graphs from causality-consistent temporal dependencies and frequency-domain covariate correlations and uses them as structural anchors. A history-conditioned residual learning mechanism then refines the two priors according to the current local signal state while preserving their structural constraints. The refined temporal and covariate graphs are processed by independent graph convolutional branches, and their complementary representations are fused for fault classification. Experiments on two public bearing fault diagnosis datasets show that PHDGCN achieves accuracy/macro-F1 values of 93.75%/93.34% and 89.47%/87.08%, respectively, outperforming the compared conventional and graph-based methods. Ablation and graph-structure visualizations further show the contribution of prior-guided dynamic refinement and the relation-specific adjustment behavior of the two graph branches.