Critical equation of seedling block falling off in transplanting process and the optimization experiment of rape blanket seedling transplanter

A novel pneumatic recovery method was proposed to curb the problem of high losses caused by side-cutting in a rapeseed combine harvester header. The influence of recovery method and material status changes on the recovery effect was studied via the CFD-DEM (computational fluid dynamic- discrete element method) coupling simulation. The effect of airflow action on the recovery effect was compared and analyzed, and the composite pneumatic recovery method was determined. In addition, the influence of material status changes and material feeding rate on the recovery effect was explored, and the critical condition of material blockage in the recovery device was configured. As such, the relationship model between air velocity and recovery rate was constructed and the air distribution ratio of the flow field in the device under this condition was optimized, had verified the rationality of this pneumatic recovery method was verified by a series of field tests. The average rapeseed recovery rate of 92.95% was achieved with the application of the recovery device, and the total loss rate of the header reduced by 52.26%, which is of great significance in reducing the total loss rate of the combine harvesters and improving the operation performance of machinery. The research results can provide a reference for the design of the header structure of a rape combine harvester. Keywords: rapeseed, combine harvester, side-cutting loss, coupled simulation, pneumatic recovery DOI: 10.25165/j.ijabe.20221502.6267 Citation: Jiang T, Zhang M, Guan Z H, Mu S L, Wu C Y, Wang G, et al. Simulation and analysis of the pneumatic recovery for side-cutting loss of combine harvesters with CFD-DEM coupling approach. Int J Agric & Biol Eng, 2022; 15(2): 117–126.

Zero-shot learning (ZSL) is a powerful and promising learning paradigm for classifying instances that have not been seen in training. Although graph convolutional networks (GCNs) have recently shown great potential for the ZSL tasks, these models cannot adjust the constant connection weights between the nodes in knowledge graph and the neighbor nodes contribute equally to classify the central node. In this study, we apply an attention mechanism to adjust the connection weights adaptively to learn more important information for classifying unseen target nodes. First, we propose an attention graph convolutional network for zero-shot learning (AGCNZ) by integrating the attention mechanism and GCN directly. Then, in order to prevent the dilution of knowledge from distant nodes, we apply the dense graph propagation (DGP) model for the ZSL tasks and propose an attention dense graph propagation model for zero-shot learning (ADGPZ). Finally, we propose a modified loss function with a relaxation factor to further improve the performance of the learned classifier. Experimental results under different pre-training settings verified the effectiveness of the proposed attention-based models for ZSL.