Photovoltaic industry trade has become a global trade activity, and a wide range of photovoltaic trade relations have been formed between countries. In order to further strengthen and balance trade relations, this paper analyzes global photovoltaic (PV) trade from the perspective of complex networks. We employ network indicators and the cascading process of risk propagation to analyze the evolution features and the vulnerability of the PV trade network. Firstly, we establish the global PV trade networks from 2000 to 2021 based on the PV trade flow between countries. We then explore evolution features and analyze the influencing factors of the trade network structure. Finally, we simulate the cascading process of risk propagation on the trade network based on an improved bootstrap percolation model. The evolution features reveal the following three results: (1) the scale of global PV trade continues to grow, and the participation of some countries has increased significantly; (2) the global PV trade network has small-world characteristics, and the related products have high circulation efficiency; and (3) the global PV trade network has a core-periphery structure, while a few countries drive most of the trade. China, Germany, and the U.S. are the top PV traders. Some Asian countries, such as Vietnam, are gradually increasing their share of the market. The QAP regression analysis shows that the gaps in GDP and electricity access rate are the biggest facilitating and hindering factors in the global PV trade, respectively. The simulation results show that the global PV trade network is vulnerable to targeted risk and is robust to randomness risk.

To avoid the step of manual feature engineering when predicting crystal properties, a graph convolutional neural network based on the dual attention mechanism, named DA-CGCNN, is proposed. It fuses both the channel attention mechanism and self-attention mechanism, named the dual attention mechanism, benefiting from capturing the complex features of each atom and dependencies between atomic nodes better. It is found to have comparable or superior performance to other advanced graph neural network (GNN) models by predicting five properties of the crystal: formation energy, total energy, bandgap, Fermi energy, and density. In addition, cross-property transfer learning is conducted on the computed properties from four small-sample crystal materials. The results show better performance on transferring prediction from these four samples. The proposed model in this study significantly improves the accuracy of crystal property prediction and demonstrates excellent prediction performance by incorporating transfer learning techniques. In summary, this work is important in accelerating the prediction of crystalline material properties and the discovery and design of crystalline materials.

With the rapid development of the Internet, businesses in the traditional catering industry are increasingly shifting toward the Online-to-Offline mode, as on-demand food delivery platforms continue to grow rapidly. Within these takeout systems, riders have a role throughout the order fulfillment process. Their behaviors involve multiple key time points, and accurately predicting these critical moments in advance is essential for enhancing both user retention and operational efficiency on such platforms. This paper first proposes a time chain simulation method, which simulates the order fulfillment in segments with an incremental process by combining dynamic and static information in the data. Subsequently, a GRU-Transformer architecture is presented, which is based on the Transformer incorporating the advantages of the Gated Recurrent Unit, thus working in concert with the time chain simulation and enabling efficient parallel prediction before order creation. Extensive experiments conducted on a real-world takeout food order dataset demonstrate that the Mean Squared Error of the prediction results of GRU-Transformer with time chain simulation is reduced by about 9.78% compared to the Transformer. Finally, according to the temporal inconsistency analysis, it can be seen that GRU-Transformer with time chain simulation still has a stable performance during peak periods, which is valuable for the intelligent takeout system.

In the current digital economy era, the takeout industry is expanding, which makes human-machine collaborative delivery management between takeout riders and intelligent algorithms increasingly important. Based on this, this paper takes takeout riders as Decision Making Units to study the management efficiency of the delivery algorithm from the perspective of input and output. Firstly, a comprehensive evaluation index system for input and output is constructed. Secondly, the entropy method is used to obtain the weights of the delivery input indicators at all levels and the comprehensive input index. Then, the output-oriented DEA-BCC model is established by combining the comprehensive input index and several delivery output indicators. Finally, the efficiency of the delivery algorithm in managing takeout riders is evaluated using the results calculated from the DEA-BCC model. Additionally, this paper also proposes suggestions for personalized human-machine collaborative delivery management in terms of quantity, quality and safety based on the slack variables of the output indicators.