Background: Serious secondary disasters caused by extreme natural weather conditions occur frequently, making it essential to establish a scientific and efficient modern emergency management system to maximize life-saving efforts. Methods: This study focuses on the uncertain environment of urban road networks and employs fuzzy theory to construct a 0–1 integer programming model for emergency evacuation paths that minimizes the average expected travel time. Results: We enhanced the neighborhood search strategy of the traditional ACO_time by incorporating the 2-opt and 3-opt perturbation mechanisms from the SA algorithm. Additionally, we utilized improved ant-volume and ant-perimeter models, along with their combinations, in the pheromone-updating mechanism of the basic ACO. The heuristic principles of the A* algorithm were integrated, introducing the joint influence of path and time into the heuristic function of the ACO algorithm. Conclusions: The IACO3 algorithm was tested on the Sioux Falls network and the Berlin Heisenheimer Center network. The computation time of the improved IACO3 algorithm was reduced by up to 20% compared to the original IACO3 algorithm in relation to the SA algorithm, with only a 4–5% increase in computation time compared to the ACO_time algorithm, which translates to an increase of merely 4–5 s. This demonstrates the superior solution efficiency of the IACO3 algorithm.

Recycling used battery packs for electric vehicles has important economic and ecological benefits. However, existing studies lack a systematic comparison of the different models. There is also a lack of exploration of symmetrical bilateral disassembly lines in comparison to other disassembly lines. Therefore, this paper takes the battery pack dismantling line of a company as the object of study. The disassembly line model is established by considering the number of workstations, disassembly time, disassembly costs, and disassembly risks. This study quantifies the risk of disassembly at each process into specific indicators for modeling. The solving algorithm adopts the improved Northern Goshawk Optimization Algorithm and the Non-dominated Sorting Genetic Algorithm. A simulation demonstration was conducted using plant simulation to compare the advantages and disadvantages of the unilateral disassembly and symmetric bilateral disassembly line programs. The results show that the optimized unilateral and symmetric bilateral disassembly line schemes increase the efficiency by 23.08% and 38.46%, respectively, compared to the original program. The symmetric bilateral disassembly line scheme is the most efficient among the programs. The optimized schemes significantly improve the overall operational efficiency of the disassembly line. Programs to promote increased efficiency in battery recycling will contribute to environmental sustainability.

Accurate vehicle trajectory prediction is crucial for autonomous driving systems, as it provides essential inputs for downstream tasks such as behavior planning and motion control. In complex traffic environments, a vehicle’s future motion depends not only on its own historical dynamics but also on interactions with surrounding vehicles, resulting in strong spatiotemporal dependencies. Although recent deep learning approaches have achieved promising performance, many existing methods focus primarily on modeling temporal sequences or spatial interactions, while insufficiently considering the heterogeneous importance of different feature channels in vehicle motion data. To address this limitation, this paper proposes a vehicle trajectory prediction framework that jointly models temporal dependencies and feature-channel relationships through a dual-attention architecture. Specifically, a temporal attention module is designed to capture long-range temporal dependencies in motion sequences, while a channel attention module learns adaptive importance weights for heterogeneous feature channels. A gated fusion mechanism is further introduced to integrate the complementary representations from the two branches. Based on the fused representation, a GRU-based decoder generates multi-step future trajectory predictions. The proposed model takes as input the motion history of a target vehicle and its six surrounding vehicle classes, and predicts trajectories over a 5 s prediction horizon using 3 s historical observations. Extensive experiments on the HighD and NGSIM datasets demonstrate that the proposed framework consistently outperforms several representative baseline methods across different prediction horizons, achieving average RMSE values of 0.78 m and 1.36 m, respectively. Ablation studies further confirm the effectiveness of the temporal attention module, channel attention mechanism, and gated fusion strategy. Overall, the proposed approach provides an effective solution for capturing both temporal dynamics and feature interactions in vehicle motion data, enabling more accurate and robust trajectory prediction in complex traffic environments.