Automatic modulation classification (AMC) is one of the most important technologies in various communication systems, including drone communications. It can be applied to confirm the legitimacy of access devices, help drone systems better identify and track signals from other communication devices, and prevent drone interference to ensure the safety and reliability of communication. However, the classification performance of previously proposed AMC approaches still needs to be improved. In this study, a dual-stream spatiotemporal fusion neural network (DSSFNN)-based AMC approach is proposed to enhance the classification accuracy for the purpose of aiding drone communication because SDDFNN can effectively mine spatiotemporal features from modulation signals through residual modules, long-short term memory (LSTM) modules, and attention mechanisms. In addition, a novel hybrid data augmentation method based on phase shift and self-perturbation is introduced to further improve performance and avoid overfitting. The experimental results demonstrate that the proposed AMC approach can achieve an average classification accuracy of 63.44%, and the maximum accuracy can reach 95.01% at SNR = 10 dB, which outperforms the previously proposed methods.
The movable fluid percentage and movable fluid porosity of rocks are important parameters for evaluating the development potential of petroleum reservoirs, which are usually determined by expensive and time-consuming low-field nuclear magnetic resonance (NMR) experiments combined with centrifugation. In this study, an NMR proxy model based on adaptive ensemble learning was proposed to predict the rock movable fluid indexes efficiently and economically. We established adaptive ensemble learning via an opposite political optimizer (AEL-OPO), which adaptively combines 33 base learners through political optimization to increase the prediction accuracy of the NMR proxy model. To improve the generalization ability of the AEL-OPO, opposition-based learning was introduced to improve the global search speed and stability of the political optimizer. Accessible petrophysical parameters such as rock density, porosity, permeability, average throat radius, and maximum throat radius were used as a training set, a validation set, and a test set. The prediction results show that our new strategy outperforms the other 33 base learners, with R2 (coefficient of determination) values of 84.64% in movable fluid percentage and 74.09% in movable fluid porosity.
Urban infrastructure resilience has become a global priority as cities concentrate growing shares of population, wealth and critical assets amid intensifying disaster risks. Yet globally applicable city-scale assessment remains constrained by three challenges: the lack of a common city-level assessment method applicable worldwide, the scarcity and largelyunstructured nature of resilience-indicator data, and the difficulty of integrating qualitative assessments with structured quantitative data from multiple sources. To address these challenges, this study develops the Global City Infrastructure Resilience Index (GCIRI). Building on the Global Infrastructure Resilience Index (GIRI), GCIRI combines a city-scale extension of the GIRI framework, multi-agent parliamentary assessment and Bayesian data fusion. Eight large language model agents assess resilience indicators from fragmented urban evidence and exchange supporting rationales to cross-check and refine their judgements. Bayesian data fusion integrates these assessments with available structured city measurements and corresponding national GIRI priors. The fused absorption scores then provide the vulnerability term required for physical-risk downscaling, after which the inherited GIRI formulation yields the city level absorption, response and recovery quantities and final GCIRI. Across 50 globally distributed benchmark cities, parliamentary deliberation reduces cross-agent score dispersion by 51.9%, compared with 0.6% when peer reasoning is scrambled. A combined comparison with two independent city rankings yields 83.9% pairwise concordance. The analysis identifies four distinct resilience modes and their associated capacity weaknesses. Exposure amplifies capacity differences, magnifying resilience weaknesses where capacity is limited while highlighting resilience advantages where capacity is strong. These findings support mode-specific resilience-development strategies.
Interpreting is a cognitively demanding task requiring the simultaneous management of multiple processes under severe time constraints. While quantifying interpreting difficulty is essential for training and assessment, existing methods struggle to reconcile predictive accuracy with explanatory power. To bridge this gap, this study introduces a novel, interpretable evaluation framework based on Kolmogorov-Arnold Networks (KAN). Unlike black-box deep learning models, KAN employ learnable univariate functions on network edges, enabling the direct extraction of explicit symbolic formulas from data. Our model achieved superior predictive accuracy (R² = 0.9893), reducing the Mean Squared Error (MSE) by 49.8% compared to a standard BP Neural Network. Global sensitivity analysis of the derived formula revealed key quantitative laws governing difficulty, most notably a non-linear threshold effect for speech rate and a synergistic composite effect between noise and context dependency. These findings provide a transparent, data-driven foundation for cognitive load theories in interpreting and offer practical implications for developing intelligent, adaptive interpreter training systems.