To enhance the accuracy and reliability of large language models (LLMs) in regulatory question-answering tasks, this study addresses the complexity and domain-specificity of regulatory texts by designing a retrieval-augmented generation (RAG) testing framework. It proposes a dimensionality reduction-based semantic similarity measurement method and a retrieval optimization approach leveraging information reasoning. Through the construction of the technical route of the intelligent knowledge management system, the semantic understanding capabilities of multiple mainstream embedding models in the text matching of financial regulations are systematically evaluated. The workflow encompasses data processing, knowledge base construction, embedding model selection, vectorization, recall parameter analysis, and retrieval performance benchmarking. Furthermore, the study innovatively introduces a multidimensional scaling (MDS) based semantic similarity measurement method and a question-reasoning processing technique. Compared to traditional cosine similarity (CS) metrics, these methods significantly improved recall accuracy. Experimental results demonstrate that, under the RAG testing framework, the mxbai-embed-large embedding model combined with MDS similarity calculation, Top-k recall, and information reasoning effectively addresses core challenges such as the structuring of regulatory texts and the generalization of domain-specific terminology. This approach provides a reusable technical solution for optimizing semantic matching in vertical-domain RAG systems, particularly for MDSs such as law and finance.

To address the data quality challenges in ship maintenance data, such as high missing rates, anomalous noise, and multi-source heterogeneity, this paper proposes a data quality enhancement method based on a multi-strategy enhanced Harris Hawks Optimization algorithm for optimizing the Kolmogorov–Arnold Stacked Denoising Autoencoder. First, leveraging the Kolmogorov–Arnold theory, the fixed activation functions of the traditional Stacked Denoising Autoencoder are reconstructed into self-learnable B-spline basis functions. Combined with a grid expansion technique, the KASDAE model is constructed, significantly enhancing its capability to represent complex nonlinear features. Second, the Harris Hawks Optimization algorithm is enhanced by incorporating a Logistic–Tent compound chaotic map, an elite hierarchy strategy, and a nonlinear logarithmic decay mechanism. These improvements effectively balance global exploration and local exploitation, thereby increasing the convergence accuracy and stability for hyperparameter optimization. Building on this, an IHHO-KASDAE collaborative cleaning framework is established to achieve the repair of anomalous data and the imputation of missing values. Experimental results on a real-world ship maintenance dataset demonstrate the effectiveness of the proposed method: it achieves an 18.3% reduction in reconstruction mean squared error under a 20% missing rate compared to the best baseline method; attains an F1-score of 0.89 and an AUC value of 0.929 under a 20% anomaly rate; and stabilizes the final fitness value of the IHHO optimizer at 0.0216, which represents improvements of 31.7%, 25.6%, and 12.2% over the Particle Swarm Optimization, Differential Evolution, and the original HHO algorithm, respectively. The proposed method outperforms traditional statistical methods, deep learning models, and other intelligent optimization algorithms in terms of reconstruction accuracy, anomaly detection robustness, and algorithmic convergence stability, thereby providing a high-quality data foundation for subsequent applications such as maintenance cost prediction and fault diagnosis.

In response to the limitations of traditional static reliability analysis methods in characterizing the reliability changes of the Integrated Power System, this paper proposes a time-varying reliability analysis framework based on a Dynamic Bayesian Network. By embedding a multi-physics coupled degradation model into the conditional probability nodes of the Dynamic Bayesian Network, a joint stochastic differential equation for the degradation process was constructed, and the dynamic correlation between continuous degradation and discrete fault events throughout the entire life cycle was achieved. A modified method for modeling continuous degradation systems was proposed, which effectively solves the numerical stability problem of modeling complex degradation systems. Finally, the applicability and correctness of the model were verified through numerical examples, and the results showed that the analysis framework can be effectively applied to time-varying reliability assessment and dynamic health management of complex equipment systems such as the Integrated Power System.

Large language models (LLMs) offer new opportunities for intelligent decision support in logistics network modeling and optimization, yet existing studies mainly treat them as algorithmic assistants, direct optimizers, or end-to-end generators. Less attention has been given to organizing open-ended human–LLM interaction into a reliable decision process for strict constraints, incomplete specifications, and outputs requiring independent verification. This study develops a knowledge-based human–LLM interaction decision support system (K-HLIDSS) for logistics network modeling and optimization, using UAV emergency scheduling as a representative application setting. K-HLIDSS integrates model-element input structuring, knowledge-guided reasoning, and feedback verification to support a traceable and verifiable optimization decision chain. Six experiments combine mechanism-level benchmark validation and application-level case validation, covering structured model prior knowledge, model-element verification, workflow-based algorithm generation and solution execution, multi-stage five-algorithm generation with path-level reverse verification, real-world UAV emergency scheduling, and dynamic new-demand re-dispatching. On Cordeau benchmark instances, the four benchmark experiments show that structured prior knowledge increases model-element accuracy from 79.17% to 99.58%, and model-element verification raises it to 100.00%. Feedback verification increases the valid route output rate from 22.50% to 90.00% and reduces constraint violations from 17.50% to 1.25%. Across 33 instances, verified results selected instance by instance from five generated algorithms yield an average gap of 2.20%. In Shenzhen, China, two case experiments further demonstrate the system's capability for network modeling, route generation, and dynamic re-dispatching under realistic spatial and demand conditions. K-HLIDSS offers a process-level pathway for reliable AI-assisted logistics network optimization. Code and data are available at https://github.com/outhouse1/AI_UAV