Advanced Engineering Informatics最新文献_第10页

AutoTRIZ: Automating engineering innovation with TRIZ and large language models AutoTRIZ：使用TRIZ和大型语言模型自动化工程创新

IF 8 1区工程技术

Advanced Engineering Informatics Pub Date : 2025-04-01 DOI: 10.1016/j.aei.2025.103312

Shuo Jiang , Weifeng Li , Yuping Qian , Yangjun Zhang , Jianxi Luo

{"title":"AutoTRIZ: Automating engineering innovation with TRIZ and large language models","authors":"Shuo Jiang , Weifeng Li , Yuping Qian , Yangjun Zhang , Jianxi Luo","doi":"10.1016/j.aei.2025.103312","DOIUrl":"10.1016/j.aei.2025.103312","url":null,"abstract":"<div><div>Various ideation methods, such as morphological analysis and design-by-analogy, have been developed to aid creative problem-solving and innovation. Among them, the Theory of Inventive Problem Solving (TRIZ) stands out as one of the best-known methods. However, the complexity of TRIZ and its reliance on users’ knowledge, experience, and reasoning capabilities limit its practicality. To address this, we introduce AutoTRIZ, an artificial ideation system that integrates Large Language Models (LLMs) to automate and enhance the TRIZ methodology. By leveraging LLMs’ vast pre-trained knowledge and advanced reasoning capabilities, AutoTRIZ offers a novel, generative, and interpretable approach to engineering innovation. AutoTRIZ takes a problem statement from the user as its initial input, automatically conduct the TRIZ reasoning process and generates a structured solution report. We demonstrate and evaluate the effectiveness of AutoTRIZ through comparative experiments with textbook cases and a real-world application in the design of a Battery Thermal Management System (BTMS). Moreover, the proposed LLM-based framework holds the potential for extension to automate other knowledge-based ideation methods, such as SCAMPER, Design Heuristics, and Design-by-Analogy, paving the way for a new era of AI-driven innovation tools.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"65 ","pages":"Article 103312"},"PeriodicalIF":8.0,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143748047","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A skill vector-based multi-task optimization algorithm for achieving objectives of multiple users in cloud manufacturing 基于技能向量的云制造多用户目标多任务优化算法

IF 8 1区工程技术

Advanced Engineering Informatics Pub Date : 2025-03-31 DOI: 10.1016/j.aei.2025.103295

Yixiao Jiang, Dunbing Tang, Haihua Zhu, Changchun Liu, Kai Chen, Zequn Zhang, Jie Chen

{"title":"A skill vector-based multi-task optimization algorithm for achieving objectives of multiple users in cloud manufacturing","authors":"Yixiao Jiang, Dunbing Tang, Haihua Zhu, Changchun Liu, Kai Chen, Zequn Zhang, Jie Chen","doi":"10.1016/j.aei.2025.103295","DOIUrl":"10.1016/j.aei.2025.103295","url":null,"abstract":"<div><div>Cloud Manufacturing (CMfg) is a new manufacturing mode that provides efficient manufacturing services to customers by centrally scheduling manufacturing resources distributed across various regions. In CMfg, each participant is an independent economic entity with distinct objectives and effectively achieving the objectives of customers, suppliers, and the CMfg platform under limited resources is a significant challenge. To solve this problem, this study first proposed a three-level multi-task optimization (TMTO) model. The upper-level and lower-level of the TMTO model respectively optimize the personalized objectives of customers and suppliers, as well as the objectives of the CMfg platform are optimized at the middle-level. Subsequently, a skill vector-guided multi-task optimization algorithm (SMTOA) is proposed to collaboratively optimize the objectives of all participants, with the skill vector designed to evaluate the ability of scheduling schemes to meet the objectives of all customers and suppliers. Finally, experimental cases based on an aerospace manufacturing enterprise confirm the effectiveness of the TMTO model and the advantages of SMTOA in solving the TMTO model.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"65 ","pages":"Article 103295"},"PeriodicalIF":8.0,"publicationDate":"2025-03-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143739426","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

3D-MMFN: Multi-level multimodal fusion network for 3D industrial image anomaly detection 3D- mmfn：用于三维工业图像异常检测的多级多模态融合网络

IF 8 1区工程技术

Advanced Engineering Informatics Pub Date : 2025-03-31 DOI: 10.1016/j.aei.2025.103284

Mujtaba Asad , Waqar Azeem , Aftab Ahmad Malik , He Jiang , Ahmad Ali , Jie Yang , Wei Liu

{"title":"3D-MMFN: Multi-level multimodal fusion network for 3D industrial image anomaly detection","authors":"Mujtaba Asad , Waqar Azeem , Aftab Ahmad Malik , He Jiang , Ahmad Ali , Jie Yang , Wei Liu","doi":"10.1016/j.aei.2025.103284","DOIUrl":"10.1016/j.aei.2025.103284","url":null,"abstract":"<div><div>3D-based image anomaly detection (AD) is a crucial computer vision task in industrial manufacturing. Most existing methods only focus on 2D shape-based detections. However, there is still limited research for detecting anomalies in 3D shapes using multimodal features. Some existing techniques developed for this task are mostly unsuitable for industrial defect detection for several reasons. Firstly, they rely mostly on memory banks, resulting in high storage overheads, making them difficult to deploy on production lines. Secondly, the multimodal features, in the existing 3D industrial AD algorithms, are concatenated directly which cause a significant disruption between the features and degrades the detection efficiency. Thirdly, their inference speed is not fast enough to achieve real-time detection. To address these challenges, we propose a deployment-friendly network named 3D-MMFN. Our model comprises of the following components: (1) The pre-trained feature extractors that generate local features from multi-stream inputs of RGB images, surface normal maps, and point clouds. (2) A novel point-to-pixel based fusion module that efficiently fuses multi-level multimodal features to mitigate disruption during the fusion operation. (3) An anomaly generator module that generates anomalous features from normal multimodal fused features, enabling self-supervised training of 3D-MMFN while eliminating the need for extensive memory banks. Experimental results on the MVTec3D-AD and Eyecandies dataset demonstrate the effectiveness of our proposed model, showcasing significant performance improvements over state-of-the-art methods.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"65 ","pages":"Article 103284"},"PeriodicalIF":8.0,"publicationDate":"2025-03-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143739427","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Predicting water demand for spraying operations in dry bulk ports: A hybrid approach based on data decomposition and deep learning 预测干散货港口喷洒作业的用水需求：基于数据分解和深度学习的混合方法

IF 8 1区工程技术

Advanced Engineering Informatics Pub Date : 2025-03-31 DOI: 10.1016/j.aei.2025.103313

Jiaqi Guo , Wenyuan Wang , Chi Wai Kwong , Yun Peng , Zicheng Xia , Xin Li

{"title":"Predicting water demand for spraying operations in dry bulk ports: A hybrid approach based on data decomposition and deep learning","authors":"Jiaqi Guo , Wenyuan Wang , Chi Wai Kwong , Yun Peng , Zicheng Xia , Xin Li","doi":"10.1016/j.aei.2025.103313","DOIUrl":"10.1016/j.aei.2025.103313","url":null,"abstract":"<div><div>Dust pollution from materials in dry bulk ports (DBPs) significantly impacts air quality and public health in coastal cities. Spraying operations are the primary dust control measures in ports and accurately predicting water demand for these operations helps optimize water scheduling and conserve resources. However, challenges remain in addressing non-stationary time series and improving prediction accuracy. Additionally, existing studies rarely consider the impacts of port operations on water demand for spraying. Therefore, this study proposes a hybrid approach based on data decomposition and deep learning to predict water demand for spraying operations in DBPs. Port operational data is specifically integrated into the input features. An optimal variational mode decomposition (OVMD) method is introduced to reduce data non-stationarity. Compared to other methods, OVMD adaptively selects the optimal modes and effectively mitigates mode mixing issues. The 1-D Convolutional Neural Network integrated with an Attention BILSTM model, combined with OVMD, an Artificial Neural Network, and Error Correction, is employed to capture long-term temporal dependencies. Moreover, the relationship between material surface moisture content and water consumption for spraying operations is uniquely incorporated into the prediction process. This approach is compared with benchmark models using a dataset from a DBP in northern China. The results demonstrate that the proposed method achieves superior predictive performance, with a MAE of 0.47, a RMSE of 0.71, and an R2 of 0.95. The proposed approach enables port operators to accurately determine water consumption for spraying operations, thereby promoting the intelligent and sustainable development of dust control in DBPs.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"65 ","pages":"Article 103313"},"PeriodicalIF":8.0,"publicationDate":"2025-03-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143738838","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Anomaly detection of machinery under time-varying operating conditions based on state-space and neural network modeling 基于状态空间和神经网络建模的机械时变工况异常检测

IF 8 1区工程技术

Advanced Engineering Informatics Pub Date : 2025-03-29 DOI: 10.1016/j.aei.2025.103285

Zimin Liu , Zihao Lei , Guangrui Wen , Yue Xi , Yu Su , Ke Feng , Xuefeng Chen

{"title":"Anomaly detection of machinery under time-varying operating conditions based on state-space and neural network modeling","authors":"Zimin Liu , Zihao Lei , Guangrui Wen , Yue Xi , Yu Su , Ke Feng , Xuefeng Chen","doi":"10.1016/j.aei.2025.103285","DOIUrl":"10.1016/j.aei.2025.103285","url":null,"abstract":"<div><div>Anomaly detection is critical for maintaining the health and stability of machinery. However, machines such as wind turbines often operate under time-varying conditions (TVCs), where changes in operating conditions (OCs) introduce disturbances to sensor signals, presenting significant challenges for traditional anomaly detection methods. To address this issue, this paper proposes a novel anomaly detection approach based on state-space and neural network modeling. First, from the perspective of system dynamic response, the machine operating under TVCs is treated as a dynamic response system, with OCs and health states governing the system’s dynamic response. A state-space model is then employed to explicitly model the health state, OCs, and response signals during the normal operation of machinery. Additionally, the nonlinear fitting capability of neural networks is used to parameterize the relationships between these factors. By incorporating OCs and health states into the model, the time-varying response induced by the two factors is effectively modeled as a time-invariant process. Furthermore, an alternating parameter update strategy, utilizing the extended Kalman filter, is developed to estimate both the health state and neural network parameters. Finally, a detection indicator is constructed based on the real-time neural network parameters to achieve machinery anomaly detection. The effectiveness and superiority of the proposed method are validated through simulation experiments and accelerated fatigue degradation experiments on rolling bearings under different time-varying operating conditions.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"65 ","pages":"Article 103285"},"PeriodicalIF":8.0,"publicationDate":"2025-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143725923","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Artificial intelligence-enabled predictive energy saving planning of liquid cooling system for data centers 数据中心液冷系统的人工智能预测节能规划

IF 8 1区工程技术

Advanced Engineering Informatics Pub Date : 2025-03-29 DOI: 10.1016/j.aei.2025.103283

Shuaiyin Ma , Yuyang Liu , Yang Liu , Jiaqiang Wang , Qiu Fang , Yuanfeng Huang

{"title":"Artificial intelligence-enabled predictive energy saving planning of liquid cooling system for data centers","authors":"Shuaiyin Ma , Yuyang Liu , Yang Liu , Jiaqiang Wang , Qiu Fang , Yuanfeng Huang","doi":"10.1016/j.aei.2025.103283","DOIUrl":"10.1016/j.aei.2025.103283","url":null,"abstract":"<div><div>As significant sources of energy consumption and carbon emissions, data centers have become a focal point for improving energy efficiency worldwide. To address the challenges of high computational resource demands and limited adaptability of traditional prediction models to complex conditions, this paper proposes an artificial intelligence-enabled predictive energy saving planning based on the Transformer-GRU model for predicting coolant temperature in the liquid cooling system of data centers. By integrating the self-attention mechanism of the Transformer and the time-series prediction strengths of GRU, the model performs correlation analysis and feature extraction of key parameters to achieve high-precision predictions of coolant return temperature. Experimental results demonstrate the model’s superior accuracy compared to traditional prediction models, achieving an MSE of 1.349, RMSE of 1.157, MAPE of 0.0244, and R<sup>2</sup> of 81.07 %, significantly outperforming baseline models such as Transformer-LSTM (MSE = 1.355), Informer (MSE = 1.356), Reformer (MSE = 1.353), DeepAR (MSE = 1.385), LSTM (MSE = 1.351), GRU (MSE = 1.366), and CNN-GRU (MSE = 1.363). The model maintains high predictive accuracy under fluctuating environments and complex cooling conditions, effectively reducing the operational energy consumption of the liquid cooling system. This advancement not only enhances cooling efficiency but also drives data centers toward greater intelligence and sustainability. By leveraging real-time monitoring data and predictive control, the model dynamically optimizes cooling strategies, reducing coolant and energy usage while promoting sustainable resource utilization. Additionally, this study offers implementation insights for high-performance computing environments, laying the groundwork for future research on extending model capabilities and integrating multimodal data.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"65 ","pages":"Article 103283"},"PeriodicalIF":8.0,"publicationDate":"2025-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143735143","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

High-Accuracy prediction and efficient adjustment of surface shape distortion in optical elements: Model correction based on uncertainty quantification-driven transfer learning 光学元件表面形状畸变的高精度预测和有效调整：基于不确定性量化驱动迁移学习的模型校正

IF 8 1区工程技术

Advanced Engineering Informatics Pub Date : 2025-03-29 DOI: 10.1016/j.aei.2025.103281

Zhihao Fan , Xiaokai Mu , Rongxuan Zhao , Kangcheng Yin , Qingchao Sun , Wei Sun , Kaike Yang , Wenjing Ma

{"title":"High-Accuracy prediction and efficient adjustment of surface shape distortion in optical elements: Model correction based on uncertainty quantification-driven transfer learning","authors":"Zhihao Fan , Xiaokai Mu , Rongxuan Zhao , Kangcheng Yin , Qingchao Sun , Wei Sun , Kaike Yang , Wenjing Ma","doi":"10.1016/j.aei.2025.103281","DOIUrl":"10.1016/j.aei.2025.103281","url":null,"abstract":"<div><div>The surface shape of optical elements is a key determinant of opto-mechanical system performance, as its distortion directly affects optical wavefront distortion and can even render systems unusable. To address the challenge of compromised prediction accuracy due to high data dimensionality, this paper proposes a transfer learning model driven by uncertainty quantification, enabling high-accuracy prediction and efficient adjustment of optical element surface shape distortion using small-sample experimental data. First, a transfer learning prediction model for optical surface shape distortion is developed, integrating theoretical data from mechanical models with small-sample experimental data. Second, based on the principle of equal probability density, the uncertainty quantification of assembly preload is used to correct the transfer learning prediction model, enabling accurate prediction of high-dimensional surface reconstruction data and determining assembly process parameters in conjunction with surface shape distortion patterns. Third, an inverse model for preload in the assembly process is established, and an adjustment theory based on confidence levels is proposed, leading to a digital adjustment strategy aimed at minimizing beam wavefront distortion. Finally, experimental validation is conducted to verify the effectiveness of the adjustment strategy. Experimental validation shows that after three adjustments, surface distortion decreased by 13.66%, with only a 0.60% deviation from the theoretical minimum. Compared to traditional methods, this approach improves adjustment efficiency by 483.76%, offering a precise and efficient solution for optical surface shape distortion adjustment in opto-mechanical systems.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"65 ","pages":"Article 103281"},"PeriodicalIF":8.0,"publicationDate":"2025-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143725925","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Digital thread in engineering: Concept, state of art, and enabling framework 工程中的数字线程：概念、技术状态和使能框架

IF 8 1区工程技术

Advanced Engineering Informatics Pub Date : 2025-03-28 DOI: 10.1016/j.aei.2025.103258

Shouxuan Wu , Guoxin Wang , Jinzhi Lu , Yan Yan , Yihui Gong , Mengru Dong , Dimitris Kiritsis

{"title":"Digital thread in engineering: Concept, state of art, and enabling framework","authors":"Shouxuan Wu , Guoxin Wang , Jinzhi Lu , Yan Yan , Yihui Gong , Mengru Dong , Dimitris Kiritsis","doi":"10.1016/j.aei.2025.103258","DOIUrl":"10.1016/j.aei.2025.103258","url":null,"abstract":"<div><div>The digital thread (DT), a new paradigm for the development of complex engineering systems, has garnered widespread attention in recent years. The DT serves as a key driver and game-changer for the digital transformation of engineering systems due to its advantages in promoting the collaboration of stakeholders to aid decision-making. This study provides a systematic and quantitative survey of publications about the DT. By using bibliometric analysis tools, current progress on the DT in the engineering domain is visually identified. We analyzed the distribution of research teams, professional publishing sources, current research topics, and future research trends, revealing that current research of DT emphasizes the features of lifecycle, integration, interoperability, and decision-making inspired by the data-information-knowledge-wisdom (DIKW) model. Through a systematic review of relevant publications from the perspective of these key features, we propose an enabling framework of the DT, including the concept, reference architecture, and “3I” enabling technologies (integration; interoperability; and intelligent decision technologies). Finally, we outline the current challenges faced by the DT in the engineering domain, including issues about integration challenges, professional workforce, and cost analysis. This study aims to facilitate a more comprehensive understanding of the current states and research directions of the DT in the engineering domain.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"65 ","pages":"Article 103258"},"PeriodicalIF":8.0,"publicationDate":"2025-03-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143726017","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Large language model for patent concept generation 用于专利概念生成的大型语言模型

IF 8 1区工程技术

Advanced Engineering Informatics Pub Date : 2025-03-28 DOI: 10.1016/j.aei.2025.103301

Runtao Ren , Jian Ma , Jianxi Luo

{"title":"Large language model for patent concept generation","authors":"Runtao Ren , Jian Ma , Jianxi Luo","doi":"10.1016/j.aei.2025.103301","DOIUrl":"10.1016/j.aei.2025.103301","url":null,"abstract":"<div><div>In traditional innovation practices, concept and IP generation are often iteratively integrated. Both processes demand an intricate understanding of advanced technical domain knowledge. Existing large language models (LLMs), while possessing massive pre-trained knowledge, often fall short in the innovative concept generation due to a lack of specialized knowledge necessary for the generation. To bridge this critical gap, we propose a novel knowledge finetuning (KFT) framework to endow LLM-based AI with the ability to autonomously mine, understand, and apply domain-specific knowledge and concepts for invention generation, i.e., concept and patent generation together. Our proposed PatentGPT integrates knowledge injection pre-training (KPT), domain-specific supervised finetuning (SFT), and reinforcement learning from human feedback (RLHF). Extensive evaluation shows that PatentGPT significantly outperforms the state-of-the-art models on patent-related benchmark tests. Our method not only provides new insights into data-driven innovation but also paves a new path to fine-tune LLMs for applications in the context of technology. We also discuss the managerial and policy implications of AI-generating inventions in the future.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"65 ","pages":"Article 103301"},"PeriodicalIF":8.0,"publicationDate":"2025-03-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143725922","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Towards trustworthy civil aviation hazards identification: An uncertainty-aware deep learning framework 面向可信赖的民航危险识别：一个不确定性感知深度学习框架

IF 8 1区工程技术

Advanced Engineering Informatics Pub Date : 2025-03-27 DOI: 10.1016/j.aei.2025.103280

Zhaoguo Hou , Huawei Wang , Minglan Xiong , Changwei Zhou , Yubin Yue

{"title":"Towards trustworthy civil aviation hazards identification: An uncertainty-aware deep learning framework","authors":"Zhaoguo Hou , Huawei Wang , Minglan Xiong , Changwei Zhou , Yubin Yue","doi":"10.1016/j.aei.2025.103280","DOIUrl":"10.1016/j.aei.2025.103280","url":null,"abstract":"<div><div>Accurate and trustworthy hazards identification is crucial for preventing accidents and ensuring flight safety. However, deep learning-based identification methods are limited by their black-box characteristics to provide trustworthy and interpretable results. Existing research on interpretable civil aviation hazard identification focuses on developing interpretable modules to be embedded in deep learning models to give engineering meaning to the results; or inferring the logic of the model’s decision-making based on the results. However, there is limited research on how to quantify and explain the uncertainty in the results. Quantifying and decomposing uncertainty not only provides confidence of results but also helps to identify the sources of unknown factors in the data, thereby providing guidance for improving model interpretability. Therefore, this paper proposes an uncertainty-aware deep learning framework for trustworthy civil aviation hazards identification. Firstly, a Bayesian multi-scale attention convolutional neural network with an integrated Monte Carlo dropout mechanism was designed, which can estimate the uncertainty of model predictions through internal randomness, thereby endowing the network with the uncertainty-aware ability. Secondly, a set of uncertainty quantification and decomposition schemes was established, which can achieve the confidence representation of the identification results and the separation of epistemic uncertainty and aleatoric uncertainty. Finally, an adjustable uncertainty decision threshold was constructed, which can be dynamically adjusted according to the risk level of application scenarios to achieve the optimal risk management. In out-of-distribution test scenarios with unknown hazards, comparisons with existing identification methods demonstrate that the proposed framework has superior uncertainty-aware capabilities and potential for engineering application.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"65 ","pages":"Article 103280"},"PeriodicalIF":8.0,"publicationDate":"2025-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143704918","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0