检索结果-内蒙古大学图书馆

International Conference on Industrial Artificial Intelligence (IAI)

作者： Jun Lu Min Feng Jinliang Ding Xiaoling Zhang State Key Laboratory of Synthetical Automation for Process Industries Northeastern University Shenyang China Key Laboratory of Intelligent Control and Optimization for Industrial Equipment Ministry of Education Dalian University of Technology Dalian China

The refineries usually process the blending crude oil at present. Blending crude oil leads to the increased uncertainties of physical properties of crude oil. This paper presents a robust stochastic configuration network (SCN) based prediction model for total nitrogen content in crude oil. First, the robust SCN based on Bayesian framework is proposed. In the Bayesian framework, the noise and outliers are assumed to be the mixture of two Gaussian distributions to improve the robustness of SCN. Then, to solve the problem that there are no analytical solutions of parameters due to the assumption of mixture of Gaussian distributions, parameters of the proposed robust SCN are estimated by the expectation maximization (EM) algorithm. Experiments using the real-world data collected from a refinery are carried out. The results indicate that the proposed approach has better accuracy and robustness.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A branch-price-and-cut algorithm for the time-dependent multiple truck–drone routing problem

引用

Computers and Operations Research 2025年 182卷

作者： Zang, Xiaoning Jiang, Li Liang, Changyong Fang, Xiang School of Management Hefei University of Technology Hefei China Key Laboratory of Process Optimization and Intelligent Decision-making Ministry of Education China Business School University of Colorado Denver DenverCO United States

Time-varying traffic conditions are crucial features of urban logistics. Overlooking these conditions will pose a high coordination risk for drone-assisted routing problems. In this paper, a time-dependent multiple truck–drone routing problem (TD-MTDRP), which captures the time-varying traffic conditions as time-dependent travel times, is introduced. In the problem, trucks with time-dependent travel times travel through the urban roads, while drones fly over the urban areas, unaffected by the urban traffic, resulting in time-independent travel times. Our objective is to the total duration of all routes. We present the ready time function on a truck–drone path level, which is proven to satisfy the FIFO property. A route-based model is introduced based on this function to formulate the problem. The model decomposes the set-partitioning problem into a master problem (MP) and a time-dependent shortest truck–drone path problem with resource constraints (TDSTDPPRC). A branch-price-and-cut algorithm is also developed to solve the model. In this algorithm, a column-and-row generation is introduced to solve the MP, and a labeling algorithm incorporating new label extension and dominance rules is presented to address the TDSTDPPRC. For the numerical results, the approaches are evaluated on two sets of instances derived from the benchmark instances, demonstrating the effectiveness of the exact method in solving the TD-MTDRP. © 2025

关键词： Travel time

来源：评论

学校读者我要写书评

暂无评论

Earth Observation Satellite Scheduling Problem for Multitemporal Revisit Tasks: A Variable Neighborhood Search Algorithm

SSRN

引用

SSRN 2024年

作者： Xing, Ligang Hu, Xiaoxuan Zhu, Waiming Zhu, Xutong Xia, Wei School of Management Hefei University of Technology Hefei230009 China Key Laboratory of Process Optimization and Intelligent Decision-making Ministry of Education Hefei230009 China Engineering Research Center for Intelligent Management of Aerospace System Hefei230009 China

After natural disasters such as floods and earthquakes, Earth observation satellites (EOSs) often need to revisit affected areas multiple times to acquire multitemporal images. Such observation tasks are referred to as multitemporal revisit tasks. In this paper, we study the Earth observation satellite scheduling problem for multitemporal revisit tasks, which is a combinatorial optimization problem in continuous space. This problem exhibits more complex constraints and is more challenging to solve than the traditional Earth observation satellite scheduling problem. Firstly, we formally define the problem and formulate it with a mixed integer nonlinear programming model. Secondly, we develop a variable neighborhood search algorithm to search for the optimal solution. This algorithm embeds a dynamic greedy heuristic, which can efficiently generate a schedule for satellites. Thirdly, computational experiments demonstrate the efficiency and stability of the algorithm. Moreover, the CPU time of the algorithm increases linearly with both task count and revisit number, indicating good scalability. © 2024, The Authors. All rights reserved.

关键词： Integer programming

来源：评论

学校读者我要写书评

暂无评论

Learning Adaptive Mapless Navigation Skills Through Evidential Deep Learning

Learning Adaptive Mapless Navigation Skills Through Evidenti...

引用

IEEE International Conference on Real-time Computing and Robotics (RCAR)

作者： Cancan Zhao Yufeng Yan Xuhao Liu Shuai Zhang Bo Ouyang School of Management Hefei University of Technology Hefei China The Key Laboratory of Process Optimization and Intelligent Decision-Making(Ministry of Education) Hefei University of Technology Hefei China

Traditional navigation methods tend to design and train navigation skills based on the consideration of multi-scenario applicability and perform well in general scenarios. However, this results in sub-optimal navigation performance in some context-specific scenarios, such as navigating through open or crowded scenarios. While some existing navigation skills can achieve satisfactory performance in context-specific scenarios, e.g., using PID in open scenarios and reinforcement learning-based training methods in crowded scenarios. Most methods currently do without consideration of the integration of navigation skills and contextual understanding. To address this gap, we propose an adaptive mapless navigation method to utilize the optimal capabilities of each skill corresponding to the context-specific scenario. Specifically, we package several existing navigation skills and train a scenario classifier to enable the agent to automatically select the most appropriate skill based on the current scenario. Furthermore, we incorporate evidential deep learning (EDL) to assess classification uncertainty and ensure the safe switching of navigation skills. Our method was tested in both simulated and real scenarios, and the results of the experiments confirmed the effectiveness and generalization of our method. Moreover, the proposed method exhibits notably enhanced, context-sensitive efficiency within a hybrid multi-scenario setting, in comparison to a single navigation skill.

关键词：

来源：评论

学校读者我要写书评

暂无评论

3DGS-based Tracking of Articulated Instruments for Instruments-Tissue Interaction Estimation in Robot-assisted Minimally Invasive Surgery

3DGS-based Tracking of Articulated Instruments for Instrumen...

引用

IEEE International Conference on Robotics and Biomimetics

作者： Xiaojian Li Lili Shen Ling Li Hangjie Mo Xin Wang Xiang Ma Jieyu Zheng School of Management Hefei University of Technology Hefei China Key Laboratory of Process Optimization and Intelligent Decision-Making Hefei University of Technology Hefei China West China Hospital Sichuan University Sichuan China

ISBN: (数字)9798331509644

ISBN: (纸本)9798331509651

3D information of surgical scene potentially enhances the accuracy of surgical navigation and facilitate the automation of robot-assisted surgery. However, current research seldom considers the relative 3D position between instruments and tissue. In this paper, the instruments-tissue interaction estimation method is proposed to obtain the relative distance between instruments and tissue. The 3D Gaussian representation of instruments is established preoperatively, making the pose of instruments the only parameter to be learned and updated. Based on this, the instrument pose estimation method is proposed, which applies 3DGS (3D Gaussian Splatting) to the instrument Gaussian model to obtain rendered color images, projection masks, and depth maps. These are aligned with the input images, segmented masks and depth from stereo depth estimation to construct the target loss. By backpropagating through the target loss, the optimal instrument pose is obtained. The L1 loss and smooth loss based on TPS (the Thin Plate Spline) are used to complete the depth of tissue occluded by instruments. The instruments-tissue interaction is estimated by the nearest distance between instrument and tissue. The experimental results indicate an average error of 0.94 mm and an accuracy of 90.40% with an error less than 2mm in estimating relative 3D distance between instruments and tissue.

关键词： Solid modeling Three-dimensional displays Accuracy Minimally invasive surgery Navigation Instruments Pose estimation Real-time systems Splines (mathematics) Robots

来源：评论

学校读者我要写书评

暂无评论

Multi-sensor Deep Clustering Networks with Joint Learning Strategy for Fault Diagnosis Considering Consensus-specific Constraint

引用

IEEE Sensors Journal 2025年第11期25卷 19921-19936页

作者： Xu, Yihao Lu, Mingfeng Wang, Gang Fang, Fang University of Science and Technology of China School of Management Hefei230026 China Hefei University of Technology School of Management Hefei230009 China Hefei University of Technology School of Management Key Laboratory of Process Optimization and Intelligent Decision-Making of the Ministry of Education Ministry of Education Engineering Research Center for Intelligent Decision-Making and Information System Technologies Hefei230009 China

The deep clustering method has been successfully employed for fault diagnosis due to its remarkable ability to extract deep representation features, particularly in scenarios where labeled data is unavailable. However, the separate learning strategy utilized in deep clustering-based fault diagnosis methods leads to sub-optimal fault diagnosis results, and the consensus representation feature present in multi-sensor industrial data is not fully explored to improve the effectiveness of fault diagnosis. To address the above problems, a novel deep clustering method named Multi-sensor Deep Clustering Networks with joint learning strategy (MsDCN) is proposed for fault diagnosis, which can comprehensively utilize the industrial multi-sensor data and enable joint optimization. First, the Commonality-net uses a sensor-aware attentive encoder (SaAE) to integrate multi-sensor industrial data and extract the consensus representation feature. Second, the Individuality-net is established to extract the specific representation features of multi-sensor industrial data. Finally, a joint multisensor deep clustering learning strategy including consensus-specific constraints is designed to optimize the networks jointly and learn the consensus representation feature with consistent and complementary information. Extensive experiments on the gearbox and bearing datasets validate the proposed MsDCN. The results indicate that the proposed MsDCN can generate a reliable consensus representation feature to achieve better fault diagnosis performance than other benchmark methods. Moreover, its capability to enhance multi-sensor fault diagnosis with unlabeled data highlights its potential for real-world applications. © 2001-2012 IEEE.

关键词： Labeled data

来源：评论

学校读者我要写书评

暂无评论

Pain Intensity Recognition from Masked Facial Expressions using Swin-Transformer

Pain Intensity Recognition from Masked Facial Expressions us...

引用

IEEE International Conference on Robotics and Biomimetics

作者： Xin Yuan Shuai Zhang Cancan Zhao Xian He Bo Ouyang Shanlin Yang Key Laboratory of Process Optimization and Intelligent Decision-Making(Ministry of Education) School of Management Hefei University of Technology Hefei China

ISBN: (纸本)9781665481106

Previous studies recognize pain expressions based on the entire face, for example, Prkachin and Solomon Pain intensity (PSPI). However, the patients face is often masked by instruments in an intensive care unit (ICU), such as respirator, and gauzes, just name a few, which causes the agent cannot measure the pain intensity using PSPI directly. To tackle this problem, we explore the pain intensity recognition from masked face. First, we conducted four levels of pain measurement experiments with four types of masked face using Swin-Transformer. Experiment results show that the accuracy of pain measurement is more than 90%, even masking a large part of facial Action Units (AU) related to the pain on the UNBC-McMaster dataset. Furthermore, we pre-train the pain recognition model with masked face dataset, such that the model can capture facial features contributed to the pain. Results on binary and four-level pain intensity measurement tasks show our model outperforms recent state-of-the-art performance, achieving 97.38% and 95.25% accuracy, respectively.

关键词： Gold Pain Face recognition Instruments Ventilation Robustness Physiology

来源：评论

学校读者我要写书评

暂无评论

Multi-Objective optimization of Cloud Manufacturing Service Combination Based on Manufacturing and Logistics Coordinate

Multi-Objective Optimization of Cloud Manufacturing Service ...

引用

Chinese Control and Decision Conference, CCDC

作者： Sining Han Bingfeng Ge Wanying Wei Xiaonong Lu Zeqiang Hou Yuming Huang College of Systems Engineering National University of Defense Technology National University of Defense Technology Changsha China Key Laboratory of Process Optimization and Intelligent Decision-Making Ministry of Education University of Science and Technology of China Hefei China

ISBN: (数字)9798350387780

ISBN: (纸本)9798350387797

With the continuous emergence of new technologies, the manufacturing industry is facing unprecedented challenges. Manufacturing services typically involve various types of production equipment, diverse suppliers, and varied customer demands. optimization in this context often encompasses multi pie objectives, necessitating timely adjustments in service comb ¡nations to meet customer needs. Traditional manufacturing models evidently fall short in resource allocation within manufacturing supply chains. Cloud manufacturing, proposed as a novel manufacturing model, aims to optimize resource allocation through the combination and configuration of resources, there by enhancing the efficiency of manufacturing supply chains. Logistics, as a crucial component within manufacturing supply chains, underscores the significance of coordinating the optimizad on of manufacturing and logistics services. This research focus es on the multi-objective optimization problem of cloud manufacturing service combinations. Building upon the execution process of manufacturing and logistics services for complex product manufacturing tasks, a novel mathematical model for the optimization of manufacturing resources and logistics resources i s constructed. The model utilizes the NSGA-H algorithm to coordinate the optimization processes of manufacturing and legistics resources. Finally, a case study demonstrates the effectiveness of the proposed model.

关键词： Costs Supply chains Production equipment Pareto optimization Mathematical models Manufacturing Resource management

来源：评论

学校读者我要写书评

暂无评论

A novel centralized cross-efficiency evaluation via explainable artificial intelligence in the context of big data

引用

European Journal of Operational Research 2025年

作者： Yang, Min Wang, Zixuan Liang, Liang Management School Hefei University of Technology Hefei230009 China Key Laboratory of Process Optimization and Intelligent Decision-Making Ministry of Education Hefei230009 China Intelligent Interconnection System Anhui Provincial Laboratory Hefei230009 China NSFC Fundamental Research Center on Smart Interconnected System Engineering Hefei230009 China

Cross-efficiency evaluation in data envelopment analysis (DEA) assumes that decision making units (DMUs) have full flexibility in choosing weights according to their individual preferences. However, this total autonomy may be inapplicable in some centralized organizational scenarios. To address this problem, this paper introduces a novel centralized cross-efficiency evaluation which considers both individual and organizational preferences with assistance of explainable artificial intelligence (XAI) in the context of big data. Specifically, XAI is first applied to approach the organizational efficiency function and then calculate the marginal contribution of each variable as the variable importance, which represents the organizational preference. Furthermore, we propose a centralized secondary goal model to select the unique optimal weight profile from the candidate weights that remain self-efficiency as Pareto-optimal, such that the deviation between individual and organizational preferences is minimized. In addition, a centralization factor is introduced to ensure that the model's centralization degree corresponds to the actual centralization level in organizational management. Finally, the proposed method is applied to evaluate the efficiency of DMUs within three different centralized organizations sequentially. The results verify that the proposed method yields more discriminative and robust efficiency scores within organizations compared to several previous cross-efficiency evaluation methods. © 2025 Elsevier B.V.

关键词： Risk perception

来源：评论

学校读者我要写书评

暂无评论

Hierarchically-Trusted Evidential Fusion Method with Consistency Learning for Multimodal Language Understanding

SSRN

引用

SSRN 2024年

作者： Yang, Ying Yan-Qiu, Yang Ren, Gang School of Management Hefei University of Technology Hefei230009 China Key Laboratory of Process Optimization and Intelligent Decision-making Ministry of Education Hefei230009 China Ministry of Education Engineering Research Centre for Intelligent Decision-Making & Information System Technologies Hefei230009 China

Human communication in the real world is based on multimodal data. Multimodal language understanding tasks such as multimodal intention recognition and multimodal emotion recognition are key steps in many applications, such as dialogue systems and social media. Although many studies have explored multimodal fusion methods and applied these methods to multimodal language understanding tasks including intent recognition and sentiment analysis, it is still a knotty problem whether the classification results are trusted. Therefore, this paper proposes a hierarchically-trusted evidential fusion method (HTEF) for multimodal language understanding. We construct multiple views from multi-modal data in a hierarchical manner, and extract multiple pieces of evidence from them. Then, we introduce Dempster-Shafer evidence theory(DST) to fuse evidence obtained from multimodal data to ensure the reduction of overall uncertainty. In addition, we explore the feature-level consistency and decision-level consistency of multi-modal data to avoid the conflict across information captured from different modalities. The method is evaluated on the public multimodal intent recognition dataset MIntRec and the emotion recognition dataset MELD. The experimental results show that our model outperforms both existing single-modal and multimodal baseline models across multiple evaluation metrics, which verifies the effectiveness of the proposed method. © 2024, The Authors. All rights reserved.

关键词： Emotion Recognition

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：