Engineering Applications of Artificial Intelligence最新文献_第5页

Multi-scale information sharing and selection network with boundary attention for polyp segmentation 用于息肉分割的多尺度信息共享与选择网络（带边界关注

IF 7.5 2区计算机科学 Q1 AUTOMATION & CONTROL SYSTEMS

Engineering Applications of Artificial Intelligence

Pub Date : 2024-11-11 DOI: 10.1016/j.engappai.2024.109467

Xiaolu Kang, Zhuoqi Ma, Kang Liu, Yunan Li, Qiguang Miao

Polyp segmentation in colonoscopy images is essential in clinical practice, offering valuable information for the diagnosis of colorectal cancer and subsequent surgical procedures. Despite the relatively good performance of existing methods, polyp segmentation still faces the following challenges: (1) Varying lighting conditions in colonoscopy and differences in polyp locations, sizes, and morphologies. (2) The indistinct boundary between polyps and surrounding tissue. To tackle these challenges, we propose a Multi-scale Information Sharing and Selection Network (MISNet) for the polyp segmentation task. We have designed a Selectively Shared Fusion Module (SSFM) to facilitate information sharing and the active selection between low-level and high-level features, thus enhancing the model’s ability to capture comprehensive information. Subsequently, we have developed a Parallel Attention Module (PAM) to improve the model’s attention on boundaries, and a Balancing Weight Module (BWM) to support the continuous refinement of boundary segmentation through the bottom-up process. Extensive experiments on five benchmark datasets show competitive results compared to existing representative methods. Specifically, our method has reached the mean Dice coefficient of 0.903 and 0.918 on the Kvasir and CVC-ClinicDB datasets, 0.762 and 0.764 on the challenging CVC-ColonDB and ETIS datasets. These innovative modules in our proposed MISNet effectively address key challenges, providing a robust solution for accurate polyp segmentation in clinical diagnosis and treatment. The proposed model is available at https://github.com/q1216355254/MISNet.git.

结肠镜图像中的息肉分割在临床实践中至关重要，它为结肠直肠癌的诊断和后续手术提供了宝贵的信息。尽管现有方法的性能相对较好，但息肉分割仍面临以下挑战：(1) 结肠镜检查中光线条件的变化以及息肉位置、大小和形态的差异。(2）息肉与周围组织的边界不清晰。为了应对这些挑战，我们针对息肉分割任务提出了多尺度信息共享和选择网络（MISNet）。我们设计了一个选择性共享融合模块（SSFM），以促进信息共享以及低层次特征和高层次特征之间的主动选择，从而提高模型捕捉综合信息的能力。随后，我们又开发了并行关注模块（PAM）和平衡权重模块（BWM），前者用于提高模型对边界的关注度，后者则支持通过自下而上的过程不断完善边界分割。在五个基准数据集上进行的广泛实验表明，与现有的代表性方法相比，我们的方法具有很强的竞争力。具体来说，我们的方法在 Kvasir 和 CVC-ClinicDB 数据集上的平均骰子系数分别达到 0.903 和 0.918，在具有挑战性的 CVC-ColonDB 和 ETIS 数据集上的平均骰子系数分别达到 0.762 和 0.764。我们提出的 MISNet 中的这些创新模块有效地解决了关键难题，为临床诊断和治疗中的精确息肉分割提供了强大的解决方案。拟议的模型可在 https://github.com/q1216355254/MISNet.git 上查阅。

{"title":"Multi-scale information sharing and selection network with boundary attention for polyp segmentation","authors":"Xiaolu Kang, Zhuoqi Ma, Kang Liu, Yunan Li, Qiguang Miao","doi":"10.1016/j.engappai.2024.109467","DOIUrl":"10.1016/j.engappai.2024.109467","url":null,"abstract":"<div><div>Polyp segmentation in colonoscopy images is essential in clinical practice, offering valuable information for the diagnosis of colorectal cancer and subsequent surgical procedures. Despite the relatively good performance of existing methods, polyp segmentation still faces the following challenges: (1) Varying lighting conditions in colonoscopy and differences in polyp locations, sizes, and morphologies. (2) The indistinct boundary between polyps and surrounding tissue. To tackle these challenges, we propose a Multi-scale Information Sharing and Selection Network (MISNet) for the polyp segmentation task. We have designed a Selectively Shared Fusion Module (SSFM) to facilitate information sharing and the active selection between low-level and high-level features, thus enhancing the model’s ability to capture comprehensive information. Subsequently, we have developed a Parallel Attention Module (PAM) to improve the model’s attention on boundaries, and a Balancing Weight Module (BWM) to support the continuous refinement of boundary segmentation through the bottom-up process. Extensive experiments on five benchmark datasets show competitive results compared to existing representative methods. Specifically, our method has reached the mean Dice coefficient of 0.903 and 0.918 on the Kvasir and CVC-ClinicDB datasets, 0.762 and 0.764 on the challenging CVC-ColonDB and ETIS datasets. These innovative modules in our proposed MISNet effectively address key challenges, providing a robust solution for accurate polyp segmentation in clinical diagnosis and treatment. The proposed model is available at <span><span>https://github.com/q1216355254/MISNet.git</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50523,"journal":{"name":"Engineering Applications of Artificial Intelligence","volume":"139 ","pages":"Article 109467"},"PeriodicalIF":7.5,"publicationDate":"2024-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142659087","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Machine learning based state observer for discrete time systems evolving on Lie groups 基于机器学习的离散时间系统状态观测器在李群上演化

IF 7.5 2区计算机科学 Q1 AUTOMATION & CONTROL SYSTEMS

Engineering Applications of Artificial Intelligence

Pub Date : 2024-11-11 DOI: 10.1016/j.engappai.2024.109576

Soham Shanbhag, Dong Eui Chang

In this paper, a machine learning based observer for systems evolving on manifolds is designed such that the state of the observer is restricted to the Lie group on which the system evolves. Designing machine learning based observers for systems evolving on Lie groups using charts would require training a machine learning based observer for each chart of the Lie group, and switching between the trained models based on the state of the system. We propose a novel deep learning based technique whose predictions are restricted to certain measure 0 subsets of the Euclidean space without using charts. Using this network, we design an observer ensuring that the state of the observer is restricted to the Lie group, and predicting the state using only one trained algorithm. The deep learning network predicts an error term on the Lie algebra of the Lie group, uses the map from the Lie algebra to the group, the group operation, and the present state to estimate the state at the next epoch. This approach, being purely data driven, does not require a model of the system. The proposed algorithm provides a novel framework for constraining the output of machine learning networks to certain measure 0 subsets of a Euclidean space without training on each specific chart and without requiring switching. We show the validity of this method using Monte Carlo simulations performed of the rigid body rotation and translation system.

本文为流形上演化的系统设计了基于机器学习的观测器，观测器的状态仅限于系统演化所在的李群。为使用图表在李群上演化的系统设计基于机器学习的观测器，需要为李群的每个图表训练一个基于机器学习的观测器，并根据系统的状态在训练好的模型之间切换。我们提出了一种新颖的基于深度学习的技术，其预测仅限于欧几里得空间的某些度量为 0 的子集，而无需使用图表。利用这一网络，我们设计了一个观测器，确保观测器的状态仅限于 Lie 组，并只使用一种训练有素的算法来预测状态。深度学习网络会对李群的李代数预测一个误差项，利用从李代数到李群的映射、李群运算和当前状态来估计下一个纪元的状态。这种方法纯粹由数据驱动，不需要系统模型。所提出的算法提供了一个新颖的框架，可将机器学习网络的输出限制在欧几里得空间的某些度量为 0 的子集上，而无需在每个特定图表上进行训练，也无需切换。我们通过对刚体旋转和平移系统进行蒙特卡罗模拟，证明了这种方法的有效性。

{"title":"Machine learning based state observer for discrete time systems evolving on Lie groups","authors":"Soham Shanbhag, Dong Eui Chang","doi":"10.1016/j.engappai.2024.109576","DOIUrl":"10.1016/j.engappai.2024.109576","url":null,"abstract":"<div><div>In this paper, a machine learning based observer for systems evolving on manifolds is designed such that the state of the observer is restricted to the Lie group on which the system evolves. Designing machine learning based observers for systems evolving on Lie groups using charts would require training a machine learning based observer for each chart of the Lie group, and switching between the trained models based on the state of the system. We propose a novel deep learning based technique whose predictions are restricted to certain measure 0 subsets of the Euclidean space without using charts. Using this network, we design an observer ensuring that the state of the observer is restricted to the Lie group, and predicting the state using only one trained algorithm. The deep learning network predicts an error term on the Lie algebra of the Lie group, uses the map from the Lie algebra to the group, the group operation, and the present state to estimate the state at the next epoch. This approach, being purely data driven, does not require a model of the system. The proposed algorithm provides a novel framework for constraining the output of machine learning networks to certain measure 0 subsets of a Euclidean space without training on each specific chart and without requiring switching. We show the validity of this method using Monte Carlo simulations performed of the rigid body rotation and translation system.</div></div>","PeriodicalId":50523,"journal":{"name":"Engineering Applications of Artificial Intelligence","volume":"139 ","pages":"Article 109576"},"PeriodicalIF":7.5,"publicationDate":"2024-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142659274","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An image-to-image adversarial network to generate high resolution wind data over complex terrains from weather predictions 根据天气预测生成复杂地形上高分辨率风力数据的图像到图像对抗网络

IF 7.5 2区计算机科学 Q1 AUTOMATION & CONTROL SYSTEMS

Engineering Applications of Artificial Intelligence

Pub Date : 2024-11-11 DOI: 10.1016/j.engappai.2024.109533

Jaime Milla-Val , Carlos Montañés , Norberto Fueyo

In this work, we propose a Machine Learning method to predict detailed wind fields over extensive, complex terrains. The ability to predict local wind fields is becoming increasingly important for a range of applications, including sports in Nature, large outdoors events, light-aircraft flying, or the management of natural disasters. The intricate nature of wind dynamics, particularly in regions with complex orography such as a mountain range, presents a major challenge to traditional forecasting models. This work presents an efficient way to predict local wind conditions with a high resolution, similar to that of Computational Fluid Dynamics (CFD), in large geographical areas with complex terrain, using the results from relatively coarse (and therefore economical) data from Numerical Weather Prediction (NWP). To achieve this goal, we developed a conditional Generative Adversarial Neural network model (cGAN) to convert NWP data into CFD-like simulations. We apply the method to a rugged region in the Pyrenees mountain range in Spain. The results show that the proposed model outperforms traditional Machine Learning methods, such as Support Vector Machines (SVM), in terms of accuracy and computational efficiency. The method is four orders of magnitude faster than traditional CFD. Mean Average Errors of

1.36 m/s

for wind speed and 18.73°for wind direction are obtained with the proposed approach.

在这项工作中，我们提出了一种机器学习方法，用于预测广阔复杂地形上的详细风场。预测当地风场的能力在一系列应用中正变得越来越重要，包括自然界中的体育运动、大型户外活动、轻型飞机飞行或自然灾害管理。风的动态性质错综复杂，尤其是在山脉等地形复杂的地区，这对传统的预测模型提出了重大挑战。这项研究提出了一种有效的方法，利用数值天气预报（NWP）中相对粗糙（因此经济）的数据结果，以类似于计算流体动力学（CFD）的高分辨率预测地形复杂的大面积地理区域的局部风况。为实现这一目标，我们开发了一种条件生成对抗神经网络模型（cGAN），用于将 NWP 数据转换为类似 CFD 的模拟结果。我们将该方法应用于西班牙比利牛斯山脉的一个崎岖地区。结果表明，所提出的模型在准确性和计算效率方面都优于传统的机器学习方法，如支持向量机（SVM）。该方法比传统的 CFD 快四个数量级。采用所提方法得出的风速平均误差为 1.36 米/秒，风向平均误差为 18.73°。

{"title":"An image-to-image adversarial network to generate high resolution wind data over complex terrains from weather predictions","authors":"Jaime Milla-Val , Carlos Montañés , Norberto Fueyo","doi":"10.1016/j.engappai.2024.109533","DOIUrl":"10.1016/j.engappai.2024.109533","url":null,"abstract":"<div><div>In this work, we propose a Machine Learning method to predict detailed wind fields over extensive, complex terrains. The ability to predict local wind fields is becoming increasingly important for a range of applications, including sports in Nature, large outdoors events, light-aircraft flying, or the management of natural disasters. The intricate nature of wind dynamics, particularly in regions with complex orography such as a mountain range, presents a major challenge to traditional forecasting models. This work presents an efficient way to predict local wind conditions with a high resolution, similar to that of Computational Fluid Dynamics (CFD), in large geographical areas with complex terrain, using the results from relatively coarse (and therefore economical) data from Numerical Weather Prediction (NWP). To achieve this goal, we developed a conditional Generative Adversarial Neural network model (cGAN) to convert NWP data into CFD-like simulations. We apply the method to a rugged region in the Pyrenees mountain range in Spain. The results show that the proposed model outperforms traditional Machine Learning methods, such as Support Vector Machines (SVM), in terms of accuracy and computational efficiency. The method is four orders of magnitude <em>faster than</em> traditional CFD. <em>Mean Average Errors of</em> <span><math><mrow><mn>1</mn><mo>.</mo><mn>36</mn><mspace></mspace><mi>m/s</mi></mrow></math></span> <em>for wind speed and 18.73°for wind direction are obtained with the proposed approach.</em></div></div>","PeriodicalId":50523,"journal":{"name":"Engineering Applications of Artificial Intelligence","volume":"139 ","pages":"Article 109533"},"PeriodicalIF":7.5,"publicationDate":"2024-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142659272","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Federated Reinforcement Learning for smart and privacy-preserving energy management of residential microgrids clusters 针对住宅微电网集群的智能和隐私保护能源管理的联合强化学习

IF 7.5 2区计算机科学 Q1 AUTOMATION & CONTROL SYSTEMS

Engineering Applications of Artificial Intelligence

Pub Date : 2024-11-11 DOI: 10.1016/j.engappai.2024.109579

Mao Tan , Jie Zhao , Xiao Liu , Yongxin Su , Ling Wang , Rui Wang , Zhuocen Dai

Real-time energy management optimizes energy utilization and manages electrical loads, which is crucial for improving the operational efficiency of residential microgrids. However, existing management methods suffer from model complexity and slow training speed. To solve this problem, we introduce Federated Reinforcement Learning to manage residential microgrids by training a control strategy in a decentralized and privacy-preserving manner. Specifically, a residential microgrid energy optimization management model is first established based on the Proximal Policy Optimization (PPO) method. Then, we propose a cooperative training strategy for multiple Residential microgrids based on Federated Reinforcement Learning (RFRL). The proposed method improves the training speed of residential microgrid models by sharing parameter information, such as network weights, while protects users’ usage data. Finally, clustering analysis is introduced in the case of heterogeneous residential microgrid data. Extensive experimental evaluation shows that our method outperforms the alternative residential microgrid management methods in terms of cost efficiency.

实时能源管理可优化能源利用和管理电力负荷，对提高住宅微电网的运行效率至关重要。然而，现有的管理方法存在模型复杂、训练速度慢等问题。为了解决这个问题，我们引入了联邦强化学习，通过分散和保护隐私的方式训练控制策略来管理住宅微电网。具体来说，我们首先基于近端策略优化（PPO）方法建立了住宅微电网能源优化管理模型。然后，我们提出了一种基于联合强化学习（RFRL）的多住宅微电网合作训练策略。该方法通过共享网络权重等参数信息，提高了住宅微电网模型的训练速度，同时保护了用户的使用数据。最后，针对异构住宅微电网数据引入了聚类分析。广泛的实验评估表明，我们的方法在成本效率方面优于其他住宅微电网管理方法。

{"title":"Federated Reinforcement Learning for smart and privacy-preserving energy management of residential microgrids clusters","authors":"Mao Tan , Jie Zhao , Xiao Liu , Yongxin Su , Ling Wang , Rui Wang , Zhuocen Dai","doi":"10.1016/j.engappai.2024.109579","DOIUrl":"10.1016/j.engappai.2024.109579","url":null,"abstract":"<div><div>Real-time energy management optimizes energy utilization and manages electrical loads, which is crucial for improving the operational efficiency of residential microgrids. However, existing management methods suffer from model complexity and slow training speed. To solve this problem, we introduce Federated Reinforcement Learning to manage residential microgrids by training a control strategy in a decentralized and privacy-preserving manner. Specifically, a residential microgrid energy optimization management model is first established based on the Proximal Policy Optimization (PPO) method. Then, we propose a cooperative training strategy for multiple Residential microgrids based on Federated Reinforcement Learning (RFRL). The proposed method improves the training speed of residential microgrid models by sharing parameter information, such as network weights, while protects users’ usage data. Finally, clustering analysis is introduced in the case of heterogeneous residential microgrid data. Extensive experimental evaluation shows that our method outperforms the alternative residential microgrid management methods in terms of cost efficiency.</div></div>","PeriodicalId":50523,"journal":{"name":"Engineering Applications of Artificial Intelligence","volume":"139 ","pages":"Article 109579"},"PeriodicalIF":7.5,"publicationDate":"2024-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142659043","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Integration of in-wheel motor sensorless systems and hierarchical direct yaw moment control for distributed drive electric vehicles 集成轮内电机无传感器系统和分层直接偏航力矩控制，用于分布式驱动电动汽车

IF 7.5 2区计算机科学 Q1 AUTOMATION & CONTROL SYSTEMS

Engineering Applications of Artificial Intelligence

Pub Date : 2024-11-11 DOI: 10.1016/j.engappai.2024.109600

Xiaodong Wang , Maoping Ran , Xinglin Zhou

Ensuring robust and reliable control of distributed vehicles powered by in-wheel motor systems poses a significant challenge due to the harsh operating environments and high costs of such motor systems. Poor motor control, parameter variations, and sensor malfunction under these conditions can compromise the vehicle yaw stability. Integrating permanent magnet synchronous motor (PMSM) sensorless systems with vehicle yaw moment control offers a cost-effective solution for this issue without wheel angular speed sensors while enhancing yaw stability. In this paper, a composite nonlinear feedback sliding mode controller that can enhance the PMSM speed response is proposed. The proposed scheme exhibits a rotor speed overshoot and transient time of only 0.64% and 0.07s, respectively, which are smaller and shorter compared with other methods under motor parameter changes. Subsequently, the key states and tire-road friction coefficients required for vehicle control were estimated using sensorless rotor speeds and unscented Kalman filters, enabling the integration of the PMSM sensorless system with the vehicle yaw moment control. Additionally, a fuzzy adaptive hybrid sliding mode method is presented for yaw moment control enhancement. This method maintained the smallest sideslip angle root mean square error during double lane changes (0.4192 deg) compared with other methods. Analysis results show that different motor controllers and parameter changes significantly affect the vehicle dynamics performance. The proposed integrated scheme is feasible and effectively enhances the yaw moment control via high-performance sensorless PMSM systems.

由于轮内电机系统的运行环境恶劣且成本高昂，因此确保对由轮内电机系统驱动的分布式车辆进行稳健可靠的控制是一项重大挑战。在这些条件下，电机控制不良、参数变化和传感器故障都会影响车辆的偏航稳定性。将永磁同步电机（PMSM）无传感器系统与车辆偏航力矩控制相结合，可在不使用车轮角速度传感器的情况下为这一问题提供经济有效的解决方案，同时增强偏航稳定性。本文提出了一种可增强 PMSM 速度响应的复合非线性反馈滑模控制器。在电机参数变化的情况下，所提方案的转子速度过冲和瞬态时间分别仅为 0.64% 和 0.07s，与其他方法相比更小、更短。随后，利用无传感器转子速度和无特征卡尔曼滤波器估算了车辆控制所需的关键状态和轮胎与路面摩擦系数，从而实现了 PMSM 无传感器系统与车辆偏航力矩控制的集成。此外，还介绍了一种用于增强偏航力矩控制的模糊自适应混合滑动模式方法。与其他方法相比，该方法在双线变道时保持了最小的侧滑角均方根误差（0.4192 度）。分析结果表明，不同的电机控制器和参数变化会显著影响车辆动力学性能。所提出的集成方案是可行的，能通过高性能无传感器 PMSM 系统有效增强偏航力矩控制。

{"title":"Integration of in-wheel motor sensorless systems and hierarchical direct yaw moment control for distributed drive electric vehicles","authors":"Xiaodong Wang , Maoping Ran , Xinglin Zhou","doi":"10.1016/j.engappai.2024.109600","DOIUrl":"10.1016/j.engappai.2024.109600","url":null,"abstract":"<div><div>Ensuring robust and reliable control of distributed vehicles powered by in-wheel motor systems poses a significant challenge due to the harsh operating environments and high costs of such motor systems. Poor motor control, parameter variations, and sensor malfunction under these conditions can compromise the vehicle yaw stability. Integrating permanent magnet synchronous motor (PMSM) sensorless systems with vehicle yaw moment control offers a cost-effective solution for this issue without wheel angular speed sensors while enhancing yaw stability. In this paper, a composite nonlinear feedback sliding mode controller that can enhance the PMSM speed response is proposed. The proposed scheme exhibits a rotor speed overshoot and transient time of only 0.64% and 0.07s, respectively, which are smaller and shorter compared with other methods under motor parameter changes. Subsequently, the key states and tire-road friction coefficients required for vehicle control were estimated using sensorless rotor speeds and unscented Kalman filters, enabling the integration of the PMSM sensorless system with the vehicle yaw moment control. Additionally, a fuzzy adaptive hybrid sliding mode method is presented for yaw moment control enhancement. This method maintained the smallest sideslip angle root mean square error during double lane changes (0.4192 deg) compared with other methods. Analysis results show that different motor controllers and parameter changes significantly affect the vehicle dynamics performance. The proposed integrated scheme is feasible and effectively enhances the yaw moment control via high-performance sensorless PMSM systems.</div></div>","PeriodicalId":50523,"journal":{"name":"Engineering Applications of Artificial Intelligence","volume":"139 ","pages":"Article 109600"},"PeriodicalIF":7.5,"publicationDate":"2024-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142659140","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Dynamic flame feature-driven prediction model for basic oxygen furnace steelmaking endpoint carbon content based on three-dimensional multi-layer complex networks 基于三维多层复杂网络的碱性氧气炉炼钢终点含碳量动态火焰特征驱动预测模型

IF 7.5 2区计算机科学 Q1 AUTOMATION & CONTROL SYSTEMS

Engineering Applications of Artificial Intelligence

Pub Date : 2024-11-11 DOI: 10.1016/j.engappai.2024.109564

JianXun Liu , Hui Liu , FuGang Chen , YunKe Su , Heng Li , XiaoJun Xue

Accurate prediction of carbon content at the endpoint is crucial for the endpoint management of Basic Oxygen Furnace (BOF) steelmaking. The carbon content in the molten pool is closely related to the dynamic and static characteristics of the flame at the furnace’s mouth. However, the flame’s texture change exhibits multidirectional and multiscale properties, posing challenges for existing algorithms to effectively extract dynamic color texture features. To address this issue, this paper proposes a dynamic texture feature extraction model based on a three-dimensional multi-layer complex network (3D-MLCN). The model constructs an unbounded complex network for a single-frame flame picture by integrating spatiotemporal position information of the image region’s centroid with color information, thereby quantizing the single-frame image into a complex network with spatiotemporal properties. Subsequently, a multi-scale multi-direction weighted dynamic color texture complex network is built for the flame video at the furnace mouth, utilizing the temporal index of the video frames in combination with vertex color values to capture the time-varying features of the flame video. The proposed method quantifies network characteristics through vertex degree distribution features to obtain dynamic color texture feature descriptors. These descriptors are then combined with static color texture features and color features to construct dynamic and static feature descriptors for the flame video, enabling the prediction of the endpoint carbon content using a regression model. By analyzing the actual production data of BOF steelmaking, the prediction accuracy of carbon content within the error range of ±0.02% is 87.91%, the

R^{2}

value is 0.8547, and the RMSE value is 2.0959, which verifies the effectiveness of the proposed method.

准确预测终点的碳含量对于碱性氧气炉（BOF）炼钢的终点管理至关重要。熔池中的碳含量与炉口火焰的动态和静态特征密切相关。然而，火焰的纹理变化具有多方向和多尺度的特性，这给现有算法有效提取动态颜色纹理特征带来了挑战。针对这一问题，本文提出了一种基于三维多层复合网络（3D-MLCN）的动态纹理特征提取模型。该模型通过整合图像区域中心点的时空位置信息和颜色信息，为单帧火焰图像构建无界复合网络，从而将单帧图像量化为具有时空属性的复合网络。随后，利用视频帧的时间指数结合顶点颜色值，为炉口火焰视频建立多尺度多方向加权动态颜色纹理复合网络，以捕捉火焰视频的时变特征。该方法通过顶点度分布特征量化网络特征，从而获得动态色彩纹理特征描述符。然后将这些描述符与静态颜色纹理特征和颜色特征相结合，构建火焰视频的动态和静态特征描述符，从而利用回归模型预测终点含碳量。通过分析转炉炼钢的实际生产数据，在误差±0.02%范围内碳含量的预测准确率为 87.91%，R2 值为 0.8547，RMSE 值为 2.0959，验证了所提方法的有效性。

{"title":"Dynamic flame feature-driven prediction model for basic oxygen furnace steelmaking endpoint carbon content based on three-dimensional multi-layer complex networks","authors":"JianXun Liu , Hui Liu , FuGang Chen , YunKe Su , Heng Li , XiaoJun Xue","doi":"10.1016/j.engappai.2024.109564","DOIUrl":"10.1016/j.engappai.2024.109564","url":null,"abstract":"<div><div>Accurate prediction of carbon content at the endpoint is crucial for the endpoint management of Basic Oxygen Furnace (BOF) steelmaking. The carbon content in the molten pool is closely related to the dynamic and static characteristics of the flame at the furnace’s mouth. However, the flame’s texture change exhibits multidirectional and multiscale properties, posing challenges for existing algorithms to effectively extract dynamic color texture features. To address this issue, this paper proposes a dynamic texture feature extraction model based on a three-dimensional multi-layer complex network (3D-MLCN). The model constructs an unbounded complex network for a single-frame flame picture by integrating spatiotemporal position information of the image region’s centroid with color information, thereby quantizing the single-frame image into a complex network with spatiotemporal properties. Subsequently, a multi-scale multi-direction weighted dynamic color texture complex network is built for the flame video at the furnace mouth, utilizing the temporal index of the video frames in combination with vertex color values to capture the time-varying features of the flame video. The proposed method quantifies network characteristics through vertex degree distribution features to obtain dynamic color texture feature descriptors. These descriptors are then combined with static color texture features and color features to construct dynamic and static feature descriptors for the flame video, enabling the prediction of the endpoint carbon content using a regression model. By analyzing the actual production data of BOF steelmaking, the prediction accuracy of carbon content within the error range of ±0.02% is 87.91%, the <span><math><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup></math></span> value is 0.8547, and the RMSE value is 2.0959, which verifies the effectiveness of the proposed method.</div></div>","PeriodicalId":50523,"journal":{"name":"Engineering Applications of Artificial Intelligence","volume":"139 ","pages":"Article 109564"},"PeriodicalIF":7.5,"publicationDate":"2024-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142659273","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Explained fire resistance machine learning models for compressed steel members of trusses and bracing systems 解释了用于桁架和支撑系统受压钢构件的耐火机器学习模型

IF 7.5 2区计算机科学 Q1 AUTOMATION & CONTROL SYSTEMS

Engineering Applications of Artificial Intelligence

Pub Date : 2024-11-11 DOI: 10.1016/j.engappai.2024.109571

Luca Possidente , Carlos Couto

Trusses and bracing systems are usually constructed from monosymmetric and built-up cross-sections, which under compression stresses may buckle in torsional or flexural–torsional modes. In fire, this phenomenon is utterly important as failure in bracing systems or trusses may cause the collapse of buildings and result in loss of lives or severe economic impacts. Machine learning models, including neural networks, random forests and support vector machines, are developed considering a dataset with 21879 samples and are further assessed in this study as an alternative with greater accuracy and ease of application over existing design methods, namely the Eurocode 3 Part 1-2 and a recent proposal for its improvement. The machine learning models are explained using a combination of domain knowledge inference, partial dependence plots and SHapley Additive exPlanations. The accuracy versus safety trade-off is discussed for a better-informed model selection. The proposed approach and discussion create an additional confidence layer for applying these techniques for the fire design.

桁架和支撑系统通常是由单对称和加固的横截面建造而成，在压缩应力作用下，它们可能会以扭转或挠曲-扭转模式发生弯曲。在火灾中，这种现象非常重要，因为支撑系统或桁架的失效可能导致建筑物倒塌，造成生命损失或严重的经济影响。本研究开发了包括神经网络、随机森林和支持向量机在内的机器学习模型，这些模型考虑了 21879 个样本的数据集，并在本研究中进行了进一步评估，认为与现有的设计方法（即欧洲规范 3 第 1-2 部分和最近提出的改进建议）相比，这些模型具有更高的准确性和更高的易用性。机器学习模型结合使用了领域知识推断、部分依存图和 SHapley Additive exPlanations。讨论了准确性与安全性之间的权衡，以便选择更明智的模型。所提出的方法和讨论为将这些技术应用于消防设计提供了一个额外的信心层。

{"title":"Explained fire resistance machine learning models for compressed steel members of trusses and bracing systems","authors":"Luca Possidente , Carlos Couto","doi":"10.1016/j.engappai.2024.109571","DOIUrl":"10.1016/j.engappai.2024.109571","url":null,"abstract":"<div><div>Trusses and bracing systems are usually constructed from monosymmetric and built-up cross-sections, which under compression stresses may buckle in torsional or flexural–torsional modes. In fire, this phenomenon is utterly important as failure in bracing systems or trusses may cause the collapse of buildings and result in loss of lives or severe economic impacts. Machine learning models, including neural networks, random forests and support vector machines, are developed considering a dataset with 21879 samples and are further assessed in this study as an alternative with greater accuracy and ease of application over existing design methods, namely the Eurocode 3 Part 1-2 and a recent proposal for its improvement. The machine learning models are explained using a combination of domain knowledge inference, partial dependence plots and SHapley Additive exPlanations. The accuracy versus safety trade-off is discussed for a better-informed model selection. The proposed approach and discussion create an additional confidence layer for applying these techniques for the fire design.</div></div>","PeriodicalId":50523,"journal":{"name":"Engineering Applications of Artificial Intelligence","volume":"139 ","pages":"Article 109571"},"PeriodicalIF":7.5,"publicationDate":"2024-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142658987","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Residual-based attention Physics-informed Neural Networks for spatio-temporal ageing assessment of transformers operated in renewable power plants 基于剩余注意力的物理信息神经网络用于可再生能源发电厂变压器的时空老化评估

IF 7.5 2区计算机科学 Q1 AUTOMATION & CONTROL SYSTEMS

Engineering Applications of Artificial Intelligence

Pub Date : 2024-11-11 DOI: 10.1016/j.engappai.2024.109556

Ibai Ramirez , Joel Pino , David Pardo , Mikel Sanz , Luis del Rio , Alvaro Ortiz , Kateryna Morozovska , Jose I. Aizpurua

Transformers are crucial for reliable and efficient power system operations, particularly in supporting the integration of renewable energy. Effective monitoring of transformer health is critical to maintain grid stability and performance. Thermal insulation ageing is a key transformer failure mode, which is generally tracked by monitoring the hotspot temperature (HST). However, HST measurement is complex, costly, and often estimated from indirect measurements. Existing HST models focus on space-agnostic thermal models, providing worst-case HST estimates. This article introduces a spatio-temporal model for transformer winding temperature and ageing estimation, which leverages physics-based partial differential equations (PDEs) with data-driven Neural Networks (NN) in a Physics Informed Neural Networks (PINNs) configuration to improve prediction accuracy and acquire spatio-temporal resolution. The computational accuracy of the PINN model is improved through the implementation of the Residual-Based Attention (PINN-RBA) scheme that accelerates the PINN model convergence. The PINN-RBA model is benchmarked against self-adaptive attention schemes and classical vanilla PINN configurations. For the first time, PINN based oil temperature predictions are used to estimate spatio-temporal transformer winding temperature values, validated through PDE numerical solution and fiber optic sensor measurements. Furthermore, the spatio-temporal transformer ageing model is inferred, which supports transformer health management decision-making. Results are validated with a distribution transformer operating on a floating photovoltaic power plant.

变压器对于电力系统的可靠和高效运行至关重要，尤其是在支持可再生能源的整合方面。有效监测变压器的健康状况对于保持电网的稳定性和性能至关重要。热绝缘老化是变压器的主要故障模式，一般通过监测热点温度（HST）来跟踪。然而，HST 测量复杂、成本高，而且通常是通过间接测量估算出来的。现有的 HST 模型侧重于与空间无关的热模型，提供最坏情况下的 HST 估计值。本文介绍了一种用于变压器绕组温度和老化估算的时空模型，该模型在物理信息神经网络（PINNs）配置中利用基于物理的偏微分方程（PDEs）和数据驱动的神经网络（NNs）来提高预测精度并获得时空分辨率。通过实施基于残差的注意力（PINN-RBA）方案，加速了 PINN 模型的收敛，从而提高了 PINN 模型的计算精度。PINN-RBA 模型以自适应注意力方案和经典虚无 PINN 配置为基准。基于 PINN 的油温预测首次用于估算变压器绕组的时空温度值，并通过 PDE 数值解决方案和光纤传感器测量进行了验证。此外，还推断出变压器的时空老化模型，为变压器健康管理决策提供支持。结果通过浮动光伏电站上运行的配电变压器进行了验证。

{"title":"Residual-based attention Physics-informed Neural Networks for spatio-temporal ageing assessment of transformers operated in renewable power plants","authors":"Ibai Ramirez , Joel Pino , David Pardo , Mikel Sanz , Luis del Rio , Alvaro Ortiz , Kateryna Morozovska , Jose I. Aizpurua","doi":"10.1016/j.engappai.2024.109556","DOIUrl":"10.1016/j.engappai.2024.109556","url":null,"abstract":"<div><div>Transformers are crucial for reliable and efficient power system operations, particularly in supporting the integration of renewable energy. Effective monitoring of transformer health is critical to maintain grid stability and performance. Thermal insulation ageing is a key transformer failure mode, which is generally tracked by monitoring the hotspot temperature (HST). However, HST measurement is complex, costly, and often estimated from indirect measurements. Existing HST models focus on space-agnostic thermal models, providing worst-case HST estimates. This article introduces a spatio-temporal model for transformer winding temperature and ageing estimation, which leverages physics-based partial differential equations (PDEs) with data-driven Neural Networks (NN) in a Physics Informed Neural Networks (PINNs) configuration to improve prediction accuracy and acquire spatio-temporal resolution. The computational accuracy of the PINN model is improved through the implementation of the Residual-Based Attention (PINN-RBA) scheme that accelerates the PINN model convergence. The PINN-RBA model is benchmarked against self-adaptive attention schemes and classical vanilla PINN configurations. For the first time, PINN based oil temperature predictions are used to estimate spatio-temporal transformer winding temperature values, validated through PDE numerical solution and fiber optic sensor measurements. Furthermore, the spatio-temporal transformer ageing model is inferred, which supports transformer health management decision-making. Results are validated with a distribution transformer operating on a floating photovoltaic power plant.</div></div>","PeriodicalId":50523,"journal":{"name":"Engineering Applications of Artificial Intelligence","volume":"139 ","pages":"Article 109556"},"PeriodicalIF":7.5,"publicationDate":"2024-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142659044","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

WIGNN: An adaptive graph-structured reasoning model for credit default prediction WIGNN：用于信用违约预测的自适应图结构推理模型

IF 7.5 2区计算机科学 Q1 AUTOMATION & CONTROL SYSTEMS

Engineering Applications of Artificial Intelligence

Pub Date : 2024-11-11 DOI: 10.1016/j.engappai.2024.109597

Zhipeng Yan , Hanwen Qu , Chen Chen , Xiaoyi Lv , Enguang Zuo , Kui Wang , Xulun Cai

In credit default prediction, the main challenge is handling complex data structures and addressing data class imbalance. Given class imbalance and multi-dimensional data, general models find it difficult to fully explore the deep interdependencies within the data and the interaction effects between local and global. To overcome these challenges, this study proposes a Weighted Imbalanced Graph Neural Network (WIGNN) model that integrates adaptive graph structure inference with differential weight connectivity strategy, and the model solves the existing problems from the perspective of differential weight connectivity and graph balancing. Here, the weight connection uses the Gaussian kernel function to refine calculations and an adaptive percentile method to adjust sparsity, improving the understanding and efficiency of mining data connections. The weighted graph generated by this method can reflect the interaction between nodes and improve the model’s ability to analyse complex data structures. Based on this weighted graph, the graph imbalance module adopts a reinforcement learning-driven neighbour sampling strategy to adjust the sampling threshold automatically, optimizes the node embedding through message aggregation, and combines with a cost-sensitive matrix to improve classification accuracy and cost-effectiveness of the model on diverse credit datasets. We applied the WIGNN model to six real and class-imbalanced credit datasets, comparing it with 11 mainstream credit default prediction models. Evaluated using metrics Area Under the Curve (AUC), Geometric Mean (G-mean), and Accuracy. The results show that WIGNN significantly outperforms other models in handling class imbalance and graph sparsity, demonstrating its potential in financial credit applications.

在信用违约预测中，主要的挑战是处理复杂的数据结构和解决数据类别不平衡问题。在类不平衡和多维数据的情况下，一般模型很难充分探索数据内部深层次的相互依赖关系，以及局部和全局之间的交互影响。为了克服这些挑战，本研究提出了一种加权不平衡图神经网络（WIGNN）模型，该模型将自适应图结构推断与差分权重连接策略相结合，从差分权重连接和图平衡的角度解决了现有问题。在这里，权重连接使用高斯核函数来细化计算，并使用自适应百分位法来调整稀疏性，从而提高了对数据连接的理解和挖掘效率。这种方法生成的加权图可以反映节点之间的互动关系，提高模型分析复杂数据结构的能力。在此加权图的基础上，图不平衡模块采用强化学习驱动的邻域采样策略自动调整采样阈值，通过消息聚合优化节点嵌入，并结合成本敏感矩阵，提高模型在不同信贷数据集上的分类精度和性价比。我们将 WIGNN 模型应用于六个真实的、类不平衡的信用数据集，并与 11 个主流信用违约预测模型进行了比较。评估指标包括曲线下面积（AUC）、几何平均数（G-mean）和准确率。结果表明，WIGNN 在处理类不平衡和图稀疏性方面明显优于其他模型，证明了其在金融信贷应用中的潜力。

{"title":"WIGNN: An adaptive graph-structured reasoning model for credit default prediction","authors":"Zhipeng Yan , Hanwen Qu , Chen Chen , Xiaoyi Lv , Enguang Zuo , Kui Wang , Xulun Cai","doi":"10.1016/j.engappai.2024.109597","DOIUrl":"10.1016/j.engappai.2024.109597","url":null,"abstract":"<div><div>In credit default prediction, the main challenge is handling complex data structures and addressing data class imbalance. Given class imbalance and multi-dimensional data, general models find it difficult to fully explore the deep interdependencies within the data and the interaction effects between local and global. To overcome these challenges, this study proposes a Weighted Imbalanced Graph Neural Network (WIGNN) model that integrates adaptive graph structure inference with differential weight connectivity strategy, and the model solves the existing problems from the perspective of differential weight connectivity and graph balancing. Here, the weight connection uses the Gaussian kernel function to refine calculations and an adaptive percentile method to adjust sparsity, improving the understanding and efficiency of mining data connections. The weighted graph generated by this method can reflect the interaction between nodes and improve the model’s ability to analyse complex data structures. Based on this weighted graph, the graph imbalance module adopts a reinforcement learning-driven neighbour sampling strategy to adjust the sampling threshold automatically, optimizes the node embedding through message aggregation, and combines with a cost-sensitive matrix to improve classification accuracy and cost-effectiveness of the model on diverse credit datasets. We applied the WIGNN model to six real and class-imbalanced credit datasets, comparing it with 11 mainstream credit default prediction models. Evaluated using metrics Area Under the Curve (AUC), Geometric Mean (G-mean), and Accuracy. The results show that WIGNN significantly outperforms other models in handling class imbalance and graph sparsity, demonstrating its potential in financial credit applications.</div></div>","PeriodicalId":50523,"journal":{"name":"Engineering Applications of Artificial Intelligence","volume":"139 ","pages":"Article 109597"},"PeriodicalIF":7.5,"publicationDate":"2024-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142659275","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Guided deep reinforcement learning framework using automated curriculum scheme for accurate motion planning 利用自动课程计划指导深度强化学习框架，实现精确的运动规划

IF 7.5 2区计算机科学 Q1 AUTOMATION & CONTROL SYSTEMS

Engineering Applications of Artificial Intelligence

Pub Date : 2024-11-11 DOI: 10.1016/j.engappai.2024.109541

Deun-Sol Cho , Jae-Min Cho , Won-Tae Kim

Collaborative robotic arms in smart factories should ensure the safety and interactivity during their operation such as reaching and grasping objects. Especially, the advanced motion planner including the path planning and the motion control functions is essential for human-machine co-working. Since the traditional physics-based motion planning approaches require extreme computational resources to obtain near-optimal solutions, deep reinforcement learning algorithms have been actively adopted and have effectively solved the limitation. They, however, have the easy task preference problem, primarily taking the simpler ways for the more rewards, due to randomly training the agents how to reach the target points in the large-scale search spaces. Therefore, we propose a novel curriculum-based deep reinforcement learning framework that makes the agents learn the motion planning tasks in unbiased ways from the ones with the low complexities to the others with the high complexities. It uses the unsupervised learning algorithms to cluster the target points with the similar task complexities for generating the effective curriculum. In addition, the review and buffer flushing mechanisms are integrated into the framework to mitigate the catastrophic forgetting problem where the agent abruptly lose the previous learned knowledge upon learning new one in the curriculum. The evaluation results of the proposed framework show that the curriculum significantly enhances the success rate on the task with the highest complexity from 12% to 56% and the mechanisms improve the success rate on the tasks with the easier complexities from an average of 66% to 76.5%, despite requiring less training time.

智能工厂中的协作机械臂在伸手抓取物体等操作过程中应确保安全性和互动性。特别是，包括路径规划和运动控制功能在内的高级运动规划器对于人机协同工作至关重要。由于传统的基于物理的运动规划方法需要极大的计算资源才能获得接近最优的解决方案，深度强化学习算法已被积极采用，并有效地解决了这一限制。然而，由于要随机训练代理如何在大规模搜索空间中到达目标点，它们存在任务偏好简单的问题，主要是为了获得更多奖励而采取更简单的方法。因此，我们提出了一种新颖的基于课程的深度强化学习框架，它能让代理以无偏见的方式学习运动规划任务，从复杂度低的任务到复杂度高的任务。它使用无监督学习算法对任务复杂度相似的目标点进行聚类，从而生成有效的课程。此外，该框架还集成了复习和缓冲区冲洗机制，以缓解灾难性遗忘问题，即代理在学习课程中的新知识时突然丢失之前学习的知识。对所提框架的评估结果表明，尽管所需的训练时间较少，但课程能显著提高复杂度最高任务的成功率，从 12% 提高到 56%，而机制能提高复杂度较低任务的成功率，平均从 66% 提高到 76.5%。

{"title":"Guided deep reinforcement learning framework using automated curriculum scheme for accurate motion planning","authors":"Deun-Sol Cho , Jae-Min Cho , Won-Tae Kim","doi":"10.1016/j.engappai.2024.109541","DOIUrl":"10.1016/j.engappai.2024.109541","url":null,"abstract":"<div><div>Collaborative robotic arms in smart factories should ensure the safety and interactivity during their operation such as reaching and grasping objects. Especially, the advanced motion planner including the path planning and the motion control functions is essential for human-machine co-working. Since the traditional physics-based motion planning approaches require extreme computational resources to obtain near-optimal solutions, deep reinforcement learning algorithms have been actively adopted and have effectively solved the limitation. They, however, have the easy task preference problem, primarily taking the simpler ways for the more rewards, due to randomly training the agents how to reach the target points in the large-scale search spaces. Therefore, we propose a novel curriculum-based deep reinforcement learning framework that makes the agents learn the motion planning tasks in unbiased ways from the ones with the low complexities to the others with the high complexities. It uses the unsupervised learning algorithms to cluster the target points with the similar task complexities for generating the effective curriculum. In addition, the review and buffer flushing mechanisms are integrated into the framework to mitigate the catastrophic forgetting problem where the agent abruptly lose the previous learned knowledge upon learning new one in the curriculum. The evaluation results of the proposed framework show that the curriculum significantly enhances the success rate on the task with the highest complexity from 12% to 56% and the mechanisms improve the success rate on the tasks with the easier complexities from an average of 66% to 76.5%, despite requiring less training time.</div></div>","PeriodicalId":50523,"journal":{"name":"Engineering Applications of Artificial Intelligence","volume":"139 ","pages":"Article 109541"},"PeriodicalIF":7.5,"publicationDate":"2024-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142659270","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0