检索结果-内蒙古大学图书馆

Design Principles for Model Generalization and Scalable AI Integration in Radio Access Networks

ieee COMMUNICATIONS MAGAZINE 2025年第1期63卷 84-91页

作者： Soldati, Pablo Ghadimi, Euhanna Demirel, Burak Wang, Yu Gaigalas, Raimundas Sintorn, Mathias Ericsson AB Stockholm Sweden

This article emphasizes the pivotal role of model generalization in enhancing performance and enabling scalable AI integration within radio communications. We outline design principles for model generalization in three key domains: environment for robustness, intents for adaptability to system objectives, and control tasks for reducing AI-driven control loops. Adopting these principles can decrease the number of models deployed and increase adaptability in diverse radio communication environments. To address the challenges of model generalization in communication systems, we propose a learning architecture that leverages centralization of training and data management functionalities, combined with distributed data generation. We illustrate these concepts by designing a generalized link adaptation algorithm, demonstrating the benefits of our proposed approach.

关键词： Artificial intelligence Adaptation models data models Task analysis Training Robustness Uncertainty

来源：评论

学校读者我要写书评

暂无评论

Adaptive Quality control With Uncertainty for a Pharmaceutical Cyber-Physical System Based on data and Knowledge Integration

引用

ieee TRANSACTIONS ON INDUSTRIAL INFORMATICS 2024年第3期20卷 3339-3350页

作者： Wang, Zhengsong Tang, Shengnan Guo, Ge Yang, Yanqiu He, Dakuo Yang, Le Han, Meng Hou, Yue Northeastern Univ Coll Informat Sci & Engn Shenyang 110819 Peoples R China Northeastern Univ Coll Informat Sci & Engn State Key Lab Synthet Automat Proc Ind Shenyang 110819 Peoples R China Northeastern Univ Coll Life & Hlth Sci Shenyang 110819 Peoples R China

In the context of Pharma 4.0, i.e., the pharmaceutical version of Industry 4.0, pharmaceutical quality control (PQC) for a pharmaceutical cyber-physical system (PCPS) plays a critical role in ensuring the quality of drug products during pharmaceutical development. However, drug customization through Pharma 4.0 also introduces uncertainty embodied in ever-changing critical material attributes, which presents new challenges related to development costs and efficiency in PQC compared to traditional control modes. Although we have proposed a data-driven methodology to tackle these challenges, it ignores some visible or potential process knowledge that also contains much additional information reflecting the laws and trends governing pharmaceutical process operations. This not only sacrifices the opportunity to use this knowledge to make up for insufficiencies in the information provided by the data but also goes against the core ideology of future intelligent manufacturing. In this article, we introduce the idea of a data- and knowledge-driven approach into PQC for the first time by proposing a general data- and knowledge-driven adaptive PQC framework for a PCPS-based two phases-PQC by direct data- and knowledge-driven adaptive iterative learning control and PQC by learning from primitive data and knowledge. Next, a case study is presented to preliminarily investigate the application of the proposed framework in a simulated pharmaceutical spray fluidized bed granulation process. Finally, a series of simulation experiments are designed to verify the feasibility and effectiveness of the proposed framework.

关键词： Process control Adaptive systems Uncertainty Drugs Solid modeling Predictive models Adaptation models data- and knowledge-driven adaptive pharmaceutical quality control (PQC) framework PQC spray fluidized bed granulation (SFBG) uncertainty

来源：评论

学校读者我要写书评

暂无评论

Neural Schrödinger Bridge With Sinkhorn Losses: Application to data-driven Minimum Effort control of Colloidal Self-Assembly

引用

ieee TRANSACTIONS ON control systems TECHNOLOGY 2024年第3期32卷 960-973页

作者： Nodozi, Iman Yan, Charlie Khare, Mira Halder, Abhishek Mesbah, Ali Univ Calif Santa Cruz Dept Elect & Comp Engn Santa Cruz CA 95064 USA Univ Calif Berkeley Dept Chem & Biomol Engn Berkeley CA 94720 USA Iowa State Univ Dept Aerosp Engn Ames IA 50011 USA

We show that the minimum effort control of colloidal self-assembly (SA) can be naturally formulated in the order-parameter space as a generalized Schrodinger bridge problem (GSBP)-a class of fixed-horizon stochastic optimal control problems that originated in the works of Erwin Schrodinger in the early 1930s. In recent years, this class of problems has seen a resurgence of research activities in the control and machine-learning communities. Different from the existing literature on the theory and computation for such problems, the controlled drift and diffusion coefficients for colloidal SA are typically nonaffine in control and are difficult to obtain from physics-based modeling. We deduce the conditions of optimality for such generalized problems and show that the resulting system of equations is structurally very different from the existing results in a way that standard computational approaches no longer apply. Thus motivated, we propose a data-driven learning and control framework, named "neural Schrodinger bridge", to solve such generalized Schrodinger bridge problems by innovating on recent advances in neural networks (NNs). We illustrate the effectiveness of the proposed framework using a numerical case study of colloidal SA. We learn the controlled drift and diffusion coefficients as two NNs using molecular dynamics (MD) simulation data and then use these two to train a third network with Sinkhorn losses designed for distributional endpoint constraints, specific for this class of control problems.

关键词： Colloidal self-assembly (SA) physics-informed neural networks (PINNs) Schr & ouml dinger bridge Sinkhorn loss stochastic optimal control

来源：评论

学校读者我要写书评

暂无评论

Safe Active learning and Probabilistic Design of Experiment for Autonomous Hydraulic Excavators

Safe Active Learning and Probabilistic Design of Experiment ...

引用

ieee/RSJ International conference on Intelligent Robots and systems (IROS)

作者： Dio, Maximilian Demir, Ozan Trachte, Adrian Graichen, Knut Friedrich Alexander Univ Erlangen Nurnberg Chair Automat Control Erlangen Germany Robert Bosch GmbH Renningen Germany

ISBN: (纸本)9781665491907

Recently, data-driven and hybrid control of hydraulic cylinders for excavator assistance functions have been in the focus of many research papers. To ensure an accurate behavior, data-driven controllers and models need a large amount of data to cover all relevant operation regions, which requires a time-consuming data generation process. In this work, we introduce two learning-based methods to enhance the efficiency of this procedure: a static learning method and an active learning method. Both methods reduce the amount of required data to learn a hydraulic inverse actuation model. Compared to previous collection methods, the required data was reduced by factor 7.5, while the information content of the dataset remains nearly the same.

关键词： Design of experiments

来源：评论

学校读者我要写书评

暂无评论

Indirect Prediction Method for Remaining Useful Life of Lithium-ion Battery based on Gray Wolf Optimized Extreme learning Machine 12

Indirect Prediction Method for Remaining Useful Life of Lith...

引用

ieee 12th data driven control and learning systems conference (DDCLS)

作者： Ding Miaomiao Wang Xianghua Shandong Univ Sci & Technol Coll Elect Engn & Automat Qingdao 266590 Peoples R China

ISBN: (纸本)9798350321050

In this paper, a novel indirect method for predicting the remaining useful life (RUL) of lithium-ion battery is proposed, not requiring the direct measurement of capacity. Firstly, a variety of indirect health indicators (HI) that can characterize the degradation state of lithium-ion battery are extracted from the charging and discharging process curves of lithium-ion battery. To reduce the redundancy among indirect HIs Principal Component Analysis (PCA) is used to fuse all these HIs into one indirect HI that can represent the degradation characteristics. Then, in order to avoid the interference of noise on the prediction accuracy of the model, we use the Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) to decompose and denoise the fused indirect HI. Finally, the Gray Wolf Optimized Extreme learning Machine (GWO-ELM) model is applied to make the RUL prediction and compare its prediction results with the prediction results of other models. Finally, simulations are conducted and the results show that the proposed indirect prediction method can predict lithium-ion battery RUL more accurately.

关键词： lithium-ion battery,RUL prediction PCA,CEEMDAN,GWO-ELM

来源：评论

学校读者我要写书评

暂无评论

learning-Based Modeling and Predictive control for Unknown Nonlinear System With Stability Guarantees

引用

ieee TRANSACTIONS ON NEURAL NETWORKS AND learning systems 2025年第6期PP卷 PP页

作者： Jin, Ao Zhang, Fan Shen, Ganghui Huang, Bingxiao Huang, Panfeng Shaanxi Prov Innovat Team Intelligent Robot Techno Xian 710072 Peoples R China Northwestern Polytech Univ Sch Automat Res Ctr Intelligent Robot Xian 710072 Peoples R China Northwestern Polytech Univ Sch Astronaut Res Ctr Intelligent Robot Xian 710072 Peoples R China

This work focuses on the safety of learning-based control for unknown nonlinear system, considering the stability of learned dynamics and modeling mismatch between the learned dynamics and the true one. A learning-based scheme imposing the stability constraint is proposed in this work for modeling and stable control of unknown nonlinear system. Specifically, a linear representation of unknown nonlinear dynamics is established using the Koopman theory. Then, a deep learning approach is utilized to approximate embedding functions of Koopman operator for unknown system. For the safe manipulation of proposed scheme in the real-world applications, a stable constraint of learned dynamics and Lipschitz constraint of embedding functions are considered for learning a stable model for prediction and control. Moreover, a robust predictive control scheme is adopted to eliminate the effect of modeling mismatch between the learned dynamics and the true one, such that the stabilization of unknown nonlinear system is achieved. Finally, the effectiveness of proposed scheme is demonstrated on the tethered space robot (TSR) with unknown nonlinear dynamics.

关键词： data-driven control learning-based control model predictive control nonlinear system

来源：评论

学校读者我要写书评

暂无评论

Model Free Adaptive control for Nonlinear Multi-Agent systems under Actuator Faults and DoS Attacks 13

Model Free Adaptive Control for Nonlinear Multi-Agent System...

引用

13th ieee data driven control and learning systems conference, DDCLS 2024

作者： Chuai, Ce Yin, Haibin Wang, Fuyong Liu, Zhongxin College of Artificial Intelligence Nankai University Tianjin300350 China

ISBN: (纸本)9798350361674

A model free fault-tolerant and intrusion-tolerant control approach is developed for nonlinear multi-agent systems (MASs) characterized by unknown dynamics, facing denial-of-service (DoS) attacks and actuator faults. Initially, the equivalent dynamic linearization method based on pseudo-partial derivatives is employed to characterize actuator fault information, while models for DoS attacks on communication links follow the Bernoulli distribution. Compensation mechanisms for actuator faults and data packets loss are then introduced. Subsequently, a distributed model free adaptive control scheme is devised based on above mechanisms, and the bounded convergence of the tracking error is proven using a compressive mapping method. Numerical simulations showcase the efficacy of the control approach. © 2024 ieee.

关键词： Adaptive control systems

来源：评论

学校读者我要写书评

暂无评论

Improved Model - Free Sliding Mode control Of PMSM Based On Finite-Time Generalized Proportional Integral Observer 12

Improved Model - Free Sliding Mode Control Of PMSM Based On ...

引用

ieee 12th data driven control and learning systems conference (DDCLS)

作者： Wang, Dongdong Liu, Shuguang Liu, Xudong Zhang, Zhixin Chen, Yongzhi Qingdao Univ Sch Automat Qingdao 266071 Peoples R China Shandong Luruan Digital Technol Co LTD Jinan 250000 Peoples R China Shandong Key Lab Ind Control Technol Qingdao 266071 Peoples R China

ISBN: (纸本)9798350321050

Aiming at the problems of poor control performance and poor anti-interference performance of PMSM control system due to external load disturbance, model and parameter uncertainty, a composite control method of PMSM based on improved model-free sliding mode control is proposed. Firstly, the ultra-local model of PMSM system without considering motor parameters is established. Then, a model-free single-loop sliding mode speed controller is designed according to the ultra-local model of the system. Furthermore, for the total disturbance of the system, a finite-time generalized proportional integral observer (FT-GPIO) is established, which can estimate the total disturbance of the system more accurately and stably, and perform feedforward compensation to update the control information in real time. Finally, compared with the traditional control method, the simulation results show that the proposed control method has good dynamic performance, anti-interference ability and steady-state control accuracy.

关键词： Model-free control PMSM sliding mode control finite-time generalized proportional integral observer

来源：评论

学校读者我要写书评

暂无评论

Integral-Type Event-Triggered Model Predictive control for Manipulator systems 13

Integral-Type Event-Triggered Model Predictive Control for M...

引用

13th ieee data driven control and learning systems conference, DDCLS 2024

作者： Liu, Shifeng Ren, Xuemei Wang, Pengbiao Beijing Institute of Technology Beijing100081 China

ISBN: (纸本)9798350361674

In this paper, we study an integral-type event-triggered model predictive control method for manipulator systems. The event-triggered mechanism, which is based on integrating the error between the actual and predicted states, is applied to the system when the state is outside the terminal domain. In comparison to directly using the difference, the proposed method achieves a more effective balance between system performance and the frequency of solving optimization problems. Furthermore, the event-triggered strategy is also considered in the state-feedback controller to save computing resources. The feasibility and stability are rigorously analyzed. Finally, the simulation results of the manipulator system validate the feasibility and effectiveness of the proposed algorithm. © 2024 ieee.

关键词： Predictive control systems

来源：评论

学校读者我要写书评

暂无评论

A robust variable projection algorithm for RBF-AR model 12

A robust variable projection algorithm for RBF-AR model

引用

ieee 12th data driven control and learning systems conference (DDCLS)

作者： She, Yuexin Chen, Guangyong Gan, Min Fuzhou Univ Coll Comp & Data Sci Fuzhou 350000 Peoples R China

ISBN: (纸本)9798350321050

The radial basis function network-based autoregressive (RBF-AR) model is a powerful statistical model which can be expressed as a linear combination of nonlinear functions and frequently appears in a wide range of application fields. Variable projection algorithm is designed for solving smooth separable optimization problems with least squares form and has been used as an efficient tool for the identification of RBF-AR model. However, in real applications, the observations are usually disturbed by non-Gaussian noise or contain outliers. This often leads to nonlinear regression problems. Since there are both linear and nonlinear parameters in such problems, how to optimize such models is still challenging. In this paper, we design a robust variable projection algorithm for the identification of RBF-AR model. The proposed method takes into account the coupling of the linear and nonlinear parameters of RBF-AR model, which eliminates the linear parameters by solving a linear programming and optimizes the reduced function that only contains nonlinear parameters. Numerical results on RBF-AR model to synthetic data and real-world data confirm the effectiveness of the proposed algorithm.

关键词： variable projection RBF-AR model nonlinear regression problem robust parameter estimation

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：