检索结果-南通市图书馆

Offline Reinforcement Learning with Constrained Hybrid Action Implicit Representation Towards Wargaming Decision-Making

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

Tsinghua Science and Technology 2024年第5期29卷 1422-1440页

作者： Liwei Dong Ni Li guanghong gong Xin Lin School of Automation Science and Electrical Engineering Beihang UniversityBeijing 100191China School of Automation Science and Electrical Engineering Beihang UniversityBeijing 100191ChinaState Key Laboratory of Virtual Reality Technology and SystemsBeihang UniversityBeijing 100191Chinaand also with Zhongguancun LaboratoryBeijing 100191China School of Automation Science and Electrical Engineering Beihang UniversityBeijing 100191Chinaand also with State Key Laboratory of Virtual Reality Technology and SystemsBeihang UniversityBeijing 100191China

Reinforcement Learning(RL)has emerged as a promising data-driven solution for wargaming ***,two domain challenges still exist:(1)dealing with discrete-continuous hybrid wargaming control and(2)accelerating RL deployment with rich offline *** RL methods fail to handle these two issues simultaneously,thereby we propose a novel offline RL method targeting hybrid action space.A new constrained action representation technique is developed to build a bidirectional mapping between the original hybrid action space and a latent space in a semantically consistent *** allows learning a continuous latent policy with offline RL with better exploration feasibility and scalability and reconstructing it back to a needed hybrid ***,a novel offline RL optimization objective with adaptively adjusted constraints is designed to balance the alleviation and generalization of out-of-distribution *** method demonstrates superior performance and generality across different tasks,particularly in typical realistic wargaming scenarios.

关键词： offline Reinforcement Learning(RL) wargaming decision-making hybrid action space

A parameterized geometry design method for inward turning inlet compatible waverider

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Chinese Journal of Aeronautics 2013年第5期26卷 1135-1146页

作者： Tian Chao Li Ni gong guanghong Su Zeya Department of Automation Science and Electrical Engineering Beihang University

Intensive studies have been carried out on generations of waverider geometry and hypersonic inlet geometry. However, integration efforts of waverider and related air-intake system are restricted majorly around the X43A-like or conical flow field induced configuration, which adopts mainly the two-dimensional air-breathing technology and limits the judicious visions of developing new aerodynamic profiles for hypersonic designers. A novel design approach for integrating the inward turning inlet with the traditional parameterized waverider is proposed. The proposed method is an alternative means to produce a compatible configuration by linking the off-the-shelf results on both traditional waverider techniques and inward turning inlet techniques. A series of geometry generations and optimization solutions is proposed to enhance the lift-to-drag ratio. A quantitative but efficient aerodynamic performance evaluation approach （the hypersonic flow panel method） with lower computational cost is employed to play the role of objective function for opti- mization purpose. The produced geometry compatibility with a computational fluid dynamics （CFD） solver is also verified for detailed flow field investigation. Optimization results and other numerical validations are obtained for the feasibility demonstration of the proposed method.

关键词： Hypersonic inlets Hypersonic vehicles Inward turning inlet Optimization Parameterization Waverider

Multimodal fusion of EEG and fMRI for epilepsy detection

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

International Journal of Modeling, Simulation, and Scientific Computing 2018年第2期9卷 23-35页

作者： Xiashuang Wang guanghong gong Ni Li School of Automation Science and Electrical Engineering Beihang University XueYuan Road No.37 HaiDian DistrictBeijing 100191P.R.China

Technology of brain–computer interface(BCI)provides a new way of communication and control without language or physical *** signal tracking and positioning is the basis of BCI research,while brain modeling affects the treatment analysis of(EEG)and functional magnetic resonance imaging(fMRI)*** paper proposes human ellipsoid brain modeling ***,we use non-parametric spectral estimation method of time–frequency analysis to deal with simulation and real EEG of epilepsy patients,which utilizes both the high spatial and the high time resolution to improve the doctor’s diagnostic efficiency.

关键词： Brain model time-frequency analysis medical diagnosis multimodal fusion.

A comprehensive performance evaluation framework of complex products based on a fuzzy AHP and DS theory

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

International Journal of Modeling, Simulation, and Scientific Computing 2016年第3期7卷 181-198页

作者： Yuhong Li guanghong gong Ni Li School of Automation Science Beihang University(BUAA)Beijing 100191 P.R.China

With the development of computer technique,performance evaluation of complex products is playing an increasingly critical role in ensuring product quality and improving development *** extensible comprehensive performance evaluation framework with the integration of effective group decision-making algorithms could be a supporting tool to achieve an efficient evaluation process and reduce comprehensive evaluation *** paper aims to provide a evaluation framework with friendly interactive operation and extensive expansibility,which adopts a multi-expert evaluation approach based on fuzzy,analytical hierarchy process(FAHP)and Dempstere–Shafer(DS)theory(FADS)in order to consider experts’relative importance *** addition,an extensible evaluation process and related auxiliary functions are implemented in the framework,including the establishment of an assessment index system,integration and calls of multiple types of testing data preprocessing methods and index assessment methods suitable for small sample data,graphical result display and data analysis,***,performance evaluation cases of two models of airborne radar anti-jamming are presented to verify the feasibility and expansibility of our assessment *** group decision-making method shows its effectiveness compared with the experimental evaluation results by the FAHP researched method.

关键词： Evaluation framework small sample data analytical hierarchy process fuzzy Dempstere-Shafer theory.

Air-combat behavior data mining based on truncation method

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Systems Engineering and Electronics 2010年第5期21卷 827-834页

作者： Yunfei Yin guanghong gong Liang Han College of Computer Science Chongqing University Chongqing 400044 E R. China School of Automation Science and Electrical Engineering Beihang University Beijing 100191 P. R. China

This paper considers the problem of applying data mining techniques to aeronautical *** truncation method,which is one of the techniques in the aeronautical data mining,can be used to efficiently handle the air-combat behavior *** technique of air-combat behavior data mining based on the truncation method is proposed to discover the air-combat rules or *** simulation platform of the air-combat behavior data mining that supports two fighters is *** simulation experimental results show that the proposed air-combat behavior data mining technique based on the truncation method is feasible whether in efficiency or in effectiveness.

关键词： air-combat truncation method behavior mining basic fighter maneuvers(BFMs) data mining.

Adaptive Boundary and Semantic Composite Segmentation Method for Individual Objects in Aerial Images

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

Computer Modeling in Engineering & Sciences 2023年第9期136卷 2237-2265页

作者： Ying Li guanghong gong Dan Wang Ni Li School of Automation Science and Electrical Engineering Beihang UniversityBeijing100191China Shenyuan Honors College Beihang UniversityBeijing100191China State Key Laboratory of Virtual Reality Technology and Systems Beijing100191China

There are two types of methods for image *** is traditional image processing methods,which are sensitive to details and boundaries,yet fail to recognize semantic *** other is deep learning methods,which can locate and identify different objects,but boundary identifications are not accurate *** of them cannot generate entire segmentation *** order to obtain accurate edge detection and semantic information,an Adaptive Boundary and Semantic Composite Segmentation method(ABSCS)is *** method can precisely semantic segment individual objects in large-size aerial images with limited GPU *** includes adaptively dividing and modifying the aerial images with the proposed principles and methods,using the deep learning method to semantic segment and preprocess the small divided pieces,using three traditional methods to segment and preprocess original-size aerial images,adaptively selecting traditional results tomodify the boundaries of individual objects in deep learning results,and combining the results of different *** object semantic segmentation experiments are conducted by using the AeroScapes dataset,and their results are analyzed qualitatively and *** experimental results demonstrate that the proposed method can achieve more promising object boundaries than the original deep learning *** work also demonstrates the advantages of the proposed method in applications of point cloud semantic segmentation and image inpainting.

关键词： Semantic segmentation aerial images composite method traditional image processing deep learning

A model of organizational performance based on the big-five factor theory and ABMS method

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

International Journal of Modeling, Simulation, and Scientific Computing 2016年第2期7卷 38-47页

作者： Mo Hao gong guanghong Li Ni Kong Haipeng School of Automation Science and Electrical Engineering Beihang University Beijing 100191P.R.China

How to predict and change organizational performance has been a focus problem drawing economists and managers’attention for a long *** Big-five Factor theory is very popular among psychologists and ABMS(Agent-Based Modeling and Simulation)method is widely used in Systems *** integrating Big-five Factor and ABMS,we proposed a model to predict organizational *** rules were built up to describe one-on-one cooperation in the five dimensions of personalities and an unsymmetrical network model was established to depict cooperation relationships between team *** is more,a series of cases was studied and the result was proved to be rational.

关键词： Organizational performance Big-five Factor ABMS unsymmetrical network model.

Learning robust features by extended generative stochastic networks

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

International Journal of Modeling, Simulation, and Scientific Computing 2018年第1期9卷 135-145页

作者： Da Teng Xiao Song guanghong gong Junhua Zhou School of Automation Science and Electrical Engineering Beihang University P.R.China State Key Laboratory of Intelligent Manufacturing System Technology P.R.China

Deep neural networks have achieved state-of-the-art performance on many object recognition tasks,but they are vulnerable to small adversarial *** this paper,several extensions of generative stochastic networks(GSNs)are proposed to improve the robustness of neural networks to random noise and adversarial *** results show that compared to normal GSN method,the extensions using adversarial examples,lateral connections and feedforward networks can improve the performance of GSNs by making the models more resistant to overfitting and noise.

关键词： Machine learning deep learning neural networks adversarial examples

Boltzmann machines with clusters of stochastic binary units

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

International Journal of Modeling, Simulation, and Scientific Computing 2016年第2期7卷 187-195页

作者： Da Teng Zhang Li guanghong gong Liang Han School of Automation Beihang University Beijing 100191P.R.China School of Information Yunnan University Kunming 650091P.R.China

The original restricted Boltzmann machines(RBMs)are extended by replacing the binary visible and hidden variables with clusters of binary units,and a new learning algorithm for training deep Boltzmann machine of this new variant is *** sum of binary units of each cluster is approximated by a Gaussian *** demonstrate that the proposed Boltzmann machines can achieve good performance in the MNIST handwritten digital recognition task.

关键词： Restricted Boltzmann machines machine learning unsupervised learning