检索结果-南通市图书馆

编队航天器协同绕飞非合作目标的全驱预设性能控制

在线全文

学校读者我要写书评

暂无评论

航空学报 2024年第1期45卷 79-93页

作者：殷泽阳邢友朋韩飞魏才盛廖宇新中南大学自动化学院长沙410083 上海航天控制技术研究所上海201109

针对利用多星编队协同绕飞空间非合作目标的控制问题,考虑外部干扰和目标非期望轨道机动影响,提出了一种基于全驱系统理论的分布式约定时间预设性能控制方法。首先,基于视线坐标系建立服务航天器编队绕飞非合作目标的相对运动方程,并将... 详细信息

针对利用多星编队协同绕飞空间非合作目标的控制问题,考虑外部干扰和目标非期望轨道机动影响,提出了一种基于全驱系统理论的分布式约定时间预设性能控制方法。首先,基于视线坐标系建立服务航天器编队绕飞非合作目标的相对运动方程,并将其转化为全驱系统形式以降低控制器设计复杂度;其次,结合拓扑理论设计一种分布式约定时间预设性能控制器,保证多星编队绕飞的瞬态和稳态控制性能;此外,引入扩张状态观测器观测并合理补偿系统中存在的不确定项与干扰,提高多星编队绕飞的控制精度与鲁棒性。最后,通过两组绕飞仿真验证了所设计控制方法的有效性。

关键词：全驱系统理论编队绕飞非合作目标预设性能控制扩张状态观测器

在线全文

学校读者我要写书评

暂无评论

面向密集场景的空天地网络资源分配算法

电子与信息学报 2024年第5期46卷 1968-1976页

作者：张鸿廖彧歆王汝言吴大鹏杜慧敏重庆邮电大学通信与信息工程学院重庆400065 先进网络与智能互联技术重庆市高校重点实验室重庆400065 泛在感知与互联重庆市重点实验室重庆400065

空天地网络具有覆盖范围大、吞吐量高、弹性强等优点。该文针对大量用户并发接入、网络负载不均衡所引发的网络拥塞、服务质量恶化等问题,提出一种面向密集场景的资源分配算法。首先以用户需求为中心,根据不同类型用户任务的偏好来构建... 详细信息

空天地网络具有覆盖范围大、吞吐量高、弹性强等优点。该文针对大量用户并发接入、网络负载不均衡所引发的网络拥塞、服务质量恶化等问题,提出一种面向密集场景的资源分配算法。首先以用户需求为中心,根据不同类型用户任务的偏好来构建用户效用函数,然后基于匹配博弈的网络选择算法和结合对偶上升法的功率控制算法来实现负载均衡,优化资源分配方案。实验表明,相较于传统策略,所提策略整体用户接入率至少提高35%,时延和吞吐量方面性能提升超过50%;在密集场景下,能更有效地均衡负载,提升网络性能。

关键词：空天地一体化网络资源分配博弈论密集场景

NiTi形状记忆合金的功能特性及其应用发展

在线全文

学校读者我要写书评

暂无评论

材料工程 2024年第2期52卷 60-77页

作者：杨超廖雨欣卢海洲颜安蔡潍锶李鹏旭华南理工大学国家金属材料近净成形工程技术研究中心广州510640 广东技术师范大学机电学院广州510665

NiTi形状记忆合金(shape memory alloys,SMAs)作为一种智能材料,具有良好的超弹性、形状记忆效应和生物相容性等功能特性,被广泛应用于航空航天、医疗器械和工程建筑等领域。其中超弹性在宏观上表现为发生较大的变形仍能恢复原形状,且... 详细信息

NiTi形状记忆合金(shape memory alloys,SMAs)作为一种智能材料,具有良好的超弹性、形状记忆效应和生物相容性等功能特性,被广泛应用于航空航天、医疗器械和工程建筑等领域。其中超弹性在宏观上表现为发生较大的变形仍能恢复原形状,且其远大于常见金属可恢复的弹性应变。形状记忆效应则是温度激励下奥氏体和马氏体两相的相互转变,根据宏观变形分为单程、双程和全程形状记忆效应。而NiTi SMAs的生物相容性体现在低弹性模量和低生物毒性等方面,可应用于正畸、矫正、心血管支架等医疗器件。为充分发挥NiTi SMAs的功能,研究者们不断开发NiTi SMAs相关的智能结构。本文简要综述了近年来研究和发展NiTi SMAs的不同功能特性及其对应的智能结构典型应用,详细介绍和讨论了NiTi SMAs的功能特性、关注问题和应用领域。同时,也对NiTi SMAs阻尼性能和储氢特性进行了阐述。最后,展望了NiTi SMAs在各领域应用上尚需重点关注的问题:利用增材制造技术调控微观结构实现超弹性的稳定性提升;通过建立本构模型为形状记忆效应的稳定应用提供理论指导,并进一步优化结构实现形状记忆效应的宏观放大;提高NiTi SMAs在生物环境里的耐腐蚀性和医疗应用推广。因此,推动NiTi SMAs在不同应用领域的个性化和功能定制化,尚需大量的跨学科研究。

关键词： NiTi形状记忆合金功能特性智能结构

Cooperative fault detection and recovery in the GNSS positioning of mobile agent swarms based on relative distance measurements

同方期刊数据库博看期刊评论

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Aeronautics 2022年第5期35卷 129-144页

作者： Qifeng CHEN Yunhe MENG Song LI yuxin liao School of Aeronautics and Astronautics Central South UniversityChangsha 410083China School of Physics and Astronomy Sun Yat-sen UniversityZhuhai 519082China

Relative measurements are exploited to cooperatively detect and recover faults in the positioning of Mobile Agent(MA)Swarms(MASs).First,a network vertex fault detection method based on edge testing is *** each edge,a property that has a functional relationship with the properties of its two vertices is measured and *** on the edge testing results of the network,the maximum likelihood principle is used to identify the vertex fault ***,an edge distance testing method based on the noncentral chi-square distribution is developed for detecting faults in the Global Navigation Satellite System(GNSS)positioning of ***,a recovery strategy for faults in the positioning of MASs based on distance measurement is *** effectiveness of the proposed methods is validated by a simulation case in which an MAS passes through a GNSS spoofing *** proposed methods are conducive to increasing the robustness of the positioning of MASs in complex *** main novelties include the following:(A)network vertex fault detection is based on concrete probability analysis rather than simple majority voting,and(B)the relation of detectability and recoverability of MAS positioning faults with the structure of the relative measurement network is first disclosed.

关键词： Fault detection and recovery Global Navigation Satellite System Mobile agents Network topology Positioning

深度强化学习求解移动机器人端到端导航问题的研究综述

在线全文

学校读者我要写书评

暂无评论

计算机工程与应用 2024年第14期60卷 1-13页

作者：何丽姚佳程廖雨鑫张文智卢赵清袁亮肖文东新疆大学智能制造现代产业学院(机械工程学院) 乌鲁木齐830017 北京化工大学信息科学与技术学院北京100029

自主导航是移动机器人完成复杂任务的前提和基础,传统的自主导航系统依赖于地图的精度,无法适应高度复杂的作业和服务场景。移动机器人不依赖先验地图信息,通过深度强化学习与环境交互学习能够自主决策的端到端导航方法成为新的研究热... 详细信息

自主导航是移动机器人完成复杂任务的前提和基础,传统的自主导航系统依赖于地图的精度,无法适应高度复杂的作业和服务场景。移动机器人不依赖先验地图信息,通过深度强化学习与环境交互学习能够自主决策的端到端导航方法成为新的研究热点。大多数现有的分类方法不能全面地总结端到端导航问题的挑战和机遇,根据端到端导航系统的特点,将导航问题的挑战归结为导航智能体感知能力差、学习效率低和导航策略泛化能力弱等关键问题,阐述了端到端导航系统的研究现状和发展趋势,分别详细介绍了近年来针对这些关键问题的代表性研究成果,并对其优势和不足进行了归纳总结。最后,从视觉语言导航、多智能体协同导航、融合超分辨率重建图像的端到端导航和可解释性端到端导航等方面展望了移动机器人端到端导航的未来发展趋势,为移动机器人端到端导航的研究和应用提供一定的思路。

关键词：端到端导航深度强化学习感知能力学习效率泛化能力

同方期刊数据库

Defect-designed Mo-doped BiVO_(4)photoanode for efficient photoelectrochemical degradation of phenol

在线全文

学校读者我要写书评

暂无评论

Journal of Materials Science & Technology 2023年第34期165卷 225-234页

作者： Mufeng Yu Yiming Tang yuxin liao Wanhan He Xin-xin Lu Xin Li School of Environment Guangdong Provincial Key Laboratory of Chemical Pollution and Environmental Safety&MOE Key Laboratory of Theoretical Chemistry of EnvironmentSouth China Normal UniversityGuangzhou 510006China PetroChina Shenzhen New Energy Research Institute Shenzhen 518052China Institute of Biomass Engineering Key Laboratory of Energy Plants Resource and UtilizationMinistry of Agriculture and Rural AffairsSouth China Agricultural UniversityGuangzhou 510642China

The monoclinic scheelite BiVO_(4)has impressive properties such as a narrow energy band gap,exceptional stability,and extended absorption in visible light,making it a suitable ***,the BiVO_(4)material encounters challenges such as the high recombination rate of photogenerated electronhole pairs and poor photoelectron conductivity,which limits photocatalytic *** address this problem,we developed Mo-doped BiVO_(4)films on FTO substrates for photoelectrocatalytic degradation of *** exposed to visible light,the Mo-BiVO_(4)film attained a 70%degradation of phenol in 120 min with a 1.2 V ***/AgCl bias—a 3.7 times improvement from pristine BiVO_(4).Mo-doping facilitates better migration and separation of electron-hole pairs and increases the concentration of photogenerated carriers,leading to an upward shift of the valence band potential direction,and an improvement in oxidation ***,density-functional theory(DFT)calculations were used to explain how Modoping with BiVO_(4)improves the adsorption energy to phenol degradation intermediates,emphasizing its effectiveness in promoting phenol ***,with the inclusion of DFT calculations,this work provides a more comprehensive understanding of the mechanism underlying the enhancement of photocatalytic activity by Mo-doped BiVO_(4),which is crucial information for the further development of effective and efficient photoanodes.

关键词： BiVO_(4) Doping Photoelectrocatalysis Phenol DFT calculations

Nest complexity correlates with larger brain size but smaller body mass across bird species

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Integrative Zoology 2024年第3期19卷 496-504页

作者： Shaobin LI yuxin LIU Xiaolong DU Guopan LI Wenbo liao College of Life Science Yangtze UniversityJingzhouChina Key Laboratory of Southwest China Wildlife Resources Conservation(Ministry of Education) China West Normal UniversityNanchongChina

Amniotes differ substantially in absolute and relative brain size after controlling for allometry,and numerous hypotheses have been proposed to explain brain size *** size is thought to correlate with processing capacity and the brain’s ability to support complex manipulation such as nest-building *** increased complexity of nest structure is supposed to be a measure of an ability to manipulate nesting material into the required *** degree of nest-structure complexity is also supposed to be associated with body mass,partly because small species lose heat faster and delicate and insulated nests are more crucial for temperature control of eggs during incubation by small ***,we conducted comparative analyses to test these hypotheses by investigating whether the complexity of species-typical nest structure can be explained by brain size and body mass(a covariate also to control for allometric effects on brain size)across 1353 bird species from 147 *** with these hypotheses,our results revealed that avian brain size increases as the complexity of the nest structure increases after controlling for a significant effect of body size,and also that a negative relationship exists between nest complexity and body mass.

关键词： body size brain size comparative analysis nest building nest type reproductive investment

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

基于深度强化学习的变外形飞行器姿态控制

航天控制 2024年第2期42卷 55-61页

作者：丁溶曹承钰李繁飙谢启超廖宇新中南大学自动化学院长沙410083

针对外界干扰和模型不确定性影响下的高超声速变外形飞行器,提出了一种基于深度强化学习的多通道端到端姿态控制方法。首先,建立了高超声速变外形飞行器姿态控制模型;其次,将飞行器姿态控制问题转化为马尔科夫决策过程;再者,基于双延迟... 详细信息

针对外界干扰和模型不确定性影响下的高超声速变外形飞行器,提出了一种基于深度强化学习的多通道端到端姿态控制方法。首先,建立了高超声速变外形飞行器姿态控制模型;其次,将飞行器姿态控制问题转化为马尔科夫决策过程;再者,基于双延迟深度确定性策略梯度算法开展了智能体训练,并将端到端生成的飞行控制指令进行了在线部署;最后,通过基本性能仿真和适应性仿真验证了所提出方法的有效性和泛化能力。

关键词：高超声速飞行器变外形飞行器深度强化学习姿态控制

Indirect Radau pseudospectral method for the receding horizon control problem

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Aeronautics 2016年第1期29卷 215-227页

作者： liao yuxin Li Huifeng Bao Weimin School of Astronautics Beihang University Science and Technology Committee China Aerospace Science and Technology Corporation

To solve the receding horizon control （RHC） problem in an online manner, a novel numerical method called the indirect Radau pseudospectral method （IRPM） is proposed in this paper. Based on calculus of variations and the first-order necessary optimality condition, the RHC problem for linear time-varying （LTV） system is transformed into the two-point boundary value problem （TPBVP）. The Radau pseudospectral approximation is employed to discretize the TPBVP into well-posed linear algebraic equations. The resulting linear algebraic equations are solved via a matrix partitioning approach afterwards to obtain the optimal feedback control law. For the nonlinear system, the linearization method or the quasi linearization method is employed to approximate the RHC problem with successive linear approximations. Subsequently, each linear problem is solved via the similar method which is used to solve the RHC problem for LTV system. Simulation results of three examples show that the IRPM is of high accuracy and of high compu- tation efficiency to solve the RHC problem and the stability of closed-loop systems is guaranteed.

关键词： Indirect Radau pseudospec-tral method Linear time-varying system Nonlinear system Receding horizon control Two-point boundary valueproblem