检索结果-南通市图书馆

Policy iteration based q.learning for linear nonzero-sum q.adratic differential games

在线全文

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2019年第5期62卷 195-213页

作者： Xinxing lI Zhihong PENg li lIANg Wenzhong zha School of Automation Beijing Institute of Technology State Key laboratory of Intelligent Control and Decision of Complex System Information Science Academy China Electronics Technology Group Corporation

In this paper, a policy iteration-based q.learning algorithm is proposed to solve infinite horizon linear nonzero-sum q.adratic differential games with completely unknown dynamics. The q.learning algorithm, which employs off-policy reinforcement learning(Rl), can learn the Nash eq.ilibrium and the corresponding value functions online, using the data sets generated by behavior policies. First, we prove eq.ivalence between the proposed off-policy q.learning algorithm and an offline PI algorithm by selecting specific initially admissible polices that can be learned online. Then, the convergence of the off-policy q.earning algorithm is proved under a mild rank condition that can be easily met by injecting appropriate probing noises into behavior policies. The generated data sets can be repeatedly used during the learning process, which is computationally effective. The simulation results demonstrate the effectiveness of the proposed q.learning algorithm.

关键词： adaptive dynamic programming ADP q.learning reinforcement learning Rl linear nonzerosum q.adratic differential games policy iteration PI off-policy

Microenvironment characterization for a novel biocompatible hybrid carbon-silicon material

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China Chemistry 2009年第8期52卷 1120-1127页

作者： SUI XiHua zha q.ngFang WU MingBo gUO YanSheng State Key laboratory of Heavy Oil China University of PetroleumDongying 257061China

The compatible carbon-silicon complex materials originated from precursor diglycerylsilane(DgS) and sugar-modified silane N-(3-triethoxysilylpropyl)gluconamide(glS) have gained substantial popularity by demonstrating admirable properties to stabilize entrapped *** microenvironment inside these materials,especially the distribution of sugar moieties inside the matrix,which is likely the most critical factor determining compatibility of these materials,still remains *** deeply investigate the biocompatibility mechanism of these materials,we have adopted two different preparation routes for these materials by introducing glS into the starting DgS sol stage,but things are different after the DgS gel is formed.A fluorescence probe rhodamine 6g is introduced herein in the DgS sol to monitor the distribution of glS moieties,as well as the evolution of the microenvironment inside resulting *** in all,the findings demonstrated that the timing of glS addition plays a critical role in controlling the evolution of the inner structure of materials,suggesting that this factor provides a promising route to tune the properties of the resulting materials.

关键词： biocompatibility carbon-silicon complex microenvironment characterization

Intelligent Fast Cell Association Scheme Based on Deep q.learning in Ultra-Dense Cellular Networks

在线全文

学校读者我要写书评

暂无评论

China Communications 2021年第2期18卷 259-270页

作者： Jinhua Pan lusheng Wang Hai lin Zhiheng zha Caihong Kai Key laboratory of Knowledge Engineering with Big Data Ministry of Education.School of Computer Science and Information EngineeringHefei University of TechnologyHefei 230601China Anhui Province Key laboratory of Industry Safety and Emergency Technology Hefei 230601China Key laboratory of Aerospace Information Security and Trusted Computing Ministry of Education.School of Cyber Science and EngineeringWuhan UniversityWuhan 430072China

To support dramatically increased traffic loads,communication networks become *** cell association(CA)schemes are timeconsuming,forcing researchers to seek fast *** paper proposes a deep q.learning based scheme,whose main idea is to train a deep neural network(DNN)to calculate the q.values of all the state-action pairs and the cell holding the maximum q.value is *** the training stage,the intelligent agent continuously generates samples through the trial-anderror method to train the DNN until *** the application stage,state vectors of all the users are inputted to the trained DNN to q.ickly obtain a satisfied CA result of a scenario with the same BS locations and user *** demonstrate that the proposed scheme provides satisfied CA results in a computational time several orders of magnitudes shorter than traditional ***,performance metrics,such as capacity and fairness,can be guaranteed.

关键词： ultra-dense cellular networks(UDCN) cell association(CA) deep q.learning proportional fairness q.learning

A Refined Regularity Criterion for 3D liq.id Crystal Eq.ations Involving Horizontal Velocity

在线全文

学校读者我要写书评

暂无评论

Journal of Mathematical Research with Applications 2021年第6期41卷 607-614页

作者： Xiaoli CHEN Xiujuan zha School of Mathematics and Statistics Jiangxi Normal University

This note investigates the global regularity of 3D liq.id crystal eq.ations in terms of the vertical derivative of uh. More precisely, we prove that if the vertical derivative of the horizontal velocity component uh satisfies ?3uh∈ lp（0, T; R3） with 2/p+3/q.3/2, 2 ≤ p ≤∞, then the local strong solution（u, d） can be smoothly extended beyond t = T.

关键词： liq.id crystal ow regularity criterion local strong solution

Evaluation of Forest Damaged Area and Severity Caused by Ice-snow Frozen Disasters over Southern China with Remote Sensing

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Chinese geographical Science 2019年第3期29卷 405-416页

作者： WANg Xuecheng YANg Fei gAO Xing WANg Wei zha Xinjie Institute of geographic Sciences and Natural Resources Research Chinese Academy of Sciences State Key Laboratory of Resources and Environmental Information System University of Chinese Academy of Sciences Jiangsu Center for Collaborative Innovation in geographical Information Resource Development and Application

The accurate assessment of forest damage is important basis for the forest post-disaster recovery process and ecosystem management. This study evaluates the spatial distribution of damaged forest and its damaged severity caused by ice-snow disaster that occurred in southern China during January 10 to February 2 in 2008. The moderate-resolution imaging spectroradiometer(MODIS)13 q. products are used, which include two vegetation indices data of NDVI(Normalized Difference Vegetation Index) and EVI(Enhanced Vegetation Index). Furtherly, after q.ality Screening(q.) and Savizky-golay(S-g) filtering of MODIS 13 q. data, four evaluation indices are obtained, which are NDVI with q.(q.NDVI), EVI with q.(q.EVI), NDVI with S-g filtering(SgNDVI) and EVI with S-g filtering(SgEVI). The study provides a new way of firstly determining the threshold for each image pixel for damaged forest evaluation, by computing the pre-disaster reference value and change threshold with vegetation index from remote sensing data. Results show obvious improvement with the new way for forest damage evaluation, evaluation result of forest damage is much close to the field survey data with standard error of only 0.95 and 1/3 less than the result that evaluated from other threshold method. Comparatively, the q.NDVI shows better performance than other three indices on evaluating forest damages. The evaluated result with q.NDVI shows that the severe, moderate, mild damaged rates of Southern China forests are 47.33%, 34.15%, 18.52%, respectively. By analyzing the influence of topographic and meteorological factors on forest-vegetation damage, we found that the precipitation on freezing days has greater impact on forest-vegetation damage, which is regarded as the most important factor. This study could be a scientific and reliable reference for evaluating the forest damages from ice-snow frozen disasters.

关键词： ice-snow disaster vegetation index forest remote sensing southern China

Force-controlled Compensation Scheme for P-q.Valve-controlled Asymmetric Cylinder used on Hydraulic q.adruped Robots

在线全文

学校读者我要写书评

暂无评论

Journal of Bionic Engineering 2020年第6期17卷 1139-1151页

作者： Yapeng Shi Mantian li Fusheng zha lining Sun Wei guo Cong Ma Zhibin li State Key laboratory of Robotics and System Harbin Institue of TechnologyHarbin 150001China Shenzhen Academy of Aerospace Technology China Aerospace ScienceShenzhen 518057China School of Informatics The University of EdinburghEdinburgh EH89ABUnited Kingdom

Under the req.irement of the force controller of hydraulic q.adruped robots,the goal of this work is to accurately track the force commands at the level of the hydraulic drive *** main contribution focuses on the development of a force-controlled compensation scheme,which is specifically aimed at the key issues affecting the hydraulic q.adrupedal *** this idea,based on a P-q.valve-controlled asymmetric cylinder,we first establish a mathematical model for the hydraulic drive unit force control *** the desired force commands,a force feed-forward algorithm is presented to improve the dynamic performance of the ***,we propose a disturbance compensation algorithm to reduce the influence induced by external disturbances due to foot-ground ***,combining with a variable gain PI controller,a series of experiments are implemented on a force control performance test platform to verify the proposed *** results demonstrate that the force-controlled compensation scheme has the ability to notably improve the force tracking accuracy,reduce the response time and redundant force.

关键词： hydraulic drive unit force tracking P-q.valve-controlled asymmetric cylinder hydraulic q.adruped robot

Effects of heterogeneity distribution on hillslope stability during rainfalls

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

Water Science and Engineering 2016年第2期9卷 134-144页

作者： Jing-sen Cai E-chuan Yan Tian-chyi Jim Yeh Yuan-yuan zha Faculty of Engineering China University of Geosciences Department of Hydrology and Water Resources University of Arizona

The objective of this study was to investigate the spatial relationship between the most likely distribution of saturated hydraulic conductivity(Ks) and the observed pressure head(P) distribution within a hillslope. The cross-correlation analysis method was used to investigate the effects of the variance of ln Ks, spatial structure anisotropy of ln Ks, and vertical infiltration flux(q. on P at some selected locations within the hillslope. The cross-correlation analysis shows that, in the unsaturated region with a uniform flux boundary, the dominant correlation between P and Ksis negative and mainly occurs around the observation location of P. A relatively high P value is located in a relatively low Kszone, while a relatively low P value is located in a relatively high Kszone. generally speaking, P is positively correlated with q.Ksat the same location in the unsaturated region. In the saturated region, the spatial distribution of Kscan significantly affect the position and shape of the phreatic surface. We therefore conclude that heterogeneity can cause some parts of the hillslope to be sensitive to external hydraulic stimuli(e.g., rainfall and reservoir level change), and other parts of the hillslope to be insensitive. This is crucial to explaining why slopes with similar geometries would show different responses to the same hydraulic stimuli, which is significant to hillslope stability analysis.

关键词： Cross-correlation analysis Heterogeneity Hillslope stability Saturated hydraulic conductivity Stochastic conceptualization Pore-water pressure

EFFECTS OF OXIDES IN INVESTMENT MOlD ON SURFACE PROPERTIES OF TITANIUM CASTINgS

在线全文

学校读者我要写书评

暂无评论

Acta Metallurgica Sinica(English letters) 2000年第2期13卷 746-752页

作者： H.g. Wang, X.M. q., J. Tian, W.q. zha, g.S. liang, Y. Y. Chen and J. Jia National Key laboratory of Precision Hot Processing of Metals, Harbin Institute of Technology, Harbin 150001, China

Effects of different facing oxides, including including yttria colloidal and powders (Y/Y), yttria stabilized zirconia colloidal and powders (ZY/ZY), zirconia colloidal and powders (Z/Z) and zirconia colloidal and zireonite powders (Z/ZS) on the q.alities of investment castings are studied. The outward appearance, microstructures, and microhardness profiles of castings made of commercial pure titanium, Ti--Al--V and Ti--Al--Zr are investigated. Castings made by shells with yttria as face materials have the least contamination and by zirconite shells have the most contamination. Thermodynamic of metal-mold reactions is also taken into consideration.

关键词： titanium alloy investment casting oxide in investment mold

Minimax q.learning design for H_(∞) control of linear discrete-time systems

在线全文

学校读者我要写书评

暂无评论

Frontiers of Information Technology & Electronic Engineering 2022年第3期23卷 438-451页

作者： Xinxing lI lele XI Wenzhong zha Zhihong PENg Information Science Academy China Electronics Technology Group CorporationBeijing 100086China School of Automation Beijing Institute of TechnologyBeijing 100081China Peng Cheng laboratory Shenzhen 518052China

The H_(∞)control method is an effective approach for attenuating the effect of disturbances on practical systems, but it is difficult to obtain the H_(∞)controller due to the nonlinear Hamilton-Jacobi-Isaacs eq.ation, even for linear systems. This study deals with the design of an H_(∞)controller for linear discrete-time systems. To solve the related game algebraic Riccati eq.ation(gARE), a novel model-free minimax q.learning method is developed, on the basis of an offline policy iteration algorithm, which is shown to be Newton’s method for solving the gARE. The proposed minimax q.learning method, which employs off-policy reinforcement learning, learns the optimal control policies for the controller and the disturbance online, using only the state samples generated by the implemented behavior policies. Different from existing q.learning methods, a novel gradient-based policy improvement scheme is proposed. We prove that the minimax q.learning method converges to the saddle solution under initially admissible control policies and an appropriate positive learning rate, provided that certain persistence of excitation(PE)conditions are satisfied. In addition, the PE conditions can be easily met by choosing appropriate behavior policies containing certain excitation noises, without causing any excitation noise bias. In the simulation study, we apply the proposed minimax q.learning method to design an H_(∞)load-freq.ency controller for an electrical power system generator that suffers from load disturbance, and the simulation results indicate that the obtained H_(∞)load-freq.ency controller has good disturbance rejection performance.

关键词： H_(∞)control Zero-sum dynamic game Reinforcement learning Adaptive dynamic programming Minimax q.learning Policy iteration