基于全景量测的聚合电网概念、调控方法及其应用

随着可再生能源发电大规模并入电网发输配用各个部分,为使源网荷在频率限定范围内最大限度地消纳可再生能源有功发电,需对其运行调控实施全局统筹。为此,在电网全景过程化量测条件下给出聚合电网调控方法。提出聚合电网、可观测点、可观测点的观测函数和调控函数的概念,在此基础上,给出聚合电网性能指标,提出可观测点、可观测点的观测函数和调控函数之间的互动关系,实现聚合电网渐进调控机制。在线调控算例对所提方法进行了验证。     

Admission control with elastic QoS for video on demand systems

In network service systems, satisfying quality of service (QoS) is one of the main objectives. Admission control and resource allocation strategy can be used to guarantee the QoS requirement. Based on partially observable Markov decision processes (POMDPs), this paper proposes a novel admission control model for video on demand (VOD) service systems with elastic QoS. Elastic QoS is also considered in resource allocation strategy. Policy gradient algorithm is often available to find the solution of POMDP problems, with a satisfactory convergence rate. Through numerical examples, it can be shown that the proposed admission control strategy has better performance than complete admission control strategy.     

几种求取电路最小测试集算法的比较研究

近年来发展的离散事件系统（DES）为数模混合电路中数字信号和模拟信号提供了一种统一的测试方法,而求取电路的最小测试集一直是该研究领域的重点和难点。本文阐述了目前学者提出的几种求取最小测试集算法的基本思想,分析了各算法的优缺点,并深入地比较了各个算法,通过具体实验说明了各算法的优劣性,最后指出了该学科今后的发展方向。     

求解POMDP的动态合并激励学习算法

把POMDP作为激励学习（Reinforcement Leaming）问题的模型,对于具有大状态空间问题的求解有比较好的适应性和有效性。但由于其求解的难度远远地超过了一般的Markov决策过程（MDP）的求解,因此还有许多问题有待解决。该文基于这样的背景,在给定一些特殊的约束条件下提出的一种求解POMDP的方法,即求解POMDP的动态合并激励学习算法。该方法利用区域的概念,在环境状态空间上建立一个区域系统,Agent在区域系统的每个区域上独自并行地实现其最优目标,加快了运算速度。然后把各组成部分的最优值函数按一定的方式整合,最后得出POMDP的最优解。     

动物双歧杆菌V9亚慢性毒性研究

目的研究动物双歧杆菌V9的亚慢性毒性,提出其无可见有害作用水平(NOAEL)。方法 80只断乳SD大鼠按体重随机分成4组,即对照组和3个受试物组,每组20只,雌雄各半。低、中、高剂量分别为1.25、2.50、5.00 g/kg BW。各剂量组灌胃给予受试物,对照组灌同体积的蒸馏水。动物单笼喂养,自由饮食,每周记录大鼠进食量,每周称2次体重以调整灌胃量,连续观察90 d,于实验中期(46 d)和实验结束采血测各项血液学和临床生化指标,并进行病理组织学检查。结果各受试物剂量组动物体重、进食量、食物利用率、临床生化指标、各脏器脏体比与对照组比较均无显著性差异;病理指标也未见由受试物引起的异常改变。血常规指标中有显著性差异的检测值均在本室历史正常对照检测范围内,无生物学意义。结论根据本次实验结果确定动物双歧杆菌V9 NOAEL在雌雄大鼠均为5.00 g/kg BW,即本实验的高剂量,相当于人体推荐量的300倍。     

Access and Sharing Contents Through the Social Network: A POMDP Approach

Users interact in the social network by exchanging useful information. The huge amount of traffic generated requires the design of an accurate model for good management. Hence, knowing the consumers' access patterns is of a great interest for content owners. We propose a study of the access and sharing patterns based on the content's popularity. We suppose that popularity is inferred by the observation of the number of views. Our goal is twofold. On the one hand, we aim to help users select the optimal action that allows content owners to decide to change or not the used social network to another to increase their profits and, on the other hand, to assist consumers, according to their area of interest, to decide to access or not a posted content. We establish a threshold structure of the optimal policies based on a trade‐off study between profits (money, access to useful information, etc.) and costs (time spent to edit or consult content, etc.). The experimental results of the proposed analytical model show that both owners and consumers maximize their utility by choosing the best strategies.     

Application of various observational techniques to the characterization of MX in F82H

Reduced activation ferritic/martensitic steels (RAFMs) have been developed for the structural material of fusion reactor blanket. Extensive researches were conducted using transmission electron microscopy (TEM) to define the precipitation behavior of MX (M: Metallic elements, X: C and/or N) precipitates such as vanadium and tantalum carbides and/or nitrides, since they strongly affect material properties. However, it is not completely defined yet. TEM can obtain detailed information, but the observable volume is limited and it needs a long time to obtain representative and statistically reliable microstructural information. It is fruitful to complement TEM results applying other characterization techniques together. In this work, we studied the precipitation behavior of MX in the RAFM, F82H, through the application of various techniques with larger observable volumes and compared with TEM past results. We additionally tried to find sub-nanometric MX precipitates that can be hardly observed through TEM. From the result, the TaX (X: C and/or N) phase surely varied depending on the chemical composition of alloy and heat treatment; some TaX precipitates were unstable during tempering, MX precipitates were intergranular, and sub-nanometric MX were not found in the matrix. These obtained results were helpful to define the precipitation behavior of MX in F82H.     

Planning for robotic exploration based on forward simulation

We address the problem of controlling a mobile robot to explore a partially known environment. The robot’s objective is the maximization of the amount of information collected about the environment. We formulate the problem as a partially observable Markov decision process (POMDP) with an information-theoretic objective function, and solve it applying forward simulation algorithms with an open-loop approximation. We present a new sample-based approximation for mutual information useful in mobile robotics. The approximation can be seamlessly integrated with forward simulation planning algorithms. We investigate the usefulness of POMDP based planning for exploration, and to alleviate some of its weaknesses propose a combination with frontier based exploration. Experimental results in simulated and real environments show that, depending on the environment, applying POMDP based planning for exploration can improve performance over frontier exploration.