Optimality and convergence of adaptive optimal control by reinforcement synthesis

Adaptive Optimal Control (AOC) by reinforcement synthesis is proposed to facilitate the application of optimal control theory in feedback controls. Reinforcement synthesis uses the critic–actor architecture of reinforcement learning to carry out sequential optimization. Optimality conditions for AOC are formulated using the discrete minimum principle. A proof of the convergence conditions for the reinforcement synthesis algorithm is presented. As the final time extends to infinity, the reinforcement synthesis algorithm is equivalent to the Dual Heuristic dynamic Programming (DHP) algorithm, a version of approximate dynamic programming. Thus, formulating DHP with the AOC approach has rigorous proofs of optimality and convergence. The efficacy of AOC by reinforcement synthesis is demonstrated by solving a linear quadratic regulator problem.     

HUMAN LIMITATIONS AND VEHICLE DESIGN∗

This paper contains a review of research on road vehicle design in relation to human physiological, anthropometric and psychological factors     

A nontrivial factor in determining current distribution in an ac HTS cable-proximity effect

A superconductor has zero resistance at the superconducting state. This unique property creates many exceptional phenomena, of which some are known and the others are not. Our experiments with multilayer high temperature superconductor (HTS) cable samples revealed a new phenomenon that alternating current had a tendency to flow in the inner and outer layers of the cables. We attribute the cause of this phenomenon to the electromagnetic interaction in an infinite electrical conductivity medium and term it su...     

A nonlinear analytical model for the squeeze-film dynamics of parallel plates subjected to axial flow

In this paper, we develop a theoretical model to predict the nonlinear fluid-structure interaction forces and the dynamics of parallel vibrating plates subjected to an axial gap flow. The gap is assumed small, when compared to the plate dimensions, the plate width being much larger than the length, so that the simplifying assumptions of 1D bulk-flow models are adequate. We thus develop a simplified theoretical squeeze-film formulation, which includes both the distributed and singular dissipative flow terms. This model is suitable for performing effective time-domain numerical simulations of vibrating systems which are coupled by the nonlinear unsteady flow forces, for instance the vibro-impact dynamics of plates with fluid gap interfaces. A linearized version of the flow model is also presented and discussed, which is appropriate for studying the complex modes and linear stability of flow/structure coupled systems as a function of the average axial gap velocity. Two applications of our formulation are presented: (1) first we study how an axial flow modifies the rigid-body motion of immersed plates falling under gravity; (2) then we compute the dynamical behavior of an immersed oscillating plate as a function of the axial gap flow velocity. Linear stability plots of oscillating plates are shown, as a function of the average fluid gap and of the axial flow velocity, for various scenarios of the loss terms. These results highlight the conditions leading to either the divergence or flutter instabilities. Numerical simulations of the nonlinear flow/structure dynamical responses are also presented, for both stable and unstable regimes. This work is of interest to a large body of real-life problems, for instance the dynamics of nuclear spent fuel racks immersed in a pool when subjected to seismic excitations, or the self-excited vibro-impact motions of valve-like components under axial flows.     

流形学习算法分析及在人脸数据库上的应用

为了更高效的处理高维数、高复杂性的非线性数据,发现其嵌入在源数据空间中的本维特征,提出了基于局部光滑逼近思想的流形学习算法,通过局部线性误差逼近最小化,实现将高维数据映射到低维空间.在FREY人脸数据库上进行降维实验,证明了该方法的可行性和有效性.     

InRout - A QoS aware route selection algorithm for industrial wireless sensor networks

Wireless sensor networks are a key enabling technology for industrial monitoring applications where the use of wireless infrastructure allows high adaptivity and low cost in terms of installation and retrofitting. To facilitate the move from the current wired designs to wireless designs, concerns regarding reliability must be satisfied. Current standardization efforts for industrial wireless systems lack specification on efficient routing protocols that mitigate reliability concerns. Consequently, this work presents the InRout route selection algorithm, where local information is shared among neighbouring nodes to enable efficient, distributed route selection while satisfying industrial application requirements and considering sensor node resource limitations. Route selection is described as a multi-armed bandit task and uses Q-learning techniques to obtain the best available solution with low overhead. A performance comparison with existing approaches demonstrates the benefits of the InRout algorithm, which satisfies typical quality of service requirements for industrial monitoring applications while considering sensor node resources. Simulation results show that InRout can provide gains ranging from 4% to 60% in the number of successfully delivered packets when compared to current approaches with much lower control overhead.     

机器学习系统毒化攻击综述

自机器学习被应用到许多关键性领域以来,机器学习系统的脆弱性也引起了人们的高度重视。其中,针对机器学习系统的毒化攻击得到了研究者的广泛关注,呈现了一些研究成果。因此,将系统地介绍当前机器学习系统毒化攻击的研究进展,对机器学习系统毒化攻击算法进行分类和总结,包括针对机器学习中的线性分类器、支持向量机、贝叶斯分类器和深度神经网络等几类常见模型的毒化攻击等攻击算法,目标是使现有的关于机器学习系统毒化攻击的研究成果更加清晰,为相关研究者的研究工作提供启发。     

抽灌井群不同布井模式热交互性研究

通过建立地下水流与温度的耦合模型,模拟分析抽灌井群在不同布控模式下地下温场的热交互性,预测地下水温度的变化趋势,进而提出合理的布井方案。模拟软件选用FlUENT6.3.26,通过不同的布井方案对地下水的热交互性进行比较分析,结果表明,抽灌井间距对地下水温度及交互性影响较大,同类井间距对其影响不显著。抽灌井数量越多热交互性越弱,为了延缓热贯通发生的时间,解决水的回灌不良的问题,可在抽水井不变的情况下适当增加回灌井数量。     

Floating-Gate Circuits for Adaptation of Saccadic Eye Movement Accuracy

In this paper we describe an analog VLSI circuit, fabricatedusing a standard 2 µm, n-well, BiCMOS process,which utilizes floating-gate structures for non-volatile, on-chip,analog parameter storage. This circuit is designed to operatein the context of a hardware model of the primate oculomotorsystem and performs visually-guided, saccadic adaptation. Thechip contains a one-dimensional array of photoreceptors and floating-gatecircuits which are used to map retinal positions to motor outputcommands. The system's functionality is demonstrated by trainingthe chip with several different mapping functions using a supervised-learningtechnique.     

建立自然型实时语音讨论系统的双令牌方法

数字语音通信服务是ＣＳＣＷ应用系统的重要组成部分。本文首先分析了自然型语音通信所需遵循的规则，建议使用双令牌来控制发言权，并描述了一个基于双令牌的发言控制协议。在此基础上，我们给出了一个自然型多点语音讨论系统ＶｏｉｃｅＣｈａｔ的设计和实现方案，最后结合某段典型的对话情景给出了应用实例