一适用于超多类手写汉字识别的新改型Adaboost算法

提出一种适用于超多类手写汉字识别的新改型Adaboost算法,采用基于描述性模型的多类分类器（modified quadratic discriminant function,MQDF）作为Adaboost基元分类器,可直接进行多类分类,无需将多类问题转化为多个两类问题处理,其训练复杂度大大低于已有的多类Adaboost算法。算法提出根据广义置信度更新样本权重,实验证明这种算法适用于大规模多类分类问题。为了降低算法的识别复杂度,提出从所有训练后得到的Adaboost基元分类器组中选择一个最优的基元分类器作为最终分类器的方法进行删减。在HCL2000及THOCR-HCD数据集上进行实验证明,所提改型Adaboost算法提高了识别率的有效性,该算法的相对错误率比现有最优算法分别下降了14.3 %,8.1 %和19.5 %。     

粮食产量影响因素的灰色关联分析及其贡献率比较

运用灰色关联分析的方法,通过对省域(河南省)、市域(安阳市)、县域(滑县)三级粮食产量影响因素(播种面积、有效灌溉面积、农村用电、农用化肥、农用柴油、农药)的灰色关联分析及其贡献率比较,结果表明:有效灌溉面积对市域、省域两级粮食产量的影响均居第一位,农用化肥施用量对县域、市域两级粮食产量的影响均居第二位,农药与其它5个因素相比,对省、市、县三级粮食产量的影响相同,都是最小.农用柴油、农村用电、农用化肥、农药对县、市两级粮食产量的贡献率均居第一、第二、第三、第四位,化肥和农药对省、市、县三级粮食产量的贡献率都据第三和第四位.在此基础上,为不同区域尺度的粮食生产提供了决策依据.表5.参11.     

Performance comparison of distributed state estimation algorithms for power systems

A newly proposed distributed dynamic state estimation algorithm based on the maximum a posteriori (MAP) technique is generalised and studied for power systems. The system model involves linear time-varying load dynamics and nonlinear measurements. The main contribution of this paper is to compare the performance and feasibility of this distributed algorithm with several existing distributed state estimation algorithms in the literature. Simulations are tested on the IEEE 39-bus and 118-bus systems under various operating conditions. The results show that this distributed algorithm performs better than distributed quasi-steady state estimation algorithms which do not use the load dynamic model. The results also show that the performance of this distributed method is very close to that by the centralized state estimation method. The merits of this algorithm over the centralized method lie in its low computational complexity and low communication load. Hence, the analysis supports the efficiency and benefits of the distributed algorithm in applications to large-scale power systems.     

Spectral Properties and Positivity of Flows in Infinite Networks with Dynamic Boundary Control Nodes

In this paper we study transport processes on infinite networks with dynamic boundary control nodes. These flows can be modeled by operator semigroups on a suitable Banach space. Using functional analytical and graph theoretical methods, we investigate its spectral properties of the system and positivity of the semigroup under appropriate assumptions on the network.     

Robust H∞ filtering for discrete-time systems with Markovian switching and time-delays

The robust H∞ filtering problem for uncertain discrete-time Markovian jump linear systems with modedependent time-delays is investigated. Attention is focused on designing a Markovian jump linear filter that ensures robust stochastic stability while achieving a prescribed H∞ performance level of the resulting filtering error system, for all admissible uncertainties. The key features of the approach include the introduction of a new type of stochastic Lyapunov functional and some free weighting matrix variables. Sufficient conditions for the solvability of this problem are obtained in terms of a set of linear matrix inequalities. Numerical examples are provided to demonstrate the reduced conservatism of the proposed approach.     

A multiyear lags input-holding-output model on education with excluding idle capital

This paper develops a multi-year lag Input-Holding-Output (I-H-O) Model on education with exclusion of the idle capital to address the reasonable education structure in support of a sustainable development strategy in China. First, the model considers the multiyear lag of human capital because the lag time of human capital is even longer and more important than that of fixed capital. Second, it considers the idle capital resulting from the output decline in education, for example, student decrease in primary school. The new generalized Leonitief dynamic inverse is deduced to obtain a positive solution on education when output declines as well as expands. After compiling the 2000 I-H-O table on education, the authors adopt modifications-by-step method to treat nonlinear coefficients, and calculate education scale, the requirement of human capital, and education expenditure from 2005 to 2020. It is found that structural imbalance of human capital is a serious problem for Chinese economic development. This research was supporting by the Key Project of National Natural Science Foundation of China under Grant No. 60874119, Innovation Funds of Chinese Academy of Sciences, Research Fund for Humanities and Social Sciences in colleges under Grant No. 06BJY102, Key project in Jiangxi Soft Science Project under Grant No. [2006]188, Jiangxi Research Fund for Humanities and Social Sciences in colleges under Grant No. 03JJ01, and Henan Research Fund for Humanities and Social Sciences in colleges under Grant No. 2008-ZD-002.     

Design and performance analysis of networked control systems with random delay

The design and performance analysis of networked control systems with random network delay in the forward channel is proposed, which are described in a state-space form. A new control scheme is used to overcome the effects of network transmission delay, which is termed networked predictive control (NPC). Furthermore, three different ways to choose control input are discussed and the performances are analyzed, respectively. Both real-time simulations and practical experiments show the effectiveness of the control scheme.     

离散T-S模糊广义系统稳定性的一个新判别条件

利用离散广义Lyapunov方法和分段模糊Lyapunov函数,研究输入采用双交叠模糊分划的离散广义模糊系统的性质,得出新的判定离散T-S模糊广义系统稳定性的充分条件,即在最大交叠规则组内寻求公共正定矩阵,并用算例说明这种方法的有效性.     

基于时间优化的设备群维持随机模型

对于在工人完成任务效率具有随机性的情况下,如何合理组织一类可拆分设备维修以加快设备维修速度,从而增强企业竞争优势的问题,建立了基于时间优化的随机维修模型并对该模型进行讨论,最终形成了一个与之等价的可以进行求解的确定型模型。通过数值算例发现,在考虑随机性对维修效率的影响后,与认为工人完成任务效率固定的确定型模型相比,任务安排可能发生变化。     

运行状态逐渐恶化的同类设备流维修管理模型

大量同类设备的维修问题具有广泛的应用背景,这样的设置具有运行状态逐渐恶化的性质。虽然有许多文献对这类设备维修问题进行研究,但很少文献在建模中联合考虑其他同类设备运行状态及维修资源有限的情况。Venkatesh等人在这方面进行了开拓性的工作。然而他们只是对一个简单的设备流维修管理问题(FMP)问题进行了研究,仍然存在不足。本文尝试将问题进行扩展,讨论了一类具有多个运作方和维修方的设备流维修优化问题。通过仿真实验,将设备流维修管理与传统的基于条件的设备维修管理进行了比较。结果显示,本文提出的设备流维修策略可以获得更多收益。