基于ISSA-SVM的钻井卡钻事故预测

为预防钻井过程中卡钻事故的发生,通过提出了一种改进麻雀搜索算法(improved sparrow search algorithm, ISSA)优化支持向量机(support vector machines, SVM)的预测模型方法(ISSA-SVM),在发现者位置更新公式中引入一种改进的自适应非线性惯性递减权重;在警戒者位置更新公式中引入莱维飞行策略。利用主成分分析法(principal component analysis, PCA)对外国某大型油田的实测钻井数据进行降维处理,并利用惩罚参数和核参数进行卡钻事故的预测。实验结果表明：ISSA-SVM的预测准确率高达85.185 2%,且收敛速度更快,可见ISSA-SVM可有效预测钻井卡钻事故。     

WSN中传感器节点的弹性神经网络任务分配方法

为解决WSN多目标跟踪节点任务分配的竞争冲突问题,提出一种融合了模糊聚类的多弹性子模自组织神经网络节点任务分配方法.通过模糊聚类估计目标数量,建立节点任务分配跟踪精度和能量消耗的综合性能指标,采用非全连接的环形弹性结构自组织神经网络优化监测联盟,用最近邻法对神经元弹性子模进行初始化,根据胜者为王原则动态调整子模的感受域,以快速锁定最优监测联盟,实现多目标的精确跟踪.实验结果表明:文中方法能有效解决多目标跟踪节点任务分配的竞争冲突问题,以及竞争冲突时的系统能耗增加与实时性问题;在随机均匀部署节点拓扑和目标直线运动模式下,文中方法的能耗较最近邻法降低了48.2%~55.9%,较未改进弹性神经网络法降低了37.4%~42.5%,且运算速度提高了19.0%~27.4%.     

基于改进RPCA的连铸过程数据可靠性研究

连铸过程数据的可靠性是连铸二冷动态控制系统成功应用的前提。针对正常连铸冷却过程慢时变、非稳态和多输入多输出特性,提出了基于改进递归主元分析方法监控连铸二冷配水过程数据的可靠性,这其中包括了用于建模的协方差阵、主元和两种报警控制限的递归更新,并通过小波除噪消除样本数据的噪声和奇异点。现场数据仿真结果表明应用此方法进行连铸过程监控可以明显减少误报的发生。     

曲线分类建模方法及其在多地区GDP预测中的应用

传统方法解决大规模时序曲线的预测建模问题,需要对每条曲线逐一建模,这使得建模工作量相当庞大,在实际应用中缺乏可操作性.为此,提出了一种解决此问题的新方法--曲线分类建模方法.该方法先减少曲线的模型种类,再进行曲线分类和分类建模,在尽可能保留原始信息的前提下大大降低了建模的工作量.文中阐述了该方法的原理和计算过程,并应用于多国家和地区GDP曲线的预测案例,说明该方法的实用性和有效性.     

Rough function model and rough membership function

Two pairs of approximation operators, which are the scale lower and upper approximations as well as the real line lower and upper approximations, are defined. Their properties and antithesis characteristics are analyzed. The rough function model is generalized based on rough set theory, and the scheme of rough function theory is made more distinct and complete. Therefore, the transformation of the real function analysis from real line to scale is achieved. A series of basic concepts in rough function model including rough numbers, rough intervals, and rough membership functions are defined in the new scheme of the rough function model. Operating properties of rough intervals similar to rough sets are obtained. The relationship of rough inclusion and rough equality of rough intervals is defined by two kinds of tools, known as the lower （upper） approximation operator in real numbers domain and rough membership functions. Their relative properties are analyzed and proved strictly, which provides necessary theoretical foundation and technical support for the further discussion of properties and practical application of the rough function model.     

复杂性研究的计算方法

计算、算法与复杂性研究有着密切的关系,计算方法已经上升为一种复杂性研究的科学方法。计算复杂性使用的研究方法基本上是计算方法,算法复杂性研究的算法描述也是一种计算方法。计算方法还在复杂适应系统理论、人工生命等复杂性分支学科中扮演了重要的角色。计算方法不但在复杂性研究中有重要的作用,而且在一般方法论意义上也已经变得不可或缺,甚至走向了计算主义。     

蒸发波导环境中基于ZMNL的海杂波仿真

针对海上蒸发波导环境中雷达目标的识别,提出了蒸发波导中基于零记忆非线性变换(ZMNL)的海杂波实现方法。该方法确定了海杂波K分布参数与蒸发波导高度间的关系,可得到不同蒸发波导高度、不同探测距离处的相关K分布杂波序列。给出了应用该方法产生杂波的具体步骤,仿真实验证明了该方法用于蒸发波导中海杂波建模的有效性。     

Robust output regulation problem for discrete-time linear systems with both input and communication delays

Recently, the robust output regulation problem for continuous-time linear systems with both input and communication time-delays was studied. This paper will further present the results on the robust output regulation problem for discrete-time linear systems with input and communication delays. The motivation of this paper comes from two aspects. First, it is known that the solvability of the output regulation problem for linear systems is dictated by two matrix equations. While, for delay-free systems, these two matrix equations are same for both continuous-time systems and discrete-time systems, they are different for continuous-time time-delay systems and discrete-time time-delay systems. Second, the stabilization methods for continuous-time timedelay systems and discrete-time time-delay systems are also somehow different. Thus, an independent treatment of the robust output regulation problem for discrete-time time-delay systems will be useful and necessary.     

New computing method of weighted coefficients for tracking a maneuvering target using PDAF in the presence of clutter

To avoid or reduce the influence of unpredictable motion mode on data association, a new computing method of weighted coefficients of measurements for PDAF is presented in which it is assumed that the current turn rate of a maneuvering target changes within a limited range and its turn may be in arbitrary direction during data association. Thus, the predicted center for computing the weighted coefficients is a curved surface in 3-D space, which differs from the predicted center for setting up a validation gate, namely, a point in 3-D space. The distance between a measurement and the curved surface is used to compute its weighted coefficient. To reduce the computational complexity of weighted coefficients, the formulas for computing the maneuvering direction angle and turn rate corresponding to a measurement are presented. Simulation results show the proposed method reduces the percentage of lost tracks and improves the state estimation accuracy in tracking a maneuvering target using PDAF in the presence of clutter.     

A differential game model for a developing and a developed country

A differential game (DG) model for a developing and a developed country is considered. Each player makes decisions about how much resource to be used to restrict the opponent's development so as to maximize his weighted sum of current consumption and final output. Current consumption is assumed to be preferred to final output for both players. The developing country is assumed to have a higher economic growth rate and a higher preference to final output, whereas the developed country is assumed to have a higher initial income and a higher efficiency in restricting his opponent. This problem is investigated under three kinds of information structures, i.e., a zerosum, a nonzero-sum, and a Stackelberg game. Open-loop equilibrium solutions are obtained for all the three cases. Economic implications of the result are provided. The research is supported by the National Natural Science Foundation of China under Grant Nos. 70771118 and 70371030 and the Scientific Research Foundation for the Returned Overseas Chinese Scholars, State Education Ministry under Grant No. 2006.331.