Research on knowledge acquisition method about the IF/THEN rules based on rough set theory

The basic principles of IF/THEN rules in rough set theory are analyzed first, and then the automatic process of knowledge acquisition is given. The numerical data is qualitatively processed by the classification of membership functions and membership degrees to get the normative decision table. The regular method of relations and the reduction algorithm of attributes are studied. The reduced relations are presented by the multi-representvalue method and its algorithm is offered. The whole knowledge acquisition process has high degree of automation and the extracted knowledge is true and reliable.     

直接力控制导弹的拟人智能自动驾驶仪设计

拟人智能控制是一种基于被控对象的物理模型以及人类相应的知识模拟人类处理控制问题的思路和一般方法解决问题的控制方法。基于拟人智能控制思想提出一种针对采用直接力控制导弹的姿态控制的控制逻辑和非线性控制律的设计方法。控制逻辑决定了气动力与直接力两种控制方式的权限分配。然后,利用拟人思想设计定性控制律。利用广义归约的方法将复杂的导弹姿态控制问题分解成若干个可直接解决的本原问题,然后依据各本原问题的主次关系以及耦合关系通过分析导弹的飞行动力学模型设计出非线性控制律结构,最后利用遗传算法将控制律中的各权值优化定量得到定量控制律。所设计出的控制逻辑和控制律在导弹进行大过载机动时保证导弹的稳定性。仿真结果表明复合控制方法可以使导弹具有更高的精度。     

考虑政府奖励机制的集群企业隐性知识传播模型研究

目的 隐性知识传播是企业获取隐性知识的有效途径,其对于提高国民经济及其自身发展十分重要。方法 首先,在均匀网络和非均匀网络中分别构建考虑政府奖励机制的集群企业隐性知识传播模型;其次,对模型的稳定性进行分析并求出传播阈值;最后,运用Matlab软件对模型参数进行数值仿真分析。结论 考虑政府奖励机制的传播率和初始传播者数量与隐性知识的传播成正比;网络结构对集群企业间隐性知识传播有显著影响,平均度越大的网络越有利于隐性知识的传播。     

Central electron-enriched NO-FeN4 sites as superior acidic oxygen reduction reaction electrocatalysts for proton exchange membrane fuel cells

Developing nobel-metal-free catalysts, especially for iron-nitrogen on carbon (FeNC) materials, has been an urgent demand for wide applications of proton exchange membrane fuel cells (PEMFCs). However, the inferior oxygen reduction reaction (ORR) activity of traditional iron-nitrogen sites in acidic conditions seriously impedes the further improvement of their performance. Herein, we synthesized FeN₄ with NO (nitric oxide) group axial modification (denoted as NO-FeN₄) on a large scale through a confined small molecule synthesis strategy. Benefitting from the strong electron-withdrawing effect of the NO group, the central electron-rich FeN₄ site exhibits ultrahigh ORR activity with a three times higher mass activity (1.1 A·g?1 at 0.85 V) compared to the traditional FeN₄ sample, as well as full four-electron reaction selectivity. Moreover, the PEMFC assembled with the as-prepared electrocatalyst also exhibits a greatly enhanced peak power density (>725 mW·cm?2). This work provides a new approach to rationally design advanced M-N_x nonnoble electrocatalysts for the ORR.     

苦草对沉积物-水界面不同形态砷的削减

以富营养化底泥为研究对象,从环境微界面的角度出发,采用室内培养试验研究了常见沉水植物苦草对沉积物间隙水中As(Ⅲ)和As(Ⅴ)的迁移转化影响及环境微界面行为,并评估了苦草对沉积物-水界面中As的去除效果。试验结果表明：苦草增加了沉积物中溶解氧的渗透深度,降低了pH值、有机质(UV₂₅₄)以及溶解态As、As(Ⅴ)、As(Ⅲ)、溶解态Fe(Ⅱ)和溶解态Mn的质量浓度;3种形态的As(溶解态As、As(Ⅴ)、As(Ⅲ))与溶解态Mn和UV₂₅₄显著相关,与溶解态Fe(Ⅱ)相关性较差;苦草对As(Ⅲ)的去除效果(69.5%)优于其对As(Ⅴ)的去除效果(64.6%)。     

工艺参数对楔横轧二次楔轧制超大断面收缩率轴类件的影响

为考察楔横轧二次楔轧制在超大断面收缩率情况下的成形规律,在H630轧机上进行了不同工艺条件下的超大断面收缩率轧制实验,得到各工艺参数对轧制质量的影响规律,发现在超大断面收缩率情况下二次楔的工艺参数对内部缺陷和拉断加工界限有决定性影响.结果表明,只要参数得当,二次楔轧制可以获得合格的大断面收缩率轴类件.     

Simulation experiments on the reaction system of CH4-MgSO4-H2O

H2S-rich gas in carbonate reservoirs is usually attributed to thermochemical sulfate reduction （TSR）. In this paper, thermal simulation experiments on the reaction system of CH4-MgSO4-H2O were carried out using autoclave at 425℃--525℃. The threshold temperature for initiating TSR is much lower than our previous studies （550℃）. Properties of the reaction products were analyzed by microcoulometry, gas-chromatography （GC）, Fourier transform-infrared spectrometry （FT-IR） and X-ray diffraction （XRD） methods. Thermodynamics and reaction kinetics of TSR processes were investigated on the basis of the experimental data. The results show that thermochemical reduction of magnesium sulfate with methane can proceed spontaneously to produce magnesium oxide, hydrogen sulfur, and carbon dioxide as the main products, and high temperature is thermodynamically favorable to the reaction. Ac- cording to the reaction model, the calculated activation energy of TSR is 101.894 kJ/mol, which is lower than that by most previous studies. Mg^2＋ may have played a role of catalytic action in the process of TSR. The elementary steps of TSR and reaction mechanism were discussed tentatively. The study can provide important information on the explanation of geochemical depth limit for natural gas and on the generation of high H2S gas in deep carbonates reservoirs.     

一种基于属性重要性的属性约简启发式算法

属性约简是粗糙集理论研究中的关键问题之一.文中定义了一种新的属性重要性度量准则,克服了多值偏向性问题,并给出一种新的属性约简算法.该算法以核属性集为初始约简集合,以新的属性重要性度量准则为启发信息,通过逐步加入相对于决策而言重要的条件属性来求取最小约简.实例分析表明该算法是有效的.     

一种新的粗糙集与支持向量分类机算法

用支持向量机的机器学习是依据结构风险最小化原则,序列最小优化(SMO)是较特殊的分解算法。对高维大样本对象,支持向量机训练算法面临耗时增大与维数灾问题,利用粗糙集(RS)对不确定数据处理能力,提出一种新的粗糙集与支持向量分类机算法RS-SMO,可以对数据集做属性约简,生成类边界集作为SMO的训练子集,比原始训练集的维数与规模大小都有一定程度的减少,可构造出具有较好时空性能的算法。用两个实用数据对象做仿真,实验结果表明算法RS-SMO比SMO的性能有大的提高,实现了结构风险最小化。     

分辨矩阵构成与约简同步的方法

针对粗集理论中分辨矩阵方法的缺陷,分析了这种方法3个过程的时间-空间性能,提出了被析取的矩阵元素不参与分辨矩阵的构成,即分辨矩阵的构成与与属性约简同步的方法,减少分辨矩阵的元素数和中间数据占用的磁盘空间,提高了数据挖掘的速度.通过实验验证了该方法时间-空间性能上的优越性.