Construction of some hypergroups from combinatorial structures

Jajcay‘s studies( 1993 ; 1994) on the automorphism groups of Cayley maps yielded a new product of groups, which he called, rotary product. Using this product, we define a hyperoperation ⊙ on the group Syme (G) , the stabilizer of the identity e ∈ G in the group Sym (G) . We prove that ( Syme (G) , ⊙) is a hypergroup and characterize the subhypergroups of this hypergroup.Finally, we show that the set of all subhypergroups of Syme ( G ) constitute a lattice under ordinary join and meet and that the minimal elements of order two of this lattice is a subgroup of Aut (G) .     

Pre-selecting young bulls before progeny-testing with marker-assisted BLUP

Suppose that a QTL affecting milk production trait or dairy cattle has been identified,and the relative distances between the QTL and the markers are hnown without error.The relative efficiency of pre-selection of young bulls before progeny-testing with marker-assisted BLUP(MBLUP) over conventional progeny-testing(i.e.random selection before progeny-testing)for five generations was studied using Monte Carlo simulation.Different numbers of QTL alleles(two or infinite),initial frequencies of the favorable QTL allele(0.1 and 0.5,when the QTL has two alleles),distances between the adjacent markers(5 cM or 20 cM),and ratios of the QTL variance to the genetic variance(10%,25%,and 50%)were considered in the simulation.The results showed that the relative efficiency of MBLUP was higher when the marker distance was 5 cM than that when the distance was 20 cM.The larger the QTL alleic effect,the lower the MBLUP efficiency.The relative superiority of MBLUP was higher when the QTL had two alleles and the frequency of the favorable QTL allele was low in the base population.The relative superiority of MBLUP decreased with th progress of generation in all cases and the decreasing rate was getting higher with the increase of marker density and QTL effect.     

拟阵交约束分划

给出了用求量大重量拟阵交算法来求定形拟阵交２－分划问题的最优解的一个充分条件。     

一类Cayley图的边Hamilton性

设G为有限群,|G|=p3,p为素数,M是G的一个生成集.证明了p3阶的Cayley图X(G,M)是边-Hamilton图.     

粒子群优化算法在脑部肿瘤图像分割中的应用

粒子群优化算法能选择适当的适应度函数,使每组粒子群根据相应的适应值搜索到最佳聚类中心,改善了FCM算法和K-means算法的不足,具有适应性强,实时性好,受噪声影响小等特点。本文将其应用于脑部肿瘤图像的分割,结果表明,粒子群聚类算法是一种很有潜力的图像分割方法。     

以培养学生创新能力为目标的计算机教学改革的思考

建立创新型国家,需要培养大批的创新型人才.只有加强培养大学生的创新能力,学生参加工作才具有可持续发展的能力.如何培养大学生具有较强实践和创新能力,必须改革传统的计算机教学模式,强化实验环节,积极参与课外科技活动,培养学生构建计算机应用系统的能力.     

改进微粒群算法及其在生料浆调配多目标优化中的应用

微粒群算法是一种简单易行、收敛快速的演化计算方法，但也存在早熟收敛的缺陷。鉴于惯性权值对控制算法收敛所起的作用，本文提出一种改进的微粒群算法，让惯性权值按类S型曲线形式随迭代次数非线性变化，并根据微粒当前的搜索状态自适应调节该曲线的曲率，有效保证算法收敛速度和全局寻优能力。改进的微粒群算法用于求解氧化铝生料浆调配过程0-1组合优化问题，与传统微粒群算法及遗传算法比较的结果表明，改进的微粒群算法具有较强的全局搜索能力且能有效避免早熟收敛问题。     

图像增强技术中的智能算法

提出了一种基于粒子群与模拟退火算法相结合的进化算法.该算法利用模拟退火算法全局收敛性好和粒子群算法收敛速度快等优点，通过交换这两种算法的信息得到最优解.将这种新算法应用于灰度图像的自适应增强，实例计算表明该算法稳定性好，在收敛速度和求解精度方面都优于遗传算法等一些其它进化算法.     

有界算子扰动C0-半群增长阶的判定

文章研究Hilbert空间中具有增长ω0的C0-半群（T（t））t≥0，在有界算子B扰动后所成半群（S（t））t≥0的增长阶ω1大于或小于给定常数ω的充分条件。     

大型地下洞室群地质信息三维可视化分析与应用

提出采用三维地质建模技术进行大型地下洞室群地质信息可视化分析的研究方法，构建了模拟真实地质形态的地下洞室群地质三维可视化模型，直观准确地描述了工程所处的地质条件和地质环境。基于所建立的地质模型，可实现与地下洞室群相关的多项地质信息的三维可视化分析。工程应用表明该方法可为应对复杂地质条件下地下洞室群设计与施工中的地质问题提供一种有效的分析手段。