一种自适应人工蜂群算法求解U型顺序相依拆卸线平衡问题

产品拆卸过程中零部件之间会相互干扰影响任务作业时间,基于该情形构建了多目标U型SDDLBP优化模型,并提出一种自适应ABC算法。所提算法设计了自适应动态邻域搜索方法,以提高局部开发能力;采用了轮盘赌与锦标赛法结合的分段选择法,以有效评价并选择蜜源进行深度开发;建立了基于当前最优解的变异操作,以提高全局探索能力快速跳出局部最优。最后,通过算例测试和实例分析验证算法的高效性。     

A coefficient identification problem for a mathematical model related to ductal carcinoma in situ

We consider a coefficient identification problem for a mathematical model with free boundary related to ductal carcinoma in situ (DCIS). This inverse problem aims to determine the nutrient consumption rate from additional measurement data at a boundary point. We first obtain a global‐in‐time uniqueness of our inverse problem. Then based on the optimization method, we present a regularization algorithm to recover the nutrient consumption rate. Finally, our numerical experiment shows the effectiveness of the proposed numerical method.     

基于实际路网情境的配送车辆调度优化

在实际路网情境下结合车道数、车道宽度、路口信号灯设置等路网物理特性,构建了考虑综合交通阻抗的多车型车辆调度模型,提出了两阶段求解策略:第1阶段设计了改进A-star精确解算法用于计算客户时间距离矩阵;第2阶段针对实际路网的特征设计了混合模拟退火算法求解调度方案。以大连市某配送中心运营实例进行路网情境仿真试验,结果表明:改进A-star算法较改进Dijkstra算法具有更短的路径搜索时间;混合模拟退火算法求解结果较实际调度方案优化了13.1% 的综合成本;路网增流、区域拥堵和路段禁行三类路网情境均能对配送方案的车辆配置、路径选择、客户服务次序、作业时间和违约费用等5方面内容产生干扰,调度计划的制定需要详细考虑这些因素的变化。     

Modified two-grid method for solving coupled Navier-Stokes/Darcy model based on Newton iteration

A new decoupled two-gird algorithm with the Newton iteration is proposed for solving the coupled Navier-Stokes/Darcy model which describes a fluid flow filtrating through porous media. Moreover the err...     

vdW-TSSCDS—An automated and global procedure for the computation of stationary points on intermolecular potential energy surfaces

We present a generalization of the transition state search using chemical dynamics simulations (TSSCDS) methodology (discussed in a previous study) which allows the topographical characterization of intermolecular potential energy surfaces (IPES) for non-covalently bound complexes (vdW-TSSCDS). Starting from a single random input geometry, we show that vdW-TSSCDS is able to globally and automatically locate stationary points of an IPES, even in limiting cases such as extremely flat regions or nontrivial topologies (eg, bifurcation points). The basic idea is the expression of the connectivity matrix in block structure, where diagonal blocks correspond to the isolated fragments and off-diagonal blocks provide the intermolecular connectivity. To this end, we introduce a new definition of bound or not, in a non-covalent sense, utilizing an extra set of van der Waals distances, which encompasses all kinds of non-covalent distances. To discuss the use of the vdW-TSSCDS method, we present a series of 2-body van der Waals systems, namely, Ar-Benzene (3D), N₂-Benzene (6D) and H₂O-Benzene (9D). Finally, we further illustrate its capabilities by presenting some applications for n-body problems (n > 2), (H₂O)₂-Benzene (12D) and (H₂O)₃-Benzene (21D), as well as to a reactive, fully-flexible, system (Benzene-NO₂)⁺ (39D) in which the simultaneous breaking/formation of both covalent and non-covalent interactions takes place.     

Sparse Multicategory Generalized Distance Weighted Discrimination in Ultra-High Dimensions

Distance weighted discrimination (DWD) is an appealing classification method that is capable of overcoming data piling problems in high-dimensional settings. Especially when various sparsity structures are assumed in these settings, variable selection in multicategory classification poses great challenges. In this paper, we propose a multicategory generalized DWD (MgDWD) method that maintains intrinsic variable group structures during selection using a sparse group lasso penalty. Theoretically, we derive minimizer uniqueness for the penalized MgDWD loss function and consistency properties for the proposed classifier. We further develop an efficient algorithm based on the proximal operator to solve the optimization problem. The performance of MgDWD is evaluated using finite sample simulations and miRNA data from an HIV study.     

Learning extended tree augmented naive structures

This work proposes an extended version of the well-known tree-augmented naive Bayes (TAN) classifier where the structure learning step is performed without requiring features to be connected to the class. Based on a modification of Edmonds' algorithm, our structure learning procedure explores a superset of the structures that are considered by TAN, yet achieves global optimality of the learning score function in a very efficient way (quadratic in the number of features, the same complexity as learning TANs). We enhance our procedure with a new score function that only takes into account arcs that are relevant to predict the class, as well as an optimization over the equivalent sample size during learning. These ideas may be useful for structure learning of Bayesian networks in general. A range of experiments shows that we obtain models with better prediction accuracy than naive Bayes and TAN, and comparable to the accuracy of the state-of-the-art classifier averaged one-dependence estimator (AODE). We release our implementation of ETAN so that it can be easily installed and run within Weka.