时间测度链上一类二阶动力方程的振动性

研究了时间测度链上的一类具有非线性中立项和变时滞的二阶非线性动力方程的振动性.通过引入参数函数和广义的Riccati变换,并借助时间测度链上的有关理论,得到了方程振动的几个充分条件.所得结果推广和改进了现有文献中相应的结果.     

混沌遗传算法对BP神经网络的改进研究

BP神经网络算法是目前应用最广泛的一种神经网络算法,但有收敛速度慢和易陷入局部极小值等缺陷.本文利用混沌遗传算法(CGA)具有混沌运动遍历性、遗传算法反演性的特性来改进BP神经网络算法.该算法的基本思想是用混沌遗传算法对BP神经网络算法的初始权值和初始阈值进行优化.把混沌变量加入遗传算法中,提高遗传算法的全局搜索能力和收敛速度;用混沌遗传算法优化后得到的最优解作为BP神经网络算法的初始权值和阈值.通过实验观察,改进后的结果与普通的BP神经网络算法的结果相比,具有更高的准确率.     

An open problem on the optimality of an asymptotic solution to Duffing’s nonlinear oscillation problem

Based on the renormalization group method, Kirkinis (2012) [8] obtained an asymptotic solution to Duffing’s nonlinear oscillation problem. Kirkinis then asked if the asymptotic solution is optimal. In this paper, an affirmative answer to the open problem is given by means of the homotopy analysis method.     

又一类三阶中立型分布时滞微分方程的振动定理

本文研究三阶中立型分布时滞微分方程的振动问题.利用Riccati变换技巧和积分平均方法,获得了方程每一解振动或者收敛到零点的新准则,最后,给出了说明所得结果应用的例子.     

Integrated scheduling and batch ordering for construction project

Multi-Mode Resource Constrained Project Scheduling Problem and material batch ordering for construction project are integrated to help project manager consider various trade-offs among several costs, such as renewable resources’ cost, material price, ordering cost, back-ordering cost, inventory holding cost and reward/penalty for early/late project completion. Therefore, we prove a mixed integer programming model and impel to calculate inventory holding cost and back order cost in objective function. Moreover, a hybrid algorithm combined adapted harmony search and genetic algorithm is proposed correspondingly. In order to inherit elitist solution and maintain population’s diversity simultaneously, we add a selection operator when the harmony memory is initialized and modify the replacement operator based on distance. Besides, genetic algorithm is adopted based on a ‘012’ coding scheme. Finally, algorithm and model performance is presented and several project instances are provided with different network structures and realizations to discuss the factors on total cost.     

Group-shop scheduling with sequence-dependent set-up and transportation times

This paper considers a group-shop scheduling problem (GSSP) with sequence-dependent set-up times (SDSTs) and transportation times. The GSSP provides a general formulation including the job-shop and the open-shop scheduling problems. The consideration of set-up and transportation times is among the most realistic assumptions made in the field of scheduling. In this paper, we study the GSSP with transportation and anticipatory SDSTs, where jobs are released at different times and there are several transporters to carry jobs. The objective is to find a job schedule that minimizes the makespan, that is, the time at which all jobs are completed and transported to the warehouse (or to the customer). The problem is formulated as a disjunctive programming problem and then prepared in a form of mixed integer linear programming (MILP). Due to the non-deterministic polynomial-time hardness (NP-hardness) of the GSSP, large instances cannot be optimally solved in a reasonable amount of time. Therefore, a genetic algorithm (GA) hybridized with an active schedule generator is proposed to tackle large-sized instances. Both Baldwinian and Lamarckian versions of the proposed hybrid algorithm are then implemented and evaluated through computational experiments.     

Mathematical analysis and solutions for multi-objective line-cell conversion problem

The line-cell (or line-seru) conversion is an innovation of assembly system applied widely in the electronics industry. Its essence is tearing out an assembly line and adopting a mini-assembly unit, called seru (or Japanese style assembly cell). In this paper, we develop a multi-objective optimization model to investigate two line-cell conversion performances: the total throughput time (TTPT) and the total labor hours (TLH). We analyze the bi-objective model to find out its mathematical characteristics such as solution space, combinatorial complexity and non-convex properties, and others. Owing to the difficulties of the model, a non-dominated sorting genetic algorithm that can solve large size problems in a reasonable time is developed. To verify the reliability of the algorithm, solutions are compared with those obtained from the enumeration method. We find that the proposed genetic algorithm is useful and can get reliable solutions in most cases.     

Hybrid algorithm for a vendor managed inventory system in a two-echelon supply chain

In this paper we address the issue of vendor managed inventory (VMI) by considering a two-echelon single vendor/multiple buyer supply chain network. We try to find the optimal sales quantity by maximizing profit, given as a nonlinear and non-convex objective function. For such complicated combinatorial optimization problems, exact algorithms and optimization commercial software such as LINGO are inefficient, especially on practical-size problems. In this paper we develop a hybrid genetic/simulated annealing algorithm to deal with this nonlinear problem. Our results demonstrate that the proposed hybrid algorithm outperforms previous methodologies and achieves more robust solutions.     

Inflow problem for the one-dimensional compressible Navier–Stokes equations under large initial perturbation

This paper is concerned with the inflow problem for the one-dimensional compressible Navier–Stokes equations. For such a problem, Matsumura and Nishihara showed in [10] that there exists boundary layer solution to the inflow problem, and that both the boundary layer solution, the rarefaction wave, and the superposition of boundary layer solution and rarefaction wave are nonlinear stable under small initial perturbation. The main purpose of this paper is to show that similar stability results for the boundary layer solution and the supersonic rarefaction wave still hold for a class of large initial perturbation which can allow the initial density to have large oscillation. The proofs are given by an elementary energy method and the key point is to deduce the desired lower and upper bounds on the density function.     

Development of the 16 X‐STR loci typing system and genetic analysis in a Shanghai Han population from China

This study developed a new multiplex PCR system that simultaneously amplifies 16 X‐STR loci in the same PCR reaction, and the polymorphism and mutation rates of these 16 X‐STR loci were explored in a Shanghai Han population from China. These loci included DXS10134, DXS10159, DXS6789, DXS6795, DXS6800, DXS6803, DXS6807, DXS6810, DXS7132, DXS7424, DXS8378, DXS9902, GATA165B12, GATA172D05, GATA31E08, and HPRTB. Samples from 591 unrelated individuals (293 males and 298 females) and 400 two‐generation families were successfully analyzed using this multiplex system. Allele frequencies and mutation rates of the 16 loci were investigated, with the comparison of allele frequency distributions among different populations performed. Polymorphism information contents of these loci were all >0.6440 except the locus DXS6800 (0.4706). Nine cases of mutations were detected in the 16 loci from the investigation of 9232 meioses. Pairwise comparisons of allele frequency distributions showed significant differences for most loci among populations from different countries and ethnic groups but not among the Han population living in other areas of China. These results suggest that the 16 X‐STR loci system provides highly informative polymorphic data for paternity testing and forensic identification in the Han population in Shanghai, China, as a complementary tool.