基于字符分割和LeNet-5网络的字符验证码识别

为了解决传统验证码识别方法效率低,精度差的问题,设计了一种先分割后识别的验证码处理方案。该方案在预处理阶段用中值滤波去噪,再利用霍夫变换对图像字符进行矫正;在字符分割阶段,利用垂直投影算法确定验证码字符块个数,以及字符坐标点,再用颜色填充算法对验证码进行初步分割,根据分割后的字符块数量对粘连字符进行二次分割;在识别阶段,我们对LeNet-5网络进行了改进,修改了输入层,并用全连接层替换了LeNet-5网络中的C5层,以此来对验证码字符进行识别;实验表明,对于非粘连验证码和粘连验证码,单张图片分割时间为0.14和0.15ms,分割准确率为98.75%和97.25%,识别准确率为99.99%和97.7%;结果表明,该算法对验证码分割和识别都有着很好的效果。     

基于HAGA的D2D-NOMA资源分配优化算法

针对无线系统带宽资源有限、基站负载压力大、传输时延长等问题,提出一种基于非正交多址接入技术的D2D系统吞吐量最大化资源分配算法。在不同用户的服务质量约束条件下,建立D2D系统吞吐量最大化资源分配模型。该模型的优化目标是一个混合整数非线性规划问题,将其解耦为信道匹配与功率分配2个子问题并分别进行处理,利用自适应惩罚函数法处理约束条件并提出一种基于爬山策略的自适应遗传算法以对问题进行求解。仿真结果表明,与GA、AGA算法相比,该算法能够有效提高D2D系统的吞吐量,且收敛性能更好。     

Efficient mutual nearest neighbor query processing for moving object trajectories

Given a set D of trajectories, a query object q, and a query time extent Γ, a mutual (i.e., symmetric) nearest neighbor (MNN) query over trajectories finds from D, the set of trajectories that are among the k₁ nearest neighbors (NNs) of q within Γ, and meanwhile, have q as one of their k₂ NNs. This type of queries is useful in many applications such as decision making, data mining, and pattern recognition, as it considers both the proximity of the trajectories to q and the proximity of q to the trajectories. In this paper, we first formalize MNN search and identify its characteristics, and then develop several algorithms for processing MNN queries efficiently. In particular, we investigate two classes of MNN queries, i.e., MNN_P and MNN_T queries, which are defined with respect to stationary query points and moving query trajectories, respectively. Our methods utilize the batch processing and reusing technology to reduce the I/O cost (i.e., number of node/page accesses) and CPU time significantly. In addition, we extend our techniques to tackle historical continuous MNN (HCMNN) search for moving object trajectories, which returns the mutual nearest neighbors of q (for a specified k₁ and k₂) at any time instance of Γ. Extensive experiments with real and synthetic datasets demonstrate the performance of our proposed algorithms in terms of efficiency and scalability.     

基于改进遗传算法的TD-SCDMA扇区优化

获得满意的覆盖是TD-SCDMA标准网络规划的重点,为此,在遗传算法的基础上提出基于PCCPCH RSCP覆盖和PCCPCH C/I覆盖的改进遗传算法。采用符合网络多扇区多参数性质的二进制矩阵式编码,为加快收敛速度,给出基于块的交叉方法,同时为防止过早的局部收敛,对适应度函数进行尺度变换。实验结果表明,该算法可以快速收敛到目标解,PCCPCH RSCP覆盖比例和PCCPCH C/I覆盖比例均有不同程度的提升。     

抗干扰的运动目标检测算法

针对改进族群进化算法(EGEA)中对族群模式的有效采样问题,提出以竞争指数为选择指标的常规选择、以族群为选择单位的常规选择以及基于竞争指数的模拟退火排序选择3种选择机制,并通过多维函数优化问题进行实验分析,结果表明3种选择机制都能够在一定程度上提高EGEA的搜索效率,且基于竞争指数的模拟退火排序选择算子的实现更简便,对EGEA综合性能的提高也更明显。     

蜂群算法在装备维修任务调度中的应用

如何在最短的时间内将损坏的装备修复好,是装备维修决策的重要内容之一。鉴于蜂群算法在任务调度特别是动态随机任务调度中的优势,将蜂群算法引入装备维修任务调度研究。测试案例的仿真结果表明,蜂群算法对于动态随机任务调度具有较强的优势。     

Segmentation of historical machine-printed documents using Adaptive Run Length Smoothing and skeleton segmentation paths

In this paper, we strive towards the development of efficient techniques in order to segment document pages resulting from the digitization of historical machine-printed sources. This kind of documents often suffer from low quality and local skew, several degradations due to the old printing matrix quality or ink diffusion, and exhibit complex and dense layout. To face these problems, we introduce the following innovative aspects: (i) use of a novel Adaptive Run Length Smoothing Algorithm (ARLSA) in order to face the problem of complex and dense document layout, (ii) detection of noisy areas and punctuation marks that are usual in historical machine-printed documents, (iii) detection of possible obstacles formed from background areas in order to separate neighboring text columns or text lines, and (iv) use of skeleton segmentation paths in order to isolate possible connected characters. Comparative experiments using several historical machine-printed documents prove the efficiency of the proposed technique.     

Hard multidimensional multiple choice knapsack problems, an empirical study

Recent advances in algorithms for the multidimensional multiple choice knapsack problems have enabled us to solve rather large problem instances. However, these algorithms are evaluated with very limited benchmark instances. In this study, we propose new methods to systematically generate comprehensive benchmark instances. Some instances with special correlation properties between parameters are found to be several orders of magnitude harder than those currently used for benchmarking the algorithms. Experiments on an existing exact algorithm and two generic solvers show that instances whose weights are uncorrelated with the profits are easier compared with weakly or strongly correlated cases. Instances with classes containing similar set of profits for items and with weights strongly correlated to the profits are the hardest among all instance groups investigated. These hard instances deserve further study and understanding their properties may shed light to better algorithms.     

Optimization of a multi-Constant Work-in-Process semiconductor assembly and test factory based on performance evaluation

This paper considers a semiconductor assembly and test factory which is a three-segment-Constant Work-in-Process (CONWIP) system with overlapping machines. In the system, three types of carts circulate for meeting the physical requirements. The optimization problem in setting the suitable total Work-in-Process (WIP) level and the distribution in the three loops from the view of the trade-off between the throughput and the WIP level for the system is addressed. In the proposed model, the system is firstly modeled as a three-loop closed queue network and we propose an approximate method to evaluate the performance. The accuracy of the evaluation method was illustrated by numerical experiments, indicating that the method is fairly precise. Secondly, a Genetic Algorithm is designed to obtain near optimal results based on the performance evaluation. The semiconductor assembly and test system case as well as the application procedure were carried out in detail.     

Multi-item inventory model of breakable items with stock-dependent demand under stock and time dependent breakability rate

Items made of glass, ceramic, etc. are normally stored in stacks and get damaged during the storage due to the accumulated stress of heaped stock. These items are known as breakable items. Here a multi-item inventory model of breakable items is developed, where demands of the items are stock dependent, breakability rates increase linearly with stock and nonlinearly with time. Due to non-linearity and complexity of the problem, the model is solved numerically and final decisions are made using Genetic Algorithm (GA). In a particular case, model is solved analytically as well as numerically and results are compared. Models are developed with both crisp and uncertain inventory costs. For uncertain inventory costs both fuzzy and stochastic parameters are considered. A chance constrained approach is followed to deal with simultaneous presence of stochastic and fuzzy parameters. Different numerical examples are used to illustrate the problem for different cases.