舰艇武器布置问题的一种协同优化算法

针对舰艇武器布置问题的特点，提出了一种基于粒子群优化和分类器系统的协同优化算法，以粒子群优化进行优化计算，用分类器系统消除约束．计算实例表明，该算法能较好地实现优化计算，并能节省大量的计算时间．     

基于多核Boosting方法的通用目标分类和识别

针对传统基于SVM分类器的多核学习方法优化参数多、优化过程复杂、计算量大的缺点,本文提出基于Real Adaboost的多核学习方法解决通用目标分类与识别问题.该方法根据核函数能将高维特征映射到低维空间的特性,采用核函数空间上的线性平面分割构建弱分类器,并用Real Adaboost学习框架对弱分类器进行学习.先用分层...     

排课系统导入数据时字段名称转换问题的研究

本文讨论了使用表驱动方法解决在开发排课系统中遇到的数据格式不一致的问题，该问题即导入数据过程中出现的相同性质的数据在不同的系统中具有不同的表示名称的问题。该方法使用简单的原理实现复杂的不同结构数据的合并操作，对于数据库系统的开发具有很强的实践意义。该方法已应用于实际的排课系统项目，受到使用者的好评。     

基于随机卷积特征和集成超限学习机的快速SAR目标识别

深度卷积神经网络在目标检测与识别等方面表现出了优异性能,但将其用于SAR目标识别时,较少的训练样本和深度模型的优化设计是必须解决的两个问题。本文设计了一种结合二维随机卷积特征和集成超限学习机的SAR目标识别算法。首先,随机生成具有不同宽度的二维卷积核,对输入图像进行卷积与池化操作,提取随机卷积特征向量。其次,为提高分类器的泛化能力,并降低训练时间,基于集成学习思想对提取的卷积特征进行随机采样,然后采用超限学习机训练基分类器。最后,通过投票表决法对基分类器的分类结果进行集成。采用MSTAR数据集进行了SAR目标识别实验,实验结果表明,由于采用的超限学习机具有快速训练能力,训练时间降低了数十倍,在无需进行数据增强的情况下,分类精度与采用数据增强和多层卷积神经网络的深度学习算法相当。提出的算法具有实现简单、需要调整参数少等优点,采用集成学习思想提高了分类器的泛化能力。

     

基于改进半监督生成对抗网络的少量标签轴承智能诊断方法EI北大核心CSCD

轴承在线监测大数据主要是由大量无标签数据和少量有标签数据组成的。已有的智能诊断方法多依赖于大量有标签数据的监督学习。针对此问题,提出一种改进半监督生成对抗网络,利用生成器与分类器的对抗学习增强分类器的辨识能力,提出了增强特征匹配算法,借助分类器的深层特征优化网络损失值计算,进而提高网络收敛速度,结合半监督学习使用少量有标签数据进一步提升分类器的学习能力,最终实现对无标签数据的正确归类。使用四组轴承试验数据的迁移学习验证和对比了改进深度网络对不同轴承、工况、故障生成模式、故障程度的辨识能力,结果表明该网络具有更高的分类准确率和更快的收敛速度。     

模糊神经网络及其在高炉参数学习专家系统中的应用

本文应用模糊神经网络对高炉参数学习系统进行研究，该神经网络采用个有线性激励的ＢＰ快速学习算法，能够实现模糊推进进行隶属函数参数和规则可信度的调整，为动态模拟专家系统知识库，本文采用了七个结构相似且相互关联的神经网络代表不同的高炉异常炉况类型。     

H∞-学习神经网络的非线性系统识别

基于H∞-滤波算法提出了一种有效的回归神经网络学习方法。该方法能有效解决系统模型和噪声统计特性存在不确定的问题。在利用神经网络进行非线性系统识别中，该方法对任意的网络结构能进行整体优化训练，具有良好的泛化能力。在非线性结构系统识别中验证了该方法的可行性与有效性。     

基于改进判别字典学习的故障诊断方法

近年来,基于稀疏表示的分类技术在模式识别中取得一定的成功。该框架中,字典的学习和分类器的训练通常是两个独立的模块,降低了方法的识别精度。针对以上问题,提出了一种特征提取和模式识别相融合的改进判别字典学习模型,将重构误差项、稀疏编码判别项及分类误差项进行了整合,并用K奇异值分解算法对目标函数进行优化,实现了字典和分类器的同步学习。该方法先对原始信号进行经验模态分解,并从分解的本征模态函数中提取时、频特征,形成故障样本;然后将训练样本输入改进模型用K奇异值分解优化;最后用习得字典及分类器权重对测试样本进行识别。实验结果表明：该算法不但适用于小样本故障问题,而且鲁棒性和分类性能都明显高于其它算法。         

基于模糊规则的非线性系统快速建模方法

描述了一个通过学习从输入输出采样数据中提取MISO模糊规则的具有插值性能的新型模糊系统。提出的模糊模型继承了Sugeno模型及其变化形式的许多优点。对于新的模糊模型采用递推最小二乘估计方法代替启发式误差反馈学习方法辨识结论参数，整个辨识过程所需的计算时间相对于其他方法最短，适用于在线辨识。仿真结果表明了该方法的有效性与实用性。     

集成轮廓跟踪

本文将跟踪看作是二分类问题,提出了一种基于Adaboost集成学习和快速水平集的轮廓跟踪算法.该方法首先在线地训练一个弱分类器的集合用以区分目标和背景,而通过Adaboost将集合中的各弱分类器组合成一个强分类器,并用于标定下一帧中的各像素的类别属性,从而确定快速水平集算法的速度函数,然后采用基于动态邻近区域快速水平集来演化目标边界曲线以实现目标的轮廓跟踪.为适应目标和背景的变化,在跟踪过程中在线训练新的弱分类器,而时间相关性则通过更新包含新弱分类器的集合来实现.实验结果表明,在摄像机运动、光照变化,部分遮挡或目标尺度变化等情况下,能实现刚体或非刚体目标的轮廓跟踪.     

工业机器人协同加工工序标准工时测定方法研究

以工业机器人为研究对象,提出工业机器人完成一道工序的标准工时模型。针对工业机器人工作特点,得出工业机器人完成一道工序的标准工时构成要素;分别讨论单台工业机器人、多台工业机器人串联、多台工业机器人并联组成一道工序的工作特征,基于可靠性等相关理论,推导这3种情况下该工序的正常工时(t)、平均故障修复时间(MTTR)、平均预防性维修时间(MPMT)的数学模型;其中从工序能力出发,将并联工业机器人转化成串联形式进行模型推导。以某条智能生产线上各工业机器人为例,与正常工时相比,基于新的标准工时得出的标准产能与实际产能的误差由4.29%下降到1.72%。     

Improved heuristics and a genetic algorithm for finding short supersequences

In this paper several heuristics and a genetic algorithm (GA) are described for the Shortest Common Supersequence (SCS) problem, an NP-complete problem with applications in production planning, mechanical engineering and data compression. While our heuristics show the same worst case behaviour as the classical Majority Merge heuristic (MM) they outperform MM on nearly all our test instances. We furthermore present a genetic algorithm based on a slightly modified version of one of the new heuristics. The resulting GA/heuristic hybrid yields significantly better results than any of the heuristics alone, though the running time is much higher. Received: 1 April 1997 / Accepted: 11 September 1997     

Improved protein secondary structure prediction using support vector machine with a new encoding scheme and an advanced tertiary classifier

Prediction of protein secondary structures is an important problem in bioinformatics and has many applications. The recent trend of secondary structure prediction studies is mostly based on the neural network or the support vector machine (SVM). The SVM method is a comparatively new learning system which has mostly been used in pattern recognition problems. In this study, SVM is used as a machine learning tool for the prediction of secondary structure and several encoding schemes, including orthogonal matrix, hydrophobicity matrix, BLOSUM62 substitution matrix, and combined matrix of these, are applied and optimized to improve the prediction accuracy. Also, the optimal window length for six SVM binary classifiers is established by testing different window sizes and our new encoding scheme is tested based on this optimal window size via sevenfold cross validation tests. The results show 2% increase in the accuracy of the binary classifiers when compared with the instances in which the classical orthogonal matrix is used. Finally, to combine the results of the six SVM binary classifiers, a new tertiary classifier which combines the results of one-versus-one binary classifiers is introduced and the performance is compared with those of existing tertiary classifiers. According to the results, the Q/sub 3/ prediction accuracy of new tertiary classifier reaches 78.8% and this is better than the best result reported in the literature.     

Sequencing algorithms for multiple arc-welding robots considering thermal distortion

This paper focuses on sequencing welding lines in multiple arc-welding robot systems, which is the problem of assigning individual welding operations to robots as well as determining their sequences with the objective of minimizing the maximum completion time. Each welding operation is denoted by a weld line with two end-points, each of which can be a possible starting point for the welding operation. Thermal distortion is explicitly considered by introducing a delay between welding operations associated with weld lines near each other. Due to the complexity of the problem, this paper suggests three types of search heuristics, genetic algorithms, simulated annealing and tabu search, each of which incorporates new methods to generate neighbourhood solutions. To show the performances of the heuristics, computational experiments are performed on a number of randomly generated test problems, and the results are reported. In particular, some of the search heuristics outperform the exiting method.     

涡流空气分级机回顾与展望

对涡流空气分级设备的历史进行了回顾 ,并列举了几种有代表性的机型 ,分析了涡流空气分级机的工作原理 ,指出了涡流空气分级机存在的问题和今后发展的方向     

Computer aided robot selection: the ‘multiple attribute decision making’ approach

This paper presents an efficient approach for a computer-based solution to the problem of selection of an ‘optimum robot’ specifically to aid industries. The approach breaks new ground by applying Multiple Attribute Decision Making (MADM) ideology to equipment selection (robots) which is of consequence because selection is an implicit aspect of design. Indeed, the approach is also applicable to a robot design problem. The robot selection procedure allows rapid convergence from a very large number to a manageable shortlist of potentially suitable robots using an ‘elimination search’ routine based on a few pertinent attributes of the robots. Subsequently, the selection procedure proceeds to rank the alternatives in the shortlist by employing a MADM method termed TOPS1S (Technique for Order Preference by Similarity to Ideal Solution). An expert system has been developed as part of the software package to assist an inexperienced user to establish priorities, and to ‘oversee’ the selection process at various stages. The procedure is also illustrated by an example.     

Analysis of the type II robotic mixed-model assembly line balancing problem

In recent years, there has been an increasing trend towards using robots in production systems. Robots are used in different areas such as packaging, transportation, loading/unloading and especially assembly lines. One important step in taking advantage of robots on the assembly line is considering them while balancing the line. On the other hand, market conditions have increased the importance of mixed-model assembly lines. Therefore, in this article, the robotic mixed-model assembly line balancing problem is studied. The aim of this study is to develop a new efficient heuristic algorithm based on beam search in order to minimize the sum of cycle times over all models. In addition, mathematical models of the problem are presented for comparison. The proposed heuristic is tested on benchmark problems and compared with the optimal solutions. The results show that the algorithm is very competitive and is a promising tool for further research.     

A two-stage path planning approach for multiple car-like robots based on PH curves and a modified harmony search algorithm

In this article, collision-avoidance path planning for multiple car-like robots with variable motion is formulated as a two-stage objective optimization problem minimizing both the total length of all paths and the task’s completion time. Accordingly, a new approach based on Pythagorean Hodograph (PH) curves and Modified Harmony Search algorithm is proposed to solve the two-stage path-planning problem subject to kinematic constraints such as velocity, acceleration, and minimum turning radius. First, a method of path planning based on PH curves for a single robot is proposed. Second, a mathematical model of the two-stage path-planning problem for multiple car-like robots with variable motion subject to kinematic constraints is constructed that the first-stage minimizes the total length of all paths and the second-stage minimizes the task’s completion time. Finally, a modified harmony search algorithm is applied to solve the two-stage optimization problem. A set of experiments demonstrate the effectiveness of the proposed approach.     

未来的数码照相机将如何发展

根据莫尔定律,以电脑为首的电子产品每九个月就会有一次升级换代,数码相机也是如此。据市场调研分析,我们认为未来数码照相机主要将在外观和摄影性能方面作改进,本文从这两个方面进行探讨。