考虑批量分割的单元装配系统工作量平衡联合决策模型

针对日本式单元化生产中批量大小不合理导致单元间工作量不平衡的问题,在单元装配系统构建过程中考虑批量分割,构建流水线向单元装配系统转化的多目标优化模型.该模型以最小化总完工时间和最小化工人总工时为目标,对转化过程中单元构建和批量分割进行联合决策.考虑到多目标优化问题特点以及解空间复杂度,增加局部搜索策略,对算法进行改进,提出INSGA-II算法,设计了基于游标的编码方式表示批量分割,满足单元数量动态变化下批量划分约束.在进化过程中不断优化各单元对应产品种类以及批量大小,平衡单元间工作量,缩短总完工时间.最后,通过数值算例验证所提出方法的有效性,结果表明在产品数量波动较大的情况下,考虑批量分割能更好地平衡单元间工作量.     

多目标粒子群算法在交叉培训规划中的应用

为了进一步提高人力资源交叉培训规划的实用性,增加了对于员工学习行为的考虑,提出了在保证任务覆盖水平的基础上,获得员工满意度最大和学习效率最高的多目标优化模型.本文针对问题的特征,采用多目标粒子群(MOPSO)算法对多目标优化模型进行了求解,并设计了多种算法策略,以适应不同的问题环境.通过数值实验,分析了不同问题规模下,针对不同性能指标算法参数和策略的适用性.最后,以柔性单元装配生产线为例,进行了数值实验,实验结果表明了模型的有效性和合理性.     

模型体结构可控装配方法

为在模型内部生成可控的多孔隙结构,提出一种针对三角网格模型的内部结构装配方法.首先对模型对象内部空间进行划分,确定目标装配区域;然后根据设计要求,采用隐函数表达的参数化结构作为模型单元填充装配区域,并通过优化局部区域的性能实现整体设计目标;最后从参数化表达的体结构模型中提取等势面,将其直接转化为三角网格体模型.实验结果表明:该方法能够构建密度、连通性和梯度可控的复杂结构,保证内部结构单元之间的平滑过渡,同时避免因大量布尔运算操作产生的错误.     

技能组合可选择的赛汝生产系统多能工配置问题

作为以人为中心的赛汝生产系统,如何合理地配置多能工是系统构建的主要问题.针对赛汝生产系统构建的基本问题,考虑工人技能组合的可选择性和工人技能水平的差异性,以最小化最大完工时间和最小化工人工作时间差异为目标,建立赛汝生产系统多能工配置的多目标优化模型.针对模型和解空间的特征,开发并采用基于NSGA-II的多目标优化算法进...     

考虑倒垛因素的轧制计划编制方法

在给定粗轧制计划的基础上考虑钢坯库倒垛优化, 编制详细的轧制计划; 建立以最小化轧制计划内钢坯出 库总倒垛次数与轧制单元之间切换机架次数为目标的多目标整数规划模型; 针对模型特征, 设计一种基于钢坯匹配的单亲遗传算法. 通过基于实际生产数据的实验验证, 相对于传统的手工计算方法, 所提出的算法在优化倒垛次数和切换机架次数上平均提升20 %, 算法和模型是可行且有效的.     

基于响应面方法的航空发动机装配系统的多目标性能优化

将响应面方法应用于复杂的航空发动机装配生产系统的性能优化中,获取系统的最大平均生产率和最小在制品水平,并解决以平均生产率和在制品水平为目标函数,以系统单机加工速度和缓冲区总大小为决策变量进行多目标性能优化问题。解决多目标优化问题的方法是利用线性加权法将多目标转换为单目标,然后应用响应面方法进行优化。实验结果表明:响应面方法能够很好地解决复杂的航空发动机装配生产线的性能优化问题,并为决策者提供多种情况下的最优解。     

考虑学习效应与物流的单元制造系统优化研究

在考虑员工学习效应与产品物流情况下,研究多技能员工生产多类型产品的单元化制造问题。通过建立0-1非线性规划模型,最小化库存成本、延期交货成本和产品在单元间的物流成本。设计具有二维结构的染色体及相应的遗传算法。以光纤连接器制造系统为例,得出最优的人员分配方案和产品生产路线。数值实验表明:在不同生产规模情形下,考虑产品物流时的运作成本比无物流时显著降低。     

多能流系统合作协同的不确定多目标决策

多能流工业生产过程具有多目标、强耦合、时变、不确定性等特点,针对此类系统的平衡调度问题,本文提出一种基于合作协同优化的不确定多目标决策方法.以钢铁企业副产煤气系统为例,针对系统未来状态的不确定性,本文在优化决策的过程中结合卡尔曼滤波方法和贝叶斯定理,提出一种考虑条件预期的不确定决策模型.该模型能够同时分析当前目标和预期目标,从而消除未来状态不确定性带来的影响.针对副产煤气系统多能流强耦合的特点,本文在优化决策过程中综合考虑单能流系统特性以及多能流系统的协同关系,基于图模型原理提出基于双向权重的协同进化方法,从"总体"–"局部"相结合的角度给出最优的决策策略.通过实际钢铁企业数据的仿真实验表明,该方法能够充分考虑未来的不确定性,同时兼顾单能流系统性能和多能流耦合关系,给出合理的调度决策方案.该方法可用于具有多目标、强耦合以及不确定性的复杂多能流系统,为其调度决策问题提供支持.     

生物能源生产多目标优化模型的研究

一个能够同时优化生物能源的生产供应链,包括环境、经济和社会方面影响的多目标优化模型,模拟和优化了综合的生物能生产系统,即用数种不同类的生物质原料以生产电能、热能和可燃气的系统。该系统包含可供用户选择的多种技术,以模块的形式体现在单元过程中。通过模型生命周期的评价(LCA),最终优化目标确定为解决最小化能量生产成本、最大化节能潜力、最小化环境负担、根据用户选择最大或最小化工人数和最大化生物能系统总效率。通过平衡环境和社会责任,本研究结果可帮助计划和生产人员有效地提高生物质系统的经济竞争力。     

混合并行机调度问题的多目标优化模型及算法

针对生产工序的合并造成一种串并联共存的生产布局,研究了一种特殊的混合并行机调度问题,并考虑以最小化总流水时间和最小化总延迟工件数量为目标的多目标调度问题,建立了混合整数规划模型.针对模型特点,设计了一种改进的非支配排序遗传算法进行求解,采用基于启发式方法的初始种群生成方式以提高种群的质量和多样性,并引入一种局域搜索策略以改善求解算法所获得的非支配解的质量及分布性.通过对大量数值算例进行仿真实验,并与典型的多目标优化算法进行比较,结果表明所提出的模型和算法在收敛性、分布性及极端点质量方面均具有优势,能够较好的解决多目标混合并行机调度问题.     

Voice Conversion Based on Maximum-Likelihood Estimation of Spectral Parameter Trajectory

In this paper, we describe a novel spectral conversion method for voice conversion (VC). A Gaussian mixture model (GMM) of the joint probability density of source and target features is employed for performing spectral conversion between speakers. The conventional method converts spectral parameters frame by frame based on the minimum mean square error. Although it is reasonably effective, the deterioration of speech quality is caused by some problems: 1) appropriate spectral movements are not always caused by the frame-based conversion process, and 2) the converted spectra are excessively smoothed by statistical modeling. In order to address those problems, we propose a conversion method based on the maximum-likelihood estimation of a spectral parameter trajectory. Not only static but also dynamic feature statistics are used for realizing the appropriate converted spectrum sequence. Moreover, the oversmoothing effect is alleviated by considering a global variance feature of the converted spectra. Experimental results indicate that the performance of VC can be dramatically improved by the proposed method in view of both speech quality and conversion accuracy for speaker individuality.     

PID based nonlinear processes control model uncertainty improvement by using Gaussian process model

Proportional-integral-derivative (PID) controller design based on the Gaussian process (GP) model is proposed in this study. The GP model, defined by its mean and covariance function, provides predictive variance in addition to the predicted mean. GP model highlights areas where prediction quality is poor, due to the lack of data, by indicating the higher variance around the predicted mean. The variance information is taken into account in the PID controller design and is used for the selection of data to improve the model at the successive stage. This results in a trade-off between safety and the performance due to the controller avoiding the region with large variance at the cost of not tracking the set point to ensure process safety. The proposed direct method evaluates the PID controller design by the gradient calculation. In order to reduce computation the characteristic of the instantaneous linearized GP model is extracted for a linearized framework of PID controller design. Two case studies on continuous and batch processes were carried out to illustrate the applicability of the proposed method.     

Portfolio optimization problems in different risk measures using genetic algorithm

This paper introduces a heuristic approach to portfolio optimization problems in different risk measures by employing genetic algorithm (GA) and compares its performance to mean–variance model in cardinality constrained efficient frontier. To achieve this objective, we collected three different risk measures based upon mean–variance by Markowitz; semi-variance, mean absolute deviation and variance with skewness. We show that these portfolio optimization problems can now be solved by genetic algorithm if mean–variance, semi-variance, mean absolute deviation and variance with skewness are used as the measures of risk. The robustness of our heuristic method is verified by three data sets collected from main financial markets. The empirical results also show that the investors should include only one third of total assets into the portfolio which outperforms than those contained more assets.     

Host load prediction with long short-term memory in cloud computing

Host load prediction is significant for improving resource allocation and utilization in cloud computing. Due to the higher variance than that in a grid, accurate prediction remains a challenge in the cloud system. In this paper, we apply a concise yet adaptive and powerful model called long short-term memory to predict the mean load over consecutive future time intervals and actual load multi-step-ahead. Two real-world load traces were used to evaluate the performance. One is the load trace in the Google data center, and the other is that in a traditional distributed system. The experiment results show that our proposed method achieves state-of-the-art performance with higher accuracy in both datasets.     

一种基于协同进化的流水线向Seru系统转化方法

Seru生产系统是一种被广泛应用于电子制造产业的新型生产模式,但由于流水线向Seru系统转化问题（Line-seru conversion）包含有Seru构建与Seru调度两个相互耦合的子问题,现有算法难以在同时兼顾解的质量与计算效率的情况下对问题进行求解.因此,本文针对流水线向Seru系统转化问题的特点,提出了一种协同进化算法,即在进化算法中加入了协同机制,将Seru构建与Seru调度子问题作为两个子种群利用该机制进行协同进化,从而弥补了现有算法的不足.并且,本文还针对问题特点设计了个体基因编码方式,从而使规划获得的Seru生产系统具有更优的生产性能及均衡性能.实验表明,采用加入了协同机制的进化算法比传统解决流水线向Seru系统转化问题的方法具有更好的性能,本文所提的方法在最小化产品流通时间和劳动时间有较好的性能表现,并且具有较高的计算效率.     

基于移动agent的分布式入侵检测系统研究*

为了提高现有分布式入侵检测系统的效率和性能,提出了一种基于移动agent的分布式入侵检测系统模型。将移动agent技术应用于入侵检测中,并给出了其移动agent间的可靠通信方法,实现了agent的协同检测。实验结果表明,由于移动agent的应用,入侵检测系统的节点成为了可移动的部件,从而使该模型具有了更强的抗攻击性和入侵检测能力。     

基于双高斯GMM的特征参数规整及其在语音识别中的应用

对特征参数概率分布的实验分析表明,在有噪声影响的情况下,特征参数通常呈现双峰分布.据此,本文提出了一种新的,基于双高斯的高斯混合模型(Gaussian mixture model,GMM)的特征参数归一化方法,以提高语音识别系统的鲁棒性.该方法采用更为细致的双高斯模型来表达特征参数的累积分布函数(CDF),并依据估计得到的CDF进行参数变换将训练和识别时的特征参数的分布都规整为标准高斯分布,从而提高识别正确率.在Aurora 2和Aurora 3数据库上的实验结果表明,本文提出的方法的性能明显好于传统的倒谱均值规整(Cepstral mean normalization,CMN)和倒谱均值方差规整(Cepstral mean and variance normalization,CMVN)方法,而与非参数化方法-直方图均衡特征规整方法的性能基本相当.     

远程控制实验室的多Agent模型研究

通过对理想远程控制实验室设计目标的论述,提出了一种基于多Agent技术的模型,给出了该模型的体系结构和Agent构造方法,强调通过增强Agent智能来提高实验室系统性能,对实时性和安全性有充分的保障体系,并支持复杂的协作实验。     

Learning from Batched Data: Model Combination Versus Data Combination

Combining models learned from multiple batches of data provide an alternative to the common practice of learning one model from all the available data (i.e. the data combination approach). This paper empirically examines the base-line behavior of the model combination approach in this multiple-data-batches scenario. We find that model combination can lead to better performance even if the disjoint batches of data are drawn randomly from a larger sample, and relate the relative performance of the two approaches to the learning curve of the classifier used. In the beginning of the curve, model combination has higher bias and variance than data combination and thus a higher error rate. As training data increases, model combination has either a lower error rate than or a comparable performance to data combination because the former achieves larger variance reduction. We also show that this result is not sensitive to the methods of model combination employed. Another interesting result is that we empirically show that the near-asymptotic performance of a single model in some classification tasks can be significantly improved by combining multiple models (derived from the same algorithm) in the multiple-data-batches scenario.