Deep reinforcement learning for selecting demand forecast models to empower Industry 3.5 and an empirical study for a semiconductor component distributor

A semiconductor distributor that plays a third-party role in the supply chain will buy diverse components from different suppliers, warehouse and resell them to a number of electronics manufacturers with vendor-managed inventories, while suffering both risks of oversupply and shortage due to demand uncertainty. However, demand fluctuation and supply chain complexity are increasing due to shortening product life cycle in the consumer electronics era and long lead time for capacity expansion for high-tech manufacturing. Focusing realistic needs of a leading distributor for semiconductor components and modules, this study aims to construct a UNISON framework based on deep reinforcement learning (RL) for dynamically selecting the optimal demand forecast model for each of the products with the corresponding demand patterns to empower smart production for Industry 3.5. Deep RL that integrates deep learning architecture and RL algorithm can learn successful policies from the dynamic and complex real world. The reward function mechanism of deep RL can reduce negative impact of demand uncertainty. An empirical study was conducted for validation showing practical viability of the proposed approach. Indeed, the developed solution has been in real settings.     

5 000 t公务船调距桨桨叶数选型研究

针对国内外调距桨桨叶数选型依靠经验为主，对不同桨叶数对推进器整体性能的影响始终缺乏定量分析的现状，以某5 000 t公务船为目标船型，通过调距桨分别配置不同的桨叶数，分析不同桨叶数对目标船型综合性能的影响。总结了公务船调距桨桨叶数选型的方法，归纳了不同叶数调距桨的优劣，可为后续新造船桨叶数的选取提供参考。     

栈式降噪自编码器在波形单元识别中的应用

为提高多功能雷达(Multi-Function Radar,MFR)波形单元的识别准确率和鲁棒性,提出一种栈式降噪自编码器(Stacked Denoising Autoencoders,SDAE)与支持向量机(Support Vector Machine,SVM)相结合的波形单元识别方法.首先摒弃传统依赖脉冲序列分析技术对MFR信号进行处理的方法,通过分析波形单元结构并借助参数间的联合变化特征,提出一种MFR波形单元分段识别模型,将传统对脉冲序列的识别转化为对MFR波形单元的识别;然后在该模型的基础上引入SDAE,对训练样本数据、SDAE隐含层神经元节点进行加噪处理,并利用这些加噪后的样本数据训练优化SDAE网络模型,提取出样本数据的深层稳健特征;最后引入SVM算法,借助SDAE挖掘出的样本深层特征,实现SVM模型的优化,得到最终的波形单元识别模型(SDAE-SVM).仿真实验表明:提出的波形单元识别方法在相同样本数目和测试误差的条件下,与SVM算法相比,能够取得较高的识别准确率,具备更优越的识别效果.证实MFR波形单元识别模型是有效的,且通过SDAE网络的引入,使得SDAE-SVM方法能够自主地挖掘原始信号的深层特征,提高波形单元识别的鲁棒性和准确率.     

An Improved Torque Sensorless Speed Control Method for Electric Assisted Bicycle With Consideration of Coordinate Conversion

In this paper, we propose an improved torque sensorless speed control method for electric assisted bicycle, this method considers the coordinate conversion. A low-pass filter is designed in disturbance observer to estimate and compensate the variable disturbance during cycling. A DC motor provides assisted power driving, the assistance method is based on the real-time wheel angular velocity and coordinate system transformation. The effect of observer is proved, and the proposed method guarantees stability under disturbances. It is also compared to the existing methods and their performances are illustrated through simulations. The proposed method improves the performance both in rapidity and stability.      

工业无线传感器网络攻击源定位任务分配优化算法

针对工业无线传感器网络中参与攻击源节点定位的任务分配问题,构建和求解多目标优化定位任务分配模型,任务分配模型中设定参考节点组合总能量消耗、距离平均标准偏差目标函数,以及空间约束和剩余能量约束条件;采用循环拥挤排序将非支配排序遗传算法(NSGA-Ⅱ)进行改进后加入基于稀疏度局部搜索的混合优化算法联合求解任务分配模型,将稀疏度最小的解作为稀疏解,再采用极限优化策略在稀疏解周围进行局部搜索使得解拥有更好的分布特性.Matlab仿真结果表明该改进的混合优化算法可以提高算法收敛速度以及降低算法复杂度,在较快的时间内选择出合适的参考节点组合,减少了定位误差,提高了定位精度.     

基于GFU和分层LSTM的组群行为识别研究方法

提出一种以"关键人物"为核心，使用门控融合单元（GFU，Gated Fusion Unit）进行特征融合的组群行为识别框架，旨在解决两个问题：①组群行为信息冗余，重点关注关键人物行为特征，忽略无关人员对组群行为的影响；②组群内部交互行为复杂，使用GFU有效融合以关键人物为核心的交互特征，再通过LSTM时序建模成为表征能力更强的组群特征.最终，通过softmax分类器进行组群行为类别分类.该算法在排球数据集上取得了86.7%的平均识别率.     

循环水替代7℃水在氯乙烯精馏装置中的应用研究

详细计算了PVC生产精馏系统换热设备的热负荷、循环水量及运行成本,认为通过调整系统的运行温度和压力,精馏装置使用循环水替代7℃水具有可行性,这将有利于突破能耗高的技术瓶颈,最大程度地发挥节能降耗、减污增效的潜能。     

数据中心内服务器能耗数据采集及特征分析

数据中心的高能耗低能效问题正受到广泛关注与研究，但目前没有公开的数据中心内服务器能耗数据集供研究人员使用，且现有过滤式特征选择并不能满足运维人员的需求。为此，提出了一套模拟数据中心内服务器运行状态的仿真环境架构，基于该架构采集了服务器运行各类任务时的多项性能指标和能耗数据。然后将基于因果关系的特征选择应用于能耗数据集的特征分析中，构建出可解释的特征子集和能耗预测结果。实验结果表明，因果特征子集大小约为过滤式特征子集大小的1/3到1/6，并且使用因果特征子集训练的模型在75%的情况下都取得了最优预测精度。     

A new DEA common-weight multi-criteria decision-making approach for technology selection

This paper addresses an advanced manufacturing technology selection problem by proposing a new common-weight multi-criteria decision-making (MCDM) approach in the evaluation framework of data envelopment analysis (DEA). We improve existing technology selection models by giving a new mathematical formulation to simplify the calculation process and to ensure its use in more general situations with multiple inputs and multiple outputs. Further, an algorithm is provided to solve the proposed model based on mixed-integer linear programming and dichotomy. Compared with previous approaches for technology selection, our approach brings new contributions. First, it guarantees that only one decision-making unit (DMU) (referring to a technology) can be evaluated as efficient and selected as the best performer while maximising the minimum efficiency among all the DMUs. Second, the number of mixed-integer linear programs to solve is independent of the number of candidates. In addition, it guarantees the uniqueness of the final optimal set of common weights. Two benchmark instances are used to compare the proposed approach with existing ones. A computational experiment with randomly generated instances is further proceeded to show that the proposed approach is more suitable for situations with large datasets.     

Assessing robustness of factor ranking for supersaturated designs

Supersaturated designs can potentially be a beneficial tool for efficiently exploring a large number of factors with a moderately sized design. However, because more factors are being considered than there are runs, the stability of the identified factors depends heavily on effect sparsity and the lack of highly influential observations. A helpful tool for the analysis of supersaturated designs is least absolute shrinkage and selection operation (LASSO), which is useful when the effects of many explanatory variables are sparse in a high‐dimensional dataset. To understand the impact of individual observations on the selected factors, the LASSO influence plot was created. This paper describes an application of this plot and its variants that can be used to identify influential points, increase understanding of the impact of individual observations on model parameters, and the robustness of results in analyses with supersaturated designs. These graphical methods can serve as a complement to other regression diagnostics techniques in the LASSO regression setting.