广西高技术产业科技创新能力评价研究

文章在分析广西高技术产业发展现状基础上,构建高技术产业科技创新能力评价指标体系,运用熵值法,基于2015—2019年面板数据对广西高技术产业的创新能力进行实证研究.研究表明:广西高技术产业创新能力在西部五省中处于较低水平;广西高技术创新产出实力略低于创新投入实力,创新技术吸收实力仍相对较弱;广西高技术四大行业创新能力存在差异,即电子及通信设备制造产业科技创新投入和产出效果最好,医药制造业和电子通信及装备制造业创新投入高,创新产出评分偏低,医疗仪器设备及仪器仪表制造产业和计算机及办公设备制造产业创新吸收能力很弱.基于以上结论,文章从加大创新资源投入、推动创新平台建设、加强人才队伍建设等方面提出了对策建议.     

基于新型空间注意力机制和迁移学习的垃圾图像分类算法

随着我国各级政府大力推动垃圾强制分类,分类回收各环节中实现标准化、自动化的垃圾分类识别需要适合云端部署的高准确率、低延时要求的细粒度图像分类模型.本文发挥深度迁移学习的优势建立了一套端到端的迁移学习网络架构GANet (garbage neural network);针对垃圾分类中类别易混淆、背景干扰等挑战,提出一种新型的像素级空间注意力机制PSATT (pixel-level spatial attention).为克服类别多和样本不平衡挑战,提出使用标签平滑正则化损失函数;为改善收敛速度以及模型稳定性与泛化性,提出了阶梯形OneCycle学习率控制方法,并给出了结合Rectified Adam (RAdam)优化方法和权重平滑处理技术的组合使用策略.实验使用了"华为云人工智能大赛.垃圾分类挑战杯"提供的按照深圳市垃圾分类标准标注的训练数据,验证了GANet在垃圾分类问题中的显著效果,获得了全国二等奖(第2名);同时,提出的PSATT机制优于对比方法,且在不同主干网络架构上均得到了提升,具有良好的通用性.本文提出的GANet架构、PSATT机制和训练策略不仅具有重要的工程参考价值,也具有较好的学术价值.     

Study on the Centralization Strategy of the Blood Allocation Among Different Departments within a Hospital

By far, the researches on how to distribute blood products among different departments in hospital have not been further studied, though the problem of blood shortage and wastage that caused by improper blood allocation is severe, which may endanger patient’s lives and impose considerable costs on hospitals. In order to solve this problem, this paper mainly studies on how to distribute the blood items among different departments within a hospital and investigates the allocation approach with the novel management method by centralizing the inventory of several different departments. By integrating the blood inventory requirements of some departments, the hospital could reduce the rate of blood shortage and wastage effectively, release the pressure of the occupancy of resources and reduce the bullwhip effect of blood products. This paper illustrates the centralization principle in hospital and formulates the mixed integer programming model to work out the optimal allocation network scheme and the optimal inventory setting for every department. And the results of the numerical example demonstrate that this centralization method could considerably reduce blood shortage and wastage in hospital by about 72% and 90% respectively. Furthermore, it could decrease the total cost by about 108,540 RMB a month on blood supply chain management in the hospital and improve the effect of some certain surgeries by transfusing the fresh blood to patients.     

A 12-bit 250-MS/s Charge-Domain Pipelined Analog-to-Digital Converter with Feed-Forward Common-Mode Charge Control

A feed-forward Common-Mode(CM) charge control circuit for a high-speed Charge-Domain(CD)pipelined Analog-to-Digital Converter(ADC) is presented herein.This study aims at solving the problem whereby the precision of CD pipelined ADCs is restricted by the variation in input CM charge,which can compensate for CM charge errors caused by a variation in CM charge input in real time.Based on the feed-forward CM charge control circuit,a 12-bit 250-MS/s CD pipelined ADC is designed and realized using a 1 P6 M 0.18-μm CMOS process.The ADC achieved a Spurious Free Dynamic Range(SFDR) of 78.1 d B and a Signal-to-Noise-and-Distortion Ratio(SNDR) of 64.6 d B for a 20.1-MHz input:a SFDR of 74.9 d B and SNDR of 62.0 d B were achieved for a 239.9-MHz input at full sampling rate.The variation in signal-to-noise ratio was less than 3 d B over a 0–1.2 V input CM voltage range.The power consumption of the prototype ADC is only 85 m W at 1.8 V supply,and it occupies an active die area of 2.24 mm~2.     

我国股票市场行业间波动溢出效应研究——基于改进的EMD去噪方法

针对我国股市噪声交易能量较大、行业指数同步性较高的特点,提出改进的EMD(empirical mode decomposition,经验模态分解)去噪方法对行业数据进行处理,进而采用BEKK-GARCH模型分析金融危机前、金融危机期间、金融危机后三个阶段行业间波动溢出效应.研究表明,在降低股价同步性方面,改进的EMD去噪方法效果更佳;行业间波动溢出效应在金融危机期间显著上升,金融危机后回落;较之金融危机前,金融危机后传统制造业受产业链上游的波动溢出效应有所降低、受科研的波动溢出效应有增强的趋势.     

高性能膜分离材料、膜过程强化关键技术及装备的研制与应用

针对影响和制约膜分离技术大规模推广应用的关键材料与技术瓶颈,中国科学院过程工程研究所"生化工程介质与设备"创新团队经过多年自主创新,在高性能分离膜材料、膜过程强化关键技术和高效膜分离装备等方面取得了系列科研成果,部分成果在调味品、水处理、中药等行业实现了产业化应用,创造了显著的经济效益和社会效益,为提高我国膜工业研发与应用水平,缩短我国膜分离行业与国外先进水平之间的差距,促进膜分离技术的大规模工业化应用做出了贡献。     

Two Innovative Coalition Formation Models for Dynamic Task Allocation in Disaster Rescues

This paper addresses the problem of multi-objective coalition formation for task allocation. In disaster rescue, due to the dynamics of environments, heterogeneity and complexity of tasks as well as limited available agents, it is hard for the single-objective and single (task)-to-single (agent) task allocation approaches to handle task allocation in such circumstances. To this end, two multi-objective coalition formation for task allocation models are proposed for disaster rescues in this paper. First, through coalition formation, the proposed models enable agents to cooperatively perform complex tasks that cannot be completed by single agent. In addition, through adjusting the weights of multiple task allocation objectives, the proposed models can employ the linear programming to generate more adaptive task allocation plans, which can satisfy different task allocation requirements in disaster rescue. Finally, through employing the multi-stage task allocation mechanism of the dynamic programming, the proposed models can handle the dynamics of tasks and agents in disaster environments. Experimental results indicate that the proposed models have good performance on coalition formation for task allocation in disaster environments, which can generate suitable task allocation plans according to various objectives of task allocation.     

低能N+注入凤仙花的变异检测

以低能N 注入凤仙花(Impatiens balsaminaL)种子后获得了高茎、矮茎2个变异品系.对其进行过氧化物同工酶(POD)检测的结果表明,2个变异品系叶片的POD分子大小、表达强度和条带数目与对照相比都发生了一定变化.随机扩增多态性DNA(random amplified polymorphic DNA,RAPD)分析结果也表明变异品系DNA链的脱氧核苷酸序列发生了变异,从25个随机引物中筛选出了9个具有多态性.实验证明适当剂量的低能N 注入能够在基因水平引起凤仙花的变异,为花卉产业的发展提供了一种遗传育种手段.     

基于DSPWVD-TFPF的通信信号噪声消除算法研究

针对通信系统低信噪比情况下,通信质量下降,传统的时频峰值滤波算法(time frequency peak filter,TFPF)在低信噪比下会出现时频聚集性下降的情况,利用联合定向平滑Wigner-Ville分布实现瞬时频率估计.这种新的分布基于定向平滑Wigner-Ville分布(directionally smoothed pseudo-Wigner-Ville distributions,DSPWVD)和一系列不同方向上的DSPWVD逐点自适应权重.仿真实验说明,在低信噪比情况下,DSPWVD-TFPF算法相比传统的TFPF,能够更好的提高时频聚集性,降低误码率(Bit Error Ratio,BER).