Vertical Federated Learning Based on Consortium Blockchain for Data Sharing in Mobile Edge Computing

The data in Mobile Edge Computing(MEC)contains tremendousmarket value,and data sharing canmaximize the usefulness of the data.However,certain data is quite sensitive,and sharing it directly may violate privacy.Vertical Federated Learning(VFL)is a secure distributed machine learning framework that completes joint model training by passing encryptedmodel parameters rather than raw data,so there is no data privacy leakage during the training process.Therefore,the VFL can build a bridge between data demander and owner to realize data sharing while protecting data privacy.Typically,the VFL requires a third party for key distribution and decryption of training results.In this article,we employ the consortium blockchain instead of the traditional third party and design a VFL architecture based on the consortium blockchain for data sharing in MEC.More specifically,we propose a V-Raft consensus algorithm based on Verifiable Random Functions(VRFs),which is a variant of the Raft.The VRaft is able to elect leader quickly and stably to assist data demander and owner to complete data sharing by VFL.Moreover,we apply secret sharing todistribute the private key to avoid the situationwhere the training result cannot be decrypted if the leader crashes.Finally,we analyzed the performance of the V-Raft and carried out simulation experiments,and the results show that compared with Raft,the V-Raft has higher efficiency and better scalability.

Optimized Design of H-Type Vertical Axis Wind Airfoil at Multiple Angles of Attack

Numerical simulations are conducted to improve the energy acquisition efficiency of H-type vertical axis wind turbines through the optimization of the related blade airfoil aerodynamic performance.The Bézier curve is initi-ally used tofit the curve profile of a NACA2412 airfoil,and the moving asymptote algorithm is then exploited to optimize the design of the considered H-type vertical-axis wind-turbine blade airfoil for a certain attack angle.The results show that the maximum lift coefficient of the optimized airfoil is 8.33%higher than that of the original airfoil.The maximum lift-to-drag ratio of the optimized airfoil exceeds the maximum lift-to-drag ratio of the ori-ginal airfoil by 11.22%.Moreover,the power coefficient is increased by 12.19%and the torque coefficient of the wind turbine is significantly improved.

Soft measurement model of ring's dimensions for vertical hot ring rolling process using neural networks optimized by genetic algorithm

Vertical hot ring rolling(VHRR) process has the characteristics of nonlinearity,time-variation and being susceptible to disturbance.Furthermore,the ring's growth is quite fast within a short time,and the rolled ring's position is asymmetrical.All of these cause that the ring's dimensions cannot be measured directly.Through analyzing the relationships among the dimensions of ring blanks,the positions of rolls and the ring's inner and outer diameter,the soft measurement model of ring's dimensions is established based on the radial basis function neural network(RBFNN).A mass of data samples are obtained from VHRR finite element(FE) simulations to train and test the soft measurement NN model,and the model's structure parameters are deduced and optimized by genetic algorithm(GA).Finally,the soft measurement system of ring's dimensions is established and validated by the VHRR experiments.The ring's dimensions were measured artificially and calculated by the soft measurement NN model.The results show that the calculation values of GA-RBFNN model are close to the artificial measurement data.In addition,the calculation accuracy of GA-RBFNN model is higher than that of RBFNN model.The research results suggest that the soft measurement NN model has high precision and flexibility.The research can provide practical methods and theoretical guidance for the accurate measurement of VHRR process.

Research and Application on Web Information Retrieval Based on Improved FP-Growth Algorithm

A kind of single linked lists named aggregative chain is introduced to the algorithm, thus improving the architecture of FP tree. The new FP tree is a one-way tree and only the pointers that point its parent at each node are kept. Route information of different nodes in a same item are compressed into aggregative chains so that the frequent patterns will be produced in aggregative chains without generating node links and conditional pattern bases. An example of Web key words retrieval is given to analyze and verify the frequent pattern algorithm in this paper.

Accelerating BP-Based Iterative Low-Density Parity-Check Decoding by Modified Vertical and Horizontal Processes

Two modified BP algorithms related to vertical and horizontal processes are proposed to accelerate iterative low-density parity- check （LDPC） decoding over an additive white Gaussian noise （AWGN） channel, where the newly updated extrinsic information is immediately used in the current decoding round. Theoretical analysis and simulation results demonstrate that both the modified approaches provide significant performance improvements over the traditional BP algorithm with almost no additional decoding complexity. The proposed algorithm with modified horizontal process offers even better performance than another algorithm with the modified horizontal process. The two modified BP algorithms are very promising in practical communications since both can achieve an excellent trade-off between the performance and decoding complexity.

Investigation of Natural Convection Heat Transfer Coefficient on Extended Vertical Base Plates

In this research, Computational analysis of the laminar natural convection on vertical surfaces has been in-vestigated. Natural convection is observed when density gradients are present in a fluid acted upon by a gravitational field. Our example of this phenomenon is the heated vertical plate exposed to air, which, far from the plate, is motionless. The CFD simulations are carried out using fluent software. Governing equa-tions are solved using a finite volume approach. Coupling between the velocity and pressure is made with SIMPLE algorithm. The resultant system of discretized linear algebraic equations is solved with an alternat-ing direction implicit scheme. Then a configuration of rectangular fins is put in different ways on the surface and natural convection heat transfer coefficient on these no slope surfaces is studied and finally optimization is done.

Recognition of vertical vowel graphemes of Korean characters based on combination of vowel graphemes

Korean characters consist of 2 dimensional distributed consonantal and vowel graphemes. The purpose of reducing the 2 dimensional characteristics of Korean characters to linear arrangements at early stage of character recognition is to decrease the complexity of following recognition task. By defining the identification codes for the vowel graphemes of Korean characters, the rules for combination of vowel graphemes are established, and a recognition algorithm based on the rules for combination of vowel graphemes, is therefore proposed for vertical vowel graphemes. The algorithm has been proved feasilbe through demonstrating simulations.

Vertical two-dimensional non-hydrostatic pressure model with single layer

The vertical two-dimensional non-hydrostatic pressure models with multiple layers can make prediction more accurate than those obtained by the hydrostatic pres- sure assumption. However, they are time-consuming and unstable, which makes them unsuitable for wider application. In this study, an efficient model with a single layer is developed. Decomposing the pressure into the hydrostatic and dynamic components and integrating the x-momentum equation from the bottom to the free surface can yield a horizontal momentum equation, in which the terms relevant to the dynamic pressure are discretized semi-implicitly. The convective terms in the vertical momentum equation are ignored, and the rest of the equation is approximated with the Keller-box scheme. The velocities expressed as the unknown dynamic pressure are substituted into the continuity equation, resulting in a tri-diagonal linear system solved by the Thomas algorithm. The validation of solitary and sinusoidal waves indicates that the present model can provide comparable results to the models with multiple layers but at much lower computation cost.

Improvment of Free Convection Heat Transfer Rateof Rectangular Heatsink on Vertical Base Plates

In this paper, the laminar heat transfer of natural convection on vertical surfaces is investigated. Most of the studies on natural convection have been considered constantly whereas velocity and temperature domain, do not change with time, transient one are used a lot. Governing equations are solved using a finite volume approach. The convective terms are discretized using the power-law scheme, whereas for diffusive terms the central difference is employed. Coupling between the velocity and pressure is made with SIMPLE algorithm. The resultant system of discretized linear algebraic equations is solved with an alternating direction implicit scheme. Then a configuration of rectangular fins is put in different ways on the surface and heat transfer of natural convection on these surfaces without sliding is studied and finally optimization is investigated.

Structural Design Optimization of a Vertical Axis Wind Turbine for Seismic Qualification and Lightweight

Recently, there is a growing interest in seismic qualification of ridges, buildings and mechanical equipment worldwide due to increase of accidents caused by earthquake. Severe earthquake can bring serious problems in the wind turbines and eventually lead to an interruption to their electric power supply. To overcome and prevent these undesirable problems, structural design optimization of a small vertical axis wind turbine has performed, in this study, for seismic qualification and lightweight by using a Genetic Algorithm (GA) subject to some design constraints such as the maximum stress limit, maximum deformation limit, and seismic acceleration gain limit. Also, the structural design optimizations were conducted for the four different initial design variable sets to confirm robustness of the optimization algorithm used. As a result, all the optimization results for the 4 different initial designs showed good agreement with each other properly. Thus the structural design optimization of a small vertical-axis wind turbine could be successfully accomplished.

顾及地球物理效应的GNSS高程时间序列AdaBoost预测和插值方法

传统的GNSS高程时间序列预测和插值方法仅考虑时间变量,具有明显的局限性。本文顾及地球物理效应的影响,通过温度、大气压强、极移等数据和GNSS高程时间序列数据构建回归问题,使用自适应提升(AdaBoost)算法建模。为了验证模型的预测和插值性能,试验选取4个GNSS站的高程时间序列进行分析。建模试验表明,相较于Prophet模型,AdaBoost模型的拟合精度提升了约35%;预测结果表明,在12个月的预测周期内,AdaBoost模型在4个GNSS站的MAE值为4.0~4.5 mm,RMSE值约为5.0~6.0 mm;插值试验表明,相较于三次样条插值方法,AdaBoost插值模型的精度约提升了15%~28%。预测和插值试验表明,顾及地球物理效应的AdaBoost模型可以应用于GNSS高程时间序列预测与插值。

基于虚拟现实的立管横焊机器人辅助焊接工艺设计与实现

焊接工艺参数的选择对焊接质量有重要影响。针对立管J型铺设窄坡口横焊熔池下淌特性,提出基于横焊焊道厚度及热输入恒定策略的焊接工艺参数匹配算法。基于Unity3D开发平台和工艺参数匹配算法,通过C#开发了立管横焊CAPP模块,扩展了虚拟现实人机交互系统的功能模块;开展立管横焊计算机辅助焊接工艺实验,焊后焊缝成形良好,无焊接缺陷。实验结果表明:CAPP模块可有效获取匹配的焊接工艺参数,能缩短焊接工艺开发周期,降低工艺开发成本,有效提升系统的智能化水平。

基于改进粒子群算法的VTI介质裂隙预测

裂隙产生的各向异性在地层介质中广泛存在,基于各向异性的裂隙参数反演和预测在一定程度上能提高裂隙反演的精度和预测的可靠性。本文通过建立基于VTI介质的反射系数方程,引入模拟退火算法中的Metropolis准则设置跳出概率对粒子群标准算法进行改进,实现对VTI介质中纵、横波速度及各向异性参数等属性的反演;并通过各向异性相关的属性、泊松比和泊松速度的联合,实现对裂隙充填物的预测。通过二层模型和Marmousi2模型对算法的稳定性和抗噪性进行测试,论证了方法的可行性;并进一步将方法应用于煤矿实际资料,对裂隙含水性进行预测,论证了方法的有效性。

改进海洋捕食者算法的机器人路径规划研究

为解决海洋捕食者算法(MPA)收敛速度慢、收敛精度低、易陷入局部最优的问题,提出一种多策略改进海洋捕食者算法(IMPA)。引入Logistic混沌映射初始化种群,增加捕食者种群多样性;基于当前迭代次数t的自适应移动步长动态调整策略,增强算法逃离局部最优的能力;在IMPA迭代后期,加入中垂线算法(MA),基于中垂线策略的游离粒子位置更新方法,能够加快更新捕食者的位置,增强算法的寻优速度和寻优精度,避免算法陷入局部最优。最后通过改变IMPA阶段转换寻优过程,进一步平衡搜索过程,加强全局与局部适应性。选用6个基准测试函数对算法性能进行测试,测试结果显示:IMPA收敛速度更快,收敛精度更高;最后将改进算法应用于移动机器人路径规划,仿真结果表明:该算法规划的路径长度更短,搜索效率更高。

短距起飞垂直降落飞行器飞行品质研究

根据短距起飞垂直降落(STOVL)飞行器的特点,结合AGARD 577飞行品质规范,从操纵效能和模态特性两方面对短距起飞垂直降落飞行器在飞行品质方面提出要求。基于短距起飞垂直降落飞行器六自由度模型仿真,对飞行器在短距起降过渡状态下的操纵效能进行评估。以等效系统拟配思想,采用遗传算法与最小二乘法的混合算法,对带L_1自适应控制器的高阶飞行器模型进行低阶等效拟配。采用该低阶模型,对飞行器在不同飞行阶段下纵向与横航向的模态特性进行评估。结果表明,该飞行器在短距起降阶段具备一级飞行品质。

基于Spark框架的FP-Growth大数据频繁项集挖掘算法

针对大数据中的频繁项集挖掘问题,提出一种基于Spark框架的FP-Growth频繁项集并行挖掘算法。首先,根据垂直布局思想将数据按照事务标志符垂直排列,以此解决扫描整个数据集的缺陷;然后,通过FPGrowth算法构建频繁模式树,并生成频繁1-项集;接着,通过扫描垂直数据集来计算项集的支持度,从而识别出非频繁项,并将其从数据集中删除以降低数据尺寸;最后,通过迭代过程来生成频繁k-项集。在标准数据集上的实验结果表明,该算法能够有效挖掘出频繁项集,在执行时间方面具有很大的优越性。

基于改进北部苍鹰算法的垂直斗式提升机减速器优化

垂直斗式提升机是一种固定装置的机械输送设备,具有输送能力强、提升高度高、运行平稳可靠、寿命长等优点。减速器是垂直斗式提升机的关键部件,为高效获得减速器的最优解,提出一种冲刺决胜策略改进的北部苍鹰优化算法。采用压力容器和悬臂梁两个经典非线性约束工程算例对改进北部苍鹰算法进行性能验证,结果表明,改进北部苍鹰算法具有良好的收敛速度。在此基础上,对垂直斗式提升机减速器进行优化设计,结果减速器的中心距由初始的720 mm减少为680.16 mm,减少了5.33%。说明改进北部苍鹰优化算法可以作为一种高效的优化算法求解实际工程问题。

基于SCE-UA算法的山区河流洪水预报研究——以寿溪河为例

在山区河流,强降雨易引发山洪灾害,给人民的生命、财产带来严重威胁,因此,及时准确的山区河流洪水预报预警非常重要。然而,由于山区河流洪水突发性强且影响山区河流洪水的原因复杂,山区小流域洪水预报精度始终有限。以四川省典型山区小流域--汶川县寿溪河流域为研究对象,采用三水源新安江模型与垂向混合产流模型,并引进SCE-UA算法进行参数优选,对该流域的洪水预报进行了深入研究。研究表明SCE-UA算法适用于山区河流洪水预报,显著提升两个模型的预报效果,使三水源新安江模型和垂向混合产流模型的合格率分别提高了20%和35%。综合分析各项指标,垂向混合产流模型的预报结果相比三水源新安江模型更好。研究成果可应用于寿溪河流域山洪灾害预警,同时也可为其他山区河流洪水预报提供参考。

X波段双偏振雷达物理量时间—高度剖面的重构方法改进及应用研究

如何利用现有雷达体扫数据重构反射率或其它物理量的时间—高度剖面,提高雷达体扫垂直分辨率并使其适用于云微物理结构的分析,是近几年来雷达气象学的重点研究内容之一。本文基于分辨率更高的X波段双偏振雷达体扫数据,对目前最新的柱垂直廓线(Columnar Vertical Profile,简称CVP)重构算法从目标区范围的选取方面进行改进,使其能够应用于水平尺度较小的局地降水云以及发展演变迅速的对流云。结果显示:对于高原地区局地降水云个例,目标区选取5 km(径向范围)×10°(方位角范围)组成的较小扇形区域,与云雷达的对比显示,改进的CVP方法重构的基本反射率(ZH)垂直廓线体现了回波的垂直结构,尤其是中高层对流泡的结构特点,相应的时间—高度序列能够较好地反映回波顶高的变化以及中高层强度逐渐减弱、低层强度逐渐增加的特点;对于华北地区发展旺盛且局地水平不均匀的对流云个例,本文改进了原始的CVP重构目标区选取方法,对高、低仰角层采用变化的径向范围并调整插值参数,改进后重构的ZH垂直廓线有效避免了低层回波水平分布相对不均匀导致的重构分层结构,显示出高、低层回波特征以及不同阶段目标区云结构的转变。进一步对比改进前后CVP方法重构建立的各偏振量时间—高度序列,改进后准确显示了个例云系微物理特征及其随时间的变化,揭示了高原地区局地降水云中对流泡的形成及其播撒作用机制,华北地区对流云成熟阶段的各偏振量垂直分布特征及其演变。

复合载荷作用下H型垂直轴风力机叶片结构多目标优化

为改善气动力、离心力和重力等复合载荷作用下的H型垂直轴风力机叶片的结构性能,提出一种多目标优化方法。以质量和最大应力最小为目标,最大变形为约束建立优化模型。通过流固耦合(Fluid-Structure-Interaction,FSI)方法,实现叶片表面压力的实时准确提取,建立复合载荷作用下的叶片有限元模型;基于最优空间填充(Optimal Space-Filling,OSF)方法和Kriging模型建立各变量对应力、质量和变形的响应面模型,进行灵敏度和变化趋势分析;最后采用多目标遗传算法(Multi-Objective Genetic Algorithm,MOGA)获得各变量的最优解,并进行结果验证。结果表明,优化后叶片质量减少了14.7%,各方位角下的最大应力减幅最大为7.8%,最大变形减幅最大为16.7%。研究结果可为复合载荷作用下叶片的结构优化设计提供参考。