基于异构网络优化与自适应负荷管理的多微电网调度策略研究

针对多微电网调度策略进行研究,提出一种基于异构网络优化与自适应负荷管理的多微电网调度策略。该策略充分利用了多微电网中不同能源网络的互补与协同特性,设计了一种基于Prophet时间序列模型的自适应负荷管理策略,并引入能源协同性指数和能源互补性指数构建异构网络优化模型,运用多目标混合蚁群算法实现多目标优化,旨在保证系统稳定运行的前提下,实现能源的高效利用和负荷平衡。实验结果表明,该方法能够提高多微电网系统的能源利用效率,降低运行成本,提高电网的稳定性和可靠性。

面向“云-边-端”算力系统的计算和传输联合优化方法

“云—边—端”资源协同优化是算力网络部署的关键难题之一.如何有效整合高性能云计算、低延时边缘计算和低成本用户设备等异构算力资源,对于算力网络建设具有重要意义.基于此,提出了一种面向“云—边—端”算力网络的计算和传输联合优化方案,旨在从应用服务模型、网络状态感知和资源联合优化3方面提供一套系统性解决方案.首先,根据通用应用服务的特性,改进了传统网络服务的链状表征模型,提出了广义图结构的通用服务模型.其次,为表征异构网络状态的动态规律,提出了用于建模计算和传输时变负载的双虚拟队列结构.再次,为降低在大规模算力网络中计算和传输资源联合优化的问题复杂度,提出了一种基于图概念的增广图模型,该模型能够将计算和传输联合优化问题转化为增广图的路由问题,简化了异构资源联合优化问题的形式化表征难度.为实际解决该问题,设计了基于波利亚重球法(Polyak heavy-ball method)的异构资源协同优化算法,并给出了算法复杂度和相关理论的性能分析.最后,通过数值仿真和原型系统实验,验证了算法理论性能的正确性,以及在服务效用和资源成本等方面对比同期3种相关解决方案的性能优势.

异构多种群粒子群优化算法在水位流量关系拟合中的应用

通过8个复杂函数对一种异构多种群粒子群优化算法进行仿真验证,并与传统单种群粒子群优化算法进行对比。针对水位流量关系拟合中相关参数难以确定的不足,利用异构多种群粒子群优化算法优化水位流量关系相关参数,以云南省龙潭站、西洋站水位流量关系拟合为例进行实例研究,并与粒子群优化算法、最小二乘法拟合结果进行对比。结果表明:异构多种群粒子群优化算法收敛精度远远优于粒子群优化算法,具有较好的计算鲁棒性和全局寻优能力。该算法对龙潭站和西洋站水位流量关系拟合的平均相对误差绝对值分别仅为0.27%和0.50%,拟合精度优于粒子群优化算法和最小二乘法。利用异构多种群粒子群优化算法优化水位流量关系可以获得更好的拟合效果。

基于Web服务的协同优化平台(英文)

提出了一种基于Web服务的协同优化平台。该平台具有开放式的架构,能够方便集成现有的通用优化工具。主要特点有:优化工具集成,能够集成本地优化软件及异地优化Web服务;优化服务包装,能够方便的将现有优化工具包装为Web服务;协同优化模型自动生成,解决了通用优化软件的缺点。最后用一个工程案例说明了通过软件智能体,通用优化软件能够访问仿真Web服务完成仿真优化,验证了协同优化平台的有效性。

基于异构模糊神经网络集成算法的自主车导向控制

设计了基于神经网络的自主车导向控制器,建立了异构模糊神经网络控制器的结构,并由实验数据产生训练样本和验证样本集.该控制器通过精确控制2个驱动轮的差动转速实现路径跟踪.实验结果表明,采用异构模糊神经网络集成算法的导向控制器能够稳定地实现跟踪导向路径的控制功能.

三维声波方程正演多级异构并行算法设计与实现

在石油勘探过程中,地震波正演为反演方法提供理论基础及理论数据,能测试反演方法的有效性,在整个反射地震学中占有很重要的地位。通过对三维声波方程正演算法的分析,结合某众核平台的特性,完成了该算法两级MPI并行+众核级并行的设计,并进行了相关优化。实验结果表明,三维声波方程正演多级异构并行算法在该平台具有较好的众核并行加速效果和良好的MPI扩展性能。

小学英语以及同课异构课的探究

作为小学生的英语启蒙教师,教师本人的基本功能直接影响到学生英语的日后发展。从狭义方面来说包括两个方面:语音的标准及语言的流利,规范的板书。小学英语对学生听和说的要求相对多一些,因此,教师语音的规范及语言的流畅至关重要。这就要求我们要不断修炼自己的语音语调基本功,确保我们说出的每个单词,每个句子没有发音上的问题。课堂语言除了要流畅,一定要确保我们说出的不是错误的句子,我们对学生的影响是潜移默化的,课堂上不够规范的课堂用语,可能会给学生日后的学习带来很大的负面影响。

联邦学习在高度数据异构场景下的泛化鲁棒性增强

联邦学习(federated learning,FL)是一种以保护客户隐私数据为中心的分布式处理网络,为解决隐私泄露问题提供了前景良好的解决方案.然而,FL的一个主要困境是高度非独立同分布(non-independent and identically distributed,non-IID)的数据会导致全局模型性能很差.尽管相关研究已经探讨了这个问题,但本文发现当面对non-IID数据、不稳定的客户端参与以及深度模型时,现有方案和标准基线FedAvg相比,只有微弱的优势或甚至更差,因此严重阻碍了FL的隐私保护应用价值.为解决这个问题,本文提出了一种对non-IID数据鲁棒的优化方案:FedUp.该方案在保留FL隐私保护特点的前提下,进一步提升了全局模型的泛化鲁棒性.FedUp的核心思路是最小化全局经验损失函数的上限来保证模型具有低的泛化误差.大量仿真实验表明,FedUp显著优于现有方案,并对高度non-IID数据以及不稳定和大规模客户端的参与具有鲁棒性.

基于神威·太湖之光的非结构网格众核优化技术

为解决高性能计算中的非结构网格离散访存问题,以神威·太湖之光国产超级计算机为平台,根据异构众核处理器SW26010的体系结构特点,提出一种基于排序思想的通用众核优化算法,以减少非结构网格计算中的随机访存。基于网格划分原理,在O(n)时间内对生成的稀疏矩阵非零元素进行并行重排序。采用一种内部映射方式对计算向量实现扩展或变换,将细粒度访存转化为无写冲突的粗粒度访存。对多个实际应用算例的通量计算进行众核优化,结果表明,相比主核上的串行算法,该算法能够获得平均10倍以上的加速效果。

面向异构通信卫星的地球站组网规划方法

经过多年的建设发展,卫星通信已经形成了多系列卫星并存、相互支撑、相互补充的通信系统。然而,由于各卫星系统建设时间跨度大,通信技术手段存在差异,如何高效利用异构的卫星通信资源成为了一个实际难题。为此,文中首次根据卫星与地球站的可互通条件构建了系统模型,并将其归结为线性约束下离散变量的非线性优化问题,优化目标是异构卫星波束间的负载均衡,并在此基础上设计了一种两阶段的启发式组网规划方法。最后,根据地球同步卫星网络特性设置了合理的仿真参数,利用Matlab平台进行仿真实验。仿真结果表明,所提方法在面对不同规模的地球站组网时,在平均负载和负载均衡性上均优于一般的枚举搜索和遗传优化算法,可以在短时间内对大规模的信道单元进行组网规划。

非结构网格下稀疏下三角方程求解器众核优化技术研究

稀疏下三角方程求解器(SpTRSV)作为基础线性代数库中一个重要的算法,在大规模科学计算中有着广泛应用。在非结构网格中,由于非结构网格具有数据存储无序性、数据强相关性以及频繁地离散访存等特点,该算法在众核架构上难以实现有效的并行。文中基于国产异构众核处理器SW26010体系结构的特点,针对非结构网格计算,提出了一种基于流水线串行-局部并行思想的通用众核优化方法。该方法能够有效减少非结构网格计算中的随机访存,提高计算效率,并且具有很好的扩展性。基于该算法对多个实际应用算例进行众核优化,实验结果表明:该方法能够实现单核组3倍以上的加速,显著降低了运行时间。

基于置信度加权的单类协同过滤推荐算法

针对推荐系统中单类协同过滤(OCCF)可解释性差、数据噪声多的缺陷,提出了一种基于置信度加权的单类协同过滤推荐算法。算法通过置信度函数将用户隐性反馈映射为置信概率,并将该函数集成到隐性反馈推荐模型(IFRM)框架中,形成了隐性反馈置信度加权推荐模型(CWIFRM);在此基础上,针对CWIFRM基于随机梯度下降提出了异构置信度优化算法。实验结果表明,该模型在多个数据集上都具有更好的推荐效果,异构置信度优化算法使推荐质量得到了进一步提高,验证了CWIFRM具有较强的适用性、可解释性和抗噪声能力。

Joint Resource Allocation Using Evolutionary Algorithms in Heterogeneous Mobile Cloud Computing Networks

The problem of joint radio and cloud resources allocation is studied for heterogeneous mobile cloud computing networks. The objective of the proposed joint resource allocation schemes is to maximize the total utility of users as well as satisfy the required quality of service(QoS) such as the end-to-end response latency experienced by each user. We formulate the problem of joint resource allocation as a combinatorial optimization problem. Three evolutionary approaches are considered to solve the problem: genetic algorithm(GA), ant colony optimization with genetic algorithm(ACO-GA), and quantum genetic algorithm(QGA). To decrease the time complexity, we propose a mapping process between the resource allocation matrix and the chromosome of GA, ACO-GA, and QGA, search the available radio and cloud resource pairs based on the resource availability matrixes for ACOGA, and encode the difference value between the allocated resources and the minimum resource requirement for QGA. Extensive simulation results show that our proposed methods greatly outperform the existing algorithms in terms of running time, the accuracy of final results, the total utility, resource utilization and the end-to-end response latency guaranteeing.

User Association and Wireless Backhaul Bandwidth Allocation for 5G Heterogeneous Networks in the Millimeter-Wave Band

The user association and wireless backhaul bandwidth allocation for a two-tier heterogeneous network （HetNet） in the mil- limeter wave （mmWave） band is proposed in this article. The two-tier HetNet is built up with a macro base station （MBS） and several small cell SBSs, where the MBS is assumed to be equipped with large-scale antenna arrays but the SBSs only have single-antenna capa- bility and they rely on the wireless link to the MBS for backhaul. The sum of logarithmic user rate, which is established according to the result of multi-user Multiple Input Mul- tiple Output （MIMO） downlink employing Zero-Force Beamforming （ZFBF）, is chosen as the network utility for the objective func- tion. And a distributed optimization algorithm based on primal and dual decomposition is used to jointly optimize the user association variable xj,z and the wireless backhaul band- width factor α. Simulation results reveal that the distributed optimization algorithm jointly optimizing two variables outperforms the con- ventional SINR-based user association strate- gies.

An Optimized Vertical Handoff Algorithm Based on Markov Process in Vehicle Heterogeneous Network

In order to solve the problem the existing vertical handoff algorithms of vehicle heterogeneous wireless network do not consider the diversification of network's status, an optimized vertical handoff algorithm based on markov process is proposed and discussed in this paper. This algorithm takes into account that the status transformation of available network will affect the quality of service(Qo S) of vehicle terminal's communication service. Firstly, Markov process is used to predict the transformation of wireless network's status after the decision via transition probability. Then the weights of evaluating parameters will be determined by fuzzy logic method. Finally, by comparing the total incomes of each wireless network, including handoff decision incomes, handoff execution incomes and communication service incomes after handoff, the optimal network to handoff will be selected. Simulation results show that: the algorithm proposed, compared to the existing algorithm, is able to receive a higher level of load balancing and effectively improves the average blocking rate, packet loss rate and ping-pang effect.

Cooperative task allocation for heterogeneous multi-UAV using multi-objective optimization algorithm

The application of multiple UAVs in complicated tasks has been widely explored in recent years.Due to the advantages of flexibility,cheapness and consistence,the performance of heterogeneous multi-UAVs with proper cooperative task allocation is superior to over the single UAV.Accordingly,several constraints should be satisfied to realize the efficient cooperation,such as special time-window,variant equipment,specified execution sequence.Hence,a proper task allocation in UAVs is the crucial point for the final success.The task allocation problem of the heterogeneous UAVs can be formulated as a multi-objective optimization problem coupled with the UAV dynamics.To this end,a multi-layer encoding strategy and a constraint scheduling method are designed to handle the critical logical and physical constraints.In addition,four optimization objectives:completion time,target reward,UAV damage,and total range,are introduced to evaluate various allocation plans.Subsequently,to efficiently solve the multi-objective optimization problem,an improved multi-objective quantum-behaved particle swarm optimization(IMOQPSO)algorithm is proposed.During this algorithm,a modified solution evaluation method is designed to guide algorithmic evolution;both the convergence and distribution of particles are considered comprehensively;and boundary solutions which may produce some special allocation plans are preserved.Moreover,adaptive parameter control and mixed update mechanism are also introduced in this algorithm.Finally,both the proposed model and algorithm are verified by simulation experiments.

A Simultaneous Optimization Method for the Structure and Compo-sition Differences of a Mass Exchange Network

This article presents a new method to calculate the composition differences （e） for tar g.etin.g the minimum total annualized cost （TAC） of a mass exchange network （MEN）,which is based on the combination of composition interval diagram （CID） with mathematical programming.The total cost target consists of the capital cost of the process units and the operating cost for mass separating agents （MS.As）. The value of total cost varies considerablv with the composition differences, so the values of e should be optimized in order to obtain minimum TAC of a MEN. This articleconsiders ε as a set of unequal variables for each equilibrium equation of a rich-lean stream pair, employing them to build the CID and mathematical model, which optimizes the structure and composition differences simultaneously. Two examples are applied to illustrate the proposed method and the results demonstrate that the approach introduced by this article is simpler and more convenient than the methods in previous literatures.

Multi-core optimization for conjugate gradient benchmark on heterogeneous processors

Developing parallel applications on heterogeneous processors is facing the challenges of 'memory wall',due to limited capacity of local storage,limited bandwidth and long latency for memory access. Aiming at this problem,a parallelization approach was proposed with six memory optimization schemes for CG,four schemes of them aiming at all kinds of sparse matrix-vector multiplication (SPMV) operation. Conducted on IBM QS20,the parallelization approach can reach up to 21 and 133 times speedups with size A and B,respectively,compared with single power processor element. Finally,the conclusion is drawn that the peak bandwidth of memory access on Cell BE can be obtained in SPMV,simple computation is more efficient on heterogeneous processors and loop-unrolling can hide local storage access latency while executing scalar operation on SIMD cores.

Power Consumption Optimization of Downlink Macro-Femto Heterogeneous Networks Based on Cross-Layer Design

Based on the cross-layer design, the power-optimization problem of Macro-Femto Heterogeneous Networks （HetNets） has been formulated. The constraints of power and re-source block allocation in the physical layer, delay and target data rate in the medium ac-cess control layer, urgent queue length in the network layer, and packet error rate in the transport layer, have been considered. The original problem is non-deterministic polyno-mial time hard, which cannot be solved practi-cally. After the restrictions of upper layers are translated into constraints with physical layer parameters, and the integer restrictions are relaxed, the original problem can be decom- posed into convex optimization subproblems. The optimal solutions of resource block allo-cation and power allocation can be obtained by using the Lagrangian optimization. Simula-tion results show that the proposed scheme is better than both the round robin algorithm and the max-rain one in terms of energy efficiency, throughput and service fairness. The round robin algorithm and the max-min one only focus on the user fairness rather than quality of service fairness. Compared to the round robin scheme （the max-min one）, the proposed scheme improves the energy efficiency 58.85% （62.41%）, the throughput 19.09% （25.25%）, the service fairness 57.69% （35.48%）.

卫星通信在网络中的具体应用

经过多年的建设发展,卫星通信已经形成了多系列卫星并存、相互支撑、相互补充的通信系统。然而,由于各卫星系统建设时间跨度大,通信技术手段存在差异,如何高效利用异构的卫星通信资源成为了一个实际难题。因此,研究全新的高通量卫星通信系统交织多址接入方法。通过抑制系统误码率,延长数据传输帧长,减少保护间隔,增加频谱利用率;并重新调整LDPC码度分布,更新交织规则,设计一套全新的多址接入方法。实验结果表明:与现阶段的两种传统接入方法相比,引入研究的方法后,高通量卫星通信系统的吞吐量更高,丢包率也进一步降低。可见研究的多址接入方法可应用到高通量卫星通信系统中。