New Latency Model for Dynamic Frequency Scaling on Network-on-Chip

Modulating both the clock frequency and supply voltage of the network-on-chip （NoC） during runtime can reduce the power consumption and heat flux, but will lead to the increase of the latency of NoC. It is necessary to find a tradeoff between power consumption and communication latency. So we propose an analytical latency model which can show us the relationship of them. The proposed model to analyze latency is based on the M/G/1 queuing model, which is suitable for dynamic frequency scaling. The experiment results show that the accuracy of this model is more than 90%.

Modeling online social networks based on preferential linking

We study the phenomena of preferential linking in a large-scale evolving online social network and find that the linear preference holds for preferential creation, preferential acceptance, and preferential attachment. Based on the linear preference, we propose an analyzable model, which illustrates the mechanism of network growth and reproduces the process of network evolution. Our simulations demonstrate that the degree distribution of the network produced by the model is in good agreement with that of the real network. This work provides a possible bridge between the micro=mechanisms of network growth and the macrostructures of online social networks.

Intelligent modeling and optimization on time-sharing power dispatching system for electrolytic zinc process

Based on real time price counting of electric power, an optimization model of time sharing power for electrolytic zinc process(EZP) was established by means of an incremental fuzzy neural network(FNN), which is adopted to approximate the relationship of current efficiency, current density and acidity. Penalty function introduced and optimal objective function reconstructed, a single loop simulated annealing algorithm(SAA) by using mutation and extending searching spaces was used to obtain optimal time sharing power scheme. Industrial practical results show that the whole system can greatly decrease the power consumption of EZP and increase the time sharing profits.

Low Power Transceiver Design Parameters for Wireless Sensor Networks

Designing low power sensor networks has been the general goal of design engineers, scientist and end users. It is desired to have a wireless sensor network (WSN) that will run on little power (if possible, none at all) thereby saving cost, and the inconveniences of having to replace batteries in some difficult to access areas of usage. Previous researches on WSN energy models have focused less on the aggregate transceiver energy consumption models as compared to studies on other components of the node, hence a large portion of energy in a WSN still get depleted through data transmission. By studying the energy consumption map of the transceiver of a WSN node in different states and within state transitions, we propose in this paper the energy consumption model of the transceiver unit of a typical sensor node and the transceiver design parameters that significantly influences this energy consumption. The contribution of this paper is an innovative energy consumption model based on simple finite automata which reveals the relationship between the aggregate energy consumption and important power parameters that characterize the energy consumption map of the transceiver in a WSN;an ideal tool to design low power WSN.

Deep Neural Network Based Behavioral Model of Nonlinear Circuits

With the rapid growth of complexity and functionality of modern electronic systems, creating precise behavioral models of nonlinear circuits has become an attractive topic. Deep neural networks (DNNs) have been recognized as a powerful tool for nonlinear system modeling. To characterize the behavior of nonlinear circuits, a DNN based modeling approach is proposed in this paper. The procedure is illustrated by modeling a power amplifier (PA), which is a typical nonlinear circuit in electronic systems. The PA model is constructed based on a feedforward neural network with three hidden layers, and then Multisim circuit simulator is applied to generating the raw training data. Training and validation are carried out in Tensorflow deep learning framework. Compared with the commonly used polynomial model, the proposed DNN model exhibits a faster convergence rate and improves the mean squared error by 13 dB. The results demonstrate that the proposed DNN model can accurately depict the input-output characteristics of nonlinear circuits in both training and validation data sets.

Nonlinear Dynamical Behavior in BS Evolution Model Based on Small-World Network Added with Nonlinear Preference

We introduce a modified small-world network adding new links with nonlinearly preferential connectioninstead of adding randomly,then we apply Bak-Sneppen(BS)evolution model on this network.We study severalimportant structural properties of our network such as the distribution of link-degree,the maximum link-degree,and thegth of the shortest path.We further argue several dynamical characteristics of the model such as the important criticalvalue f_c,the f_0 avalanche,and the mutating condition,and find that those characteristics show panticular behaviors.

ANN-Based Identification of Steady-State Behavior Parameters of Composite Power Semiconductor Device Model

The paper describes the application of an ANN based approach to the identification of the parameters relevant to the steady state behavior of composite power electronic device models of circuit simulation software. The identification of model parameters of IGBT in PSPICE using BP neural network is illustrated.

Power-Aware Wireless Communication System Design for Body Area Networks

With the explosive development of wireless communication and low power embedded techniques, Body Area Network (BAN) has opened up new frontiers in the race to provide real-time health monitoring. IEEE 802 has established a Task Group called IEEE 802.15.6 inNovember 2007 and aims to establish a communication standard optimized for low power, high reliability applied to medical and non-medical application for BANs. This paper overviews the path loss model and the communication scheme for implant-to-body surface channel presented by IEEE 802.15.6 standard. Comparing with the standard scheme where BCH (Bose-Chaudhuri-Hochquenghem) code is employing, we propose a new coding solution using convolutional code operating with Bit Interleaver based on the properties of implant-to-body surface channel. To analyze the performance of the two Error Correct Coding (ECC) schemes, we performed simulations in terms of Bit Error Rate (BER) and power consumption on MATLAB and FPGA platform, respectively. The simulation results proved that with appropriate constraint length, convolutional code has a better performance not only in BER, but also in minimization of resources and power consumption.

UEs Power Reduction Evolution with Adaptive Mechanism over LTE Wireless Networks

At present, the major drawback for mobile phones is the issue of power consumption. As one of the alternatives to decrease the power consumption of standard, power-hungry location-based services usually require the knowledge of how individual phone features consume power. A typical phone feature is that the applications related to multimedia streaming utilize more power while receiving, processing, and displaying the multimedia contents, thus contributing to the increased power consumption. There is a growing concern that current battery modules have limited capability in fulfilling the long-term energy need for the progress on the mobile phone because of increasing power consumption during multimedia streaming processes. Considering this, in this paper, we provide an offline meaning sleep-mode method to compute the minimum power consumption comparing with the power-on solution to save power by implementing energy rate adaptation(RA) mechanism based on mobile excess energy level purpose to save battery power use. Our simulation results show that our RA method preserves efficient power while achieving better throughput compared with the mechanism without rate adaptation(WRA).

Modeling Simulation Technology Research for Distribution Network Planning

This paper proposes to use the power system simulation software CYME to plan, model and simulate for an actual distribution network for improving the reliability and efficiency, enhancing the efficiency and capacity, simulating the abnormal condition of distribution network, and presenting operation program of safe, reliable and having simulation record statements. The modeling simulation results show that the software module has lots of advantages including high accuracy, ideal reliability, powerful practicality in simulation and analysis of distribution network, it only need to create once model, the model can sufficiently satisfy multifarious types of simulation analysis required for the distribution network planning.

Optimal Design for Wireless Power Transfer System with Relay Resonators

The optimization method by adjusting load and distances between two adjacent coils( or resonators) is presented on basis of wireless power transfer( WPT) system with relay resonators. 2-port network and impedance matching theory are applied to analyzing power flow of incidence and reflection in WPT system,then setting up power flow model. The maximum power transmission efficiency can be obtained when the load and distance between secondary resonator and output coil meets impedance matching at 2-port network's output port. The simulation and experimental results shown the impedance matching method can effectively improve and maintain transmission efficiency by adjusting load and distances between coils or relay resonators.

Reluctance Network Analysis for Complex Coupled Inductors

The use of reluctance networks has been a conventional practice to analyze transformer structures. Basic transformer structures can be well analyzed by using the magnetic-electric analogues discovered by Heaviside in the 19th century. However, as power transformer structures are getting more complex today, it has been recognized that changing transformer structures cannot be accurately analyzed using the current reluctance network methods. This paper presents a novel method in which the magnetic reluctance network or arbitrary complexity and the surrounding electrical networks can be analyzed as a single network. The method presented provides a straightforward mapping table for systematically linking the electric lumped elements to magnetic circuit elements. The methodology is validated by analyzing several practical transformer structures. The proposed method allows the analysis of coupled inductor of any complexity, linear or non-linear.

航空电推进电机多层波浪形拓扑及散热设计方法

针对航空电推进电机的散热问题,提出一种基于多层波浪形散热拓扑的航空电推进电机高效散热设计方法。航空电推进电机采用多层波浪形散热拓扑,建立电机等效热网络模型,确定等效热阻、对流换热系数等重要参数,完成电机温度精确计算,并通过CFD仿真验证所建立电机等效热网络模型的准确性和有效性。以此为基础,对比分析传统散热翅和多层波浪形散热拓扑对电机功率密度的影响。基于多层波浪形散热拓扑的等效热网络模型,采用遗传学习粒子群优化(GL-PSO)算法,完成航空电推进电机高效散热优化设计。优化结果表明:相比于原始方案,优化方案的机壳质量减轻15.1%,整个电机的功率密度提升0.06 kW/kg。

抽油机井恒功率变频控制系统研究

抽油机作为陆上油田最重要的机械采油装备之一,近年来已发展出一系列的变频节能技术,但其研究主要集中在硬件方面,缺乏相关基础理论,导致现场应用普适性差,应用效果参差不齐。为此,研究有杆泵变频测控机理,优选了主要部件,对测试数据提出了要求,形成了有杆泵抽油变频控制系统。构建BP神经网络系统对正常井和漏失、柱塞碰泵、气体影响等异常井进行识别,分类提出变频计算模型:考虑负扭矩和抽油机冲次界限等建立了正常井恒功率变频计算模型,实例对比分析了变频后抽油机速度、加速度变化规律,产液量一定时,负功功率和有功功率分别下降了65%和44%。建立了上行程漏失的异常井变频计算模型,针对性地提高上行程冲次,形成“上快下慢”运行模式。实例计算表明:该系统缩短上行程时间2.6 s,较好地解决有杆泵上行程漏失和变频节能的问题。研究成果可为抽油机变频技术提供理论支持,对油田节能和提产具有指导意义。

基于模型迁移和SVM的煤矿电网单相接地电容电流预测方法

针对智慧化矿山背景下煤矿电网单相接地电容电流预测精度低和智能预测实现困难的问题,提出了一种单相接地电容电流智能预测方法。采用模型迁移思想扩充矿用橡套软电缆的电容电流样本数据,以解决智能预测中电容电流数据不完备的问题;基于Sobol’敏感性方法分析了影响单相接地电容电流的关键因素及其各因素的交互关系,确定智能预测模型输入特征量;在此基础上,通过基于稀疏技术的支持向量机建立煤矿电网单相接地电容电流智能预测模型,并引入鲸鱼算法优化预测模型超参数,克服了电容电流数据样本容量小的不足。采用厂家测试数据及多家煤矿电网现场实测数据进行试验,结果表明:电缆绝缘层外径和绝缘层内径是影响单相接地电容电流的关键因素;所提方法预测煤矿电网单相接地电容电流的平均误差为2.26%,相较于现有预测方法,误差分别下降了34.19%,24.91%和7.40%。该方法实现了煤矿电网单相接地电容电流的准确预测,并为其智能化预测提供了新思路。

基于注意力机制的CNN-LSTM-XGBoost台风暴雨电力气象混合预测模型

极端台风暴雨灾害具有非线性、极差大以及多峰值等特点。为使电网及时获取预警信息,提出一种基于注意力机制的CNN-LSTM-XGBoost台风暴雨电力气象混合预测模型。首先,利用基于注意力机制的卷积神经网络(CNN)辨识关键台风暴雨灾害特征;然后,利用长短期记忆网络(LSTM)训练时间序列预测模型以挖掘台风暴雨时序特征,使用极限梯度提升算法替换模型输出层以缓解过拟合问题;最后,以2023年台风泰利为例验证所提方法的有效性。算例分析表明,所提模型具有较高的准确性,对预测精度的提升可达40.84%以上。

基于多层时序网络模型的中国机场网络特征分析

为克服静态网络模型在分析动态系统时出现网络结构特性反映不全面和数据表达失真的问题,引入时间属性,提出一种基于时序复杂网络模型的中国机场网络(CAN)结构分析方法。通过分析节点时序度、时序度中心性、时序介数中心性和时序接近中心性等网络特征指标,探究中国机场时序网络的结构特征,筛选和排序网络中的重要机场节点,并分析异质节点产生的原因。结果表明:CAN结构呈现出无标度网络特性;CAN的平均时序距离值差距较大,最大值是最小值的6.06倍;乌鲁木齐地窝堡国际机场(ZWWW)、呼和浩特白塔国际机场(ZBHH)的时序介数中心性相对其他指标较高,其在网络中起到重要的中心作用,而深圳宝安国际机场(ZGSZ)相对其他指标具有较高的时序度中心性。

基于GRU-XGBoost的短期风电功率预测研究

本文利用自适应噪声的完备经验模态分解(CEEMDAN)对风电原始序列信号进行处理后,采用GRU-XGBoost模型对非线性、非平稳的功率序列进行建模和预测,以提升模型的预测能力和泛化性。首先,通过CEEMDAN将风电功率原始序列分解为不同时间尺度的分量,然后将分解后的信号输入GRU神经网络生成预测信号,最后通过XGBoost进行校正。通过与多种预测模型进行比较,证明了该模型在预测精度方面的卓越表现。

面向算力网络的算力建模与度量技术研究

算力度量与建模技术作为算力网络发展的重要基础,通过统一建模算力需求和资源,结合网络性能指标,形成算网能力模板,为算力路由、管理和计费提供统一度量标准。深入探讨了该技术的发展趋势与挑战,分析了算力网络发展中的算力度量需求。在此基础上对算力网络各层级进行抽象建模,设计了一种基于任务基本处理单元的算力度量和管理方法,并提出了一套完善的算力度量与建模体系,为算力网络的进一步发展提供了有力支撑和新的思路。

基于低风速功率修正和损失函数改进的超短期风电功率预测

风电功率具有较强的波动性和随机性。为进一步提升风电功率的预测精度,提出一种基于低风速功率修正和损失函数改进的超短期风电功率预测模型。该模型采用卷积神经网络、自注意力机制和双向门控循环单元捕获风电功率序列的长期时序依赖关系。为了解决低风速下待风状态神经网络难以精确拟合的问题,模型通过预测风速并结合当前时段的风电功率对低风速段的预测功率进行修正。针对参数训练的稳定性问题,模型通过改进预测策略和共享权重,引入一种多元非线性的损失函数来提取序列间的关联性。结果表明,所提模型在多项误差指标中均优于对比模型,能够有效提升超短期风电功率的预测效果。