An adaptive strategy based on linear prediction of queue length to minimize congestion in Barabási-Albert scale-free networks

In this paper, we propose an adaptive strategy based on the linear prediction of queue length to minimize congestion in Barabaisi-Albert （BA） scale-free networks. This strategy uses local knowledge of traffic conditions and allows nodes to be able to self-coordinate their accepting probability to the incoming packets. We show that the strategy can delay remarkably the onset of congestion and systems avoiding the congestion can benefit from hierarchical organization of accepting rates of nodes. Furthermore, with the increase of prediction orders, we achieve larger values for the critical load together with a smooth transition from free-flow to congestion.

A Multipath Routing Algorithm Based on Traffic Prediction in Wireless Mesh Networks

The technology of QoS routing has become a great challenge in Wireless Mesh Networks (WMNs). There exist a lot of literatures on QoS routing in WMNs, but the current algorithms have some deficiencies, such as high complexity, poor scalability and flexibility. To solve the problems above, a multipath routing algorithm based on traffic prediction (MRATP) is proposed in WMNs. MRATP consists of three modules including an algo-rithm on multipath routing built, a congestion discovery mechanism based on wavelet-neural network and a load balancing algorithm via multipath. Simulation results show that MRATP has some characteristics, such as better scalability, flexibility and robustness. Compared with the current algorithms, MRATP has higher success ratio, lower end to end delay and overhead. So MRATP can guarantee the end to end QoS of WMNs.

Dynamic Wavelength and Bandwidth Allocation Using Adaptive Linear Prediction in WDM/TDM Ethernet Passive Optical Networks

Hybrid wavelength-division-multiplexing(WDM)/time-division-multiplexing(TDM) ethernet passive optical networks(EPONs) can achieve low per-subscriber cost and scalability to increase the number of subscribers. This paper discusses dynamic wavelength and bandwidth allocation(DWBA) algorithm in hybrid WDM/TDM EPONs.Based on the correlation structure of the variable bit rate(VBR) video traffic,we propose a quality-ofservice (QoS) supported DWBA using adaptive linear traffic prediction.Wavelength and timeslot are allocated dynamically by optical line terminal(OLT) to all optical network units(ONUs) based on the bandwidth requests and the guaranteed service level agreements(SLA) of all ONUs.Mean square error of the predicted average arriving rate of compound video traffic during waiting period is minimized through Wiener-Hopf equation.Simulation results show that the DWBA-adaptive-linear-prediction(DWBA-ALP) algorithm can significantly improve the QoS performances in terms of low delay and high bandwidth utilization.

Real-Time Prediction of Urban Traffic Problems Based on Artificial Intelligence-Enhanced Mobile Ad Hoc Networks(MANETS)

Traffic in today’s cities is a serious problem that increases travel times,negatively affects the environment,and drains financial resources.This study presents an Artificial Intelligence(AI)augmentedMobile Ad Hoc Networks(MANETs)based real-time prediction paradigm for urban traffic challenges.MANETs are wireless networks that are based on mobile devices and may self-organize.The distributed nature of MANETs and the power of AI approaches are leveraged in this framework to provide reliable and timely traffic congestion forecasts.This study suggests a unique Chaotic Spatial Fuzzy Polynomial Neural Network(CSFPNN)technique to assess real-time data acquired from various sources within theMANETs.The framework uses the proposed approach to learn from the data and create predictionmodels to detect possible traffic problems and their severity in real time.Real-time traffic prediction allows for proactive actions like resource allocation,dynamic route advice,and traffic signal optimization to reduce congestion.The framework supports effective decision-making,decreases travel time,lowers fuel use,and enhances overall urban mobility by giving timely information to pedestrians,drivers,and urban planners.Extensive simulations and real-world datasets are used to test the proposed framework’s prediction accuracy,responsiveness,and scalability.Experimental results show that the suggested framework successfully anticipates urban traffic issues in real-time,enables proactive traffic management,and aids in creating smarter,more sustainable cities.

Tunnelling performance prediction of cantilever boring machine in sedimentary hard-rock tunnel using deep belief network

Evaluating the adaptability of cantilever boring machine(CBM) through in-depth excavation and analysis of tunnel excavation data and rock mass parameters is the premise of mechanical design and efficient excavation in the field of underground space engineering.This paper presented a case study of tunnelling performance prediction method of CBM in sedimentary hard-rock tunnel of Karst landform type by using tunneling data and surrounding rock parameters.The uniaxial compressive strength(UCS),rock integrity factor(Kv),basic quality index([BQ]),rock quality index RQD,brazilian tensile strength(BTS) and brittleness index(BI) were introduced to construct a performance prediction database based on the hard-rock tunnel of Guiyang Metro Line 1 and Line 3,and then established the performance prediction model of cantilever boring machine.Then the deep belief network(DBN) was introduced into the performance prediction model,and the reliability of performance prediction model was verified by combining with engineering data.The study showed that the influence degree of surrounding rock parameters on the tunneling performance of the cantilever boring machine is UCS > [BQ] > BTS >RQD > Kv > BI.The performance prediction model shows that the instantaneous cutting rate(ICR) has a good correlation with the surrounding rock parameters,and the predicting model accuracy is related to the reliability of construction data.The prediction of limestone and dolomite sections of Line 3 based on the DBN performance prediction model shows that the measured ICR and predicted ICR is consistent and the built performance prediction model is reliable.The research results have theoretical reference significance for the applicability analysis and mechanical selection of cantilever boring machine for hard rock tunnel.

基于MLR–ANN算法的地应力场反演与裂缝预测

中国页岩气储层埋藏深,受构造运动影响,地应力分布规律复杂,传统方法很难准确反演区域地应力大小和方向。提出多元线性回归和人工神经网络的耦合算法,对川南长宁—建武区块的页岩气储层及周边地应力场进行反演,并采用综合破裂系数法,对储层裂缝进行预测,划分裂缝发育区域。研究表明,基于多元回归和神经网络的耦合算法能准确反演区域的地应力场分布规律。研究区的地应力以挤压应力为主,方向在NE115°左右。受构造运动产生的断层周边应力较为集中,易发育剪切裂缝,裂缝以发育和较发育程度为主。研究区在邻近龙马溪组底部的五峰组上段和构造大断层部位裂缝发育程度较高。研究成果对该区块完善页岩气开采的井网布置、压裂优化设计和套管损坏防治等有一定的参考价值。

试验环境水下声信号的特征提取方法

水下试验环境参数的反演是水声学研究领域的重要内容。而当前研究的关键是通过对水下声信号做特征提取从而获取参数信息。针对特征提取较难、模型很难拟合等问题。本文提出了一种试验环境水下声信号的特征提取方法。将水下声信号同时用梅尔频谱倒谱系数及线性预测系数处理,两者运用特征加权组合方法得到新的特征矩阵;再应用映射插值算法对特征矩阵进行处理,获得适应神经网络输入的三通道矩阵。本文选取的网络模型为残差神经网络。利用实验室所录制的对河口水库数据集测试表明,本文提出的特征提取方法普遍优于仅利用梅尔频谱倒谱系数或线性预测系数的特征处理方法。利用单频矩形脉冲信号对环境进行深度5分类,准确率平均提升2%。利用线性调频信号对环境进行深度5分类,准确率平均提升2.03%。本文提出的特征提取方法对线性调频信号在深度分类任务下处理的结果要优于单频矩形脉冲信号处理的结果。

基于贝叶斯线性回归的造价预测模型

为在项目前期准确快速地预测高速公路造价,运用贝叶斯线性回归方程对造价进行预测。首先,识别高速公路造价影响因素,建立造价预测指标体系;其次,对收集的高速公路造价数据统一处理后,基于贝叶斯线性回归方程建立造价预测模型,并与BP神经网络预测模型进行预测效果对比;最后,利用Matlab进行仿真训练和预测。结果表明:相较于BP神经网络模型,贝叶斯线性回归模型具有更高的预测精度和稳定性,其预测值误差控制在5%以内,MAPE为2.29%,决定系数为0.925,可见贝叶斯回归模型的拟合度高、预测效果好,具有良好的可行性和适用性,可以应用于高速公路项目的造价预测。

基于Transformer-LSTM网络的轴承寿命预测

轴承是旋转机械设备中的重要部件,由于工况、材质、加工方式等原因,轴承寿命时长相差许多。传统的并行或串行神经网络预测方式,对数据集有一定要求。因此,需要一种能够适用于不同数据长短的轴承剩余使用寿命预测网络。为此提出了一种能够预测不同寿命时长的Transformer-LSTM串并行神经网络预测模型。通过将Transformer解码层进行重构,并与长短期记忆时序神经网络(long short-term memory,LSTM)网络结构融合,实现轴承寿命数据的串并行预测处理。试验结果表明Transformer-LSTM神经网络能够精准预测长、中、短不同寿命时长的轴承失效时间,具有较强的模型泛化能力,提升轴承寿命预测精度与模型的泛化能力。

基于图卷积网络的交通流预测方法综述

近年来,基于深度学习的交通流预测方法一直是交通流预测领域的研究热点.与传统卷积神经网络不同,适合处理非欧几里得数据的图卷积网络在空间特征建模方面表现出了强大的能力,而反映路网空间特征的拓扑图、距离图、流量相似图等正是典型的非欧几里得数据.因此,基于图卷积网络及其变体的交通流预测方法成为交通流预测领域的一个研究热点,并取得了很多有吸引力的研究结果.本文对近年来基于图卷积网络的交通流预测模型进行了分类和总结.首先,从图卷积网络的基本定义出发,结合空域图卷积和谱域图卷积的定义详述了图卷积的基本原理.其次,根据预测模型的网络结构特点,将基于图卷积网络的交通流预测模型分为“组合型”和“改进型”两大类,并对其中最具代表性的模型结构进行了详细分析和讨论.此外,对交通流预测领域中常用于模型性能对比的典型数据集进行了综述,并以其中一个真实数据集为例开展仿真测试,展示了4个基于图卷积网络交通流预测模型的预测性能.最后,基于当前的研究现状和发展趋势,对基于图卷积网络的交通流预测方法研究领域中未来的研究热点和难点进行了开放性的讨论和展望.

融合原子交换特征信息的代谢路径预测

为获取从任意起始代谢物到目标代谢物的生化相关性较好的代谢路径,提出一种融合原子交换特征信息的基于约束的代谢路径预测算法PVA。结合代谢网络中具有的原子交换特征信息,建立一种基于约束的混合整数线性规划(MILP)代谢路径预测模型,以搜索从任意起始代谢物到目标代谢物并包含特定原子交换信息的代谢路径。实验结果表明,与同类方法相比,PVA能够有效地发现生化相关性更好的代谢路径。

一种优化FPGA布线拥塞的FHO-BP网络

超大集成电路的高度复杂化造成的布线拥塞可能导致电路的不可布性,早期的布线拥塞预测对于提高集成电路的最终设计质量非常关键,因此针对现场可编程门阵列(Field Programmable Gate Array,FPGA),引入火鹰优化(Fire Hawk Optimizer,FHO)算法机制优化反向传播(Back Propagation,BP)神经网络,提出一种基于复杂网络和FHO-BP网络的布线拥塞优化方法,将电路布局的复杂网络特征向量应用到布线拥塞度预测模型中,并利用提出的优化算法改善电路布线拥塞。实验结果表明,与经典的BP网络相比,所提FHO-BP预测模型具有更高的预测精度和收敛速度,决定系数达到92.62%,模型的平均训练时间为94.55 s,平均预测时间为0.57 s,并且利用布线拥塞优化算法对布局进行优化后的布线实际拥塞程度明显缓和。

双关系预测与特征融合的实体关系抽取模型

现有分阶段解码的实体关系抽取模型仍存在着阶段间特征融合不充分的问题,会增大曝光偏差对抽取性能的影响。为此,提出一种双关系预测和特征融合的实体关系抽取模型(entity relation extraction model with dual relation prediction and feature fusion,DRPFF),该模型使用预训练的基于Transformer的双向编码表示模型(bidirectional encoder representation from transformers,BERT)对文本进行编码,并设计两阶段的双关系预测结构来减少抽取过程中错误三元组的生成。在阶段间通过门控线性单元(gated linear unit,GLU)和条件层规范化(conditional layer normalization,CLN)组合的结构来更好地融合实体之间的特征。在NYT和WebNLG这2个公开数据集上的试验结果表明,该模型相较于基线方法取得了更好的效果。

The application of modeling and prediction with MRA wavelet network

As there are lots of non-linear systems in the real engineering, it is very important to do more researches on the modeling and prediction of non-linear systems. Based on the multi-resolution analysis (MRA) of wavelet theory, this paper combined the wavelet theory with neural network and established a MRA wavelet network with the scaling function and wavelet function as its neurons. From the analysis in the frequency domain, the results indicated that MRA wavelet network was better than other wavelet networks in the ability of approaching to the signals. An essential research was carried out on modeling and prediction with MRA wavelet network in the non-linear system. Using the lengthwise sway data received from the experiment of ship model, a model of offline prediction was established and was applied to the short-time prediction of ship motion. The simulation results indicated that the forecasting model improved the prediction precision effectively, lengthened the forecasting time and had a better prediction results than that of AR linear model.The research indicates that it is feasible to use the MRA wavelet network in the short -time prediction of ship motion.

Prediction of Water Table Based on General Regression Neural Network

Traditional methods for water table prediction have such defects as extensive calculation and reliance on the presupposition of a homogeneous and regular aquifer.Based on the fundamentals of the general regression neural network(GRNN),this article sets up a GRNN model for water level prediction.Case study indicates that this model,even with limited information,has satisfactory prediction accuracy,which,coupled with a simple model structure and relatively high calculation efficiency,mean a vast application prospect for the model.

基于时空图卷积神经网络的离港航班延误预测

对于日益频发的机场航班延误,精准的航班延误预测是最重要的防范措施之一。通过谱图卷积将机场网络从不规则的图结构转换为规则的网络结构,利用图卷积神经网络(GCN)和门控线性单元(GLU)捕获网络中的时空相关性并组成时空卷积块,提出可以预测离港航班延误状况的时空图卷积神经网络(STGCN)。遴选美国51座枢纽机场构建机场网络,并预测未来一段时间内的机场离港准点率以检验STGCN用于预测航班延误的可行性。结果表明:当预测窗口为1天时,STGCN预测结果的平均绝对误差(MAE)相对于历史平均(HA)法、长短期记忆循环神经网络(LSTM)、堆栈自编码器(SAEs)分别下降了18.19%、10.45%、6.24%;当预测窗口为2天时,MAE分别下降了9.93%、3.96%、4.37%;当预测窗口为3天时,MAE分别下降了7.02%、2.47%、9.20%。实例证明STGCN相比传统模型能够显著提升航班延误预测的精度,并为机场制定延误决策提供参考指导。

Movie Score Prediction Model Based on Movie Plots

With the rapid development of the movie industry, it is vital to evaluate and predict a movie’s quality. In this paper, a movie score prediction model is proposed based on the movie plots. Movie data was processed with the word2vec method, and the linear regression model and back propagation neural network algorithm were employed to establish the movie score prediction model. The high-quality classic movie plots of high-scoring movies summed up by big data contributed to a high synthesis of the wonderful content of the film. Experimental results show that it is effective in terms of movie evaluation and prediction, and helpful in understanding people’s preferences for movie plots.

考虑航路网络结构的离场航班延误预测模型

针对离场航班延误预测缺少对航路网络结构因素的考虑,以及传统多分类预测难以满足高精度的需求,提出了一种考虑航路网络结构的离场航班延误预测模型。首先,根据离场航班所在终端区的航路网络结构,提出了航路拥挤指标,即航路流量、航路拥挤度和航路网络拥挤度,从航路网络和网络结构2个维度量化分析了拥挤特征,构造了航路拥挤数据集;然后,基于深度神经网络(deep neural network,DNN),构建了考虑航路网络结构的离场航班延误预测模型;最后,分析各类别延误样本比例,调整焦点损失函数的平衡因子以及各模型参数,进行了不同损失函数、不同数据集和不同模型参数的对比实验。结果表明:调整平衡因子后,模型预测准确率提高了2.3个百分点,融入航路拥挤数据集后,准确率继续提高了1.52个百分点,并且最终达到93.47%。可见,本文所提模型能够对离场航班延误作出有效准确判断,为民航相关单位提供决策参考。

融合FCM-RBF的短时交通拥堵状态预测模型

针对高速公路常发性拥堵路段,提出一种融合模糊C均值聚类算法和径向基函数神经网络的短时交通拥堵状态预测模型。模型基于FCM聚类算法获取历史交通流的拥堵状态标签以及不同交通状态的聚类中心;基于RBF神经网络算法实现短时交通流参数预测。将RBF神经网络预测得到的短时交通流参数代入FCM聚类结果中,得到短时交通拥堵状态标签。通过交通流参数与交通状态的隐含关系,搭建出融合模型的基本计算架构。结果表明:FCM聚类算法训练后的分类结果更加稳定有效;RBF神经网络比对照方法具有更高的预测精度,预测相对误差基本低于1.2%;建立的FCM-RBF模型对短时交通拥堵状态预测的分类正确率达到95%,预测结果准确可靠。

基于DBN和多元线性回归的风电功率预测

提出一种结合深度置信网络(Deep Belief Network, DBN)和多元线性回归的风电功率预测方法,同时考虑风速、风向对功率的影响进行预测研究。首先运用多元线性回归模型建立风速、风向和风电场输出功率之间的映射关系,寻找最优拟合解;然后对深度置信网络进行分层预训练并在最顶层加入多元线性回归最优解,构建预测模型在监督微调下达到参数最优;最后通过某风电场风速、风向对风电功率模型进行预测,表明提出的模型有更好的拟合效果、更精确的预测结果。