网络流量预测是网络拥塞控制与网络管理的一个重要问题.网络流量时间序列具有时变、非线性特征,导致传统时间序列预测方法预测精度比较低,无法建立精确的预测模型.回声状态网络(echo state network,ESN)在非线性混沌系统预测与建模方面...网络流量预测是网络拥塞控制与网络管理的一个重要问题.网络流量时间序列具有时变、非线性特征,导致传统时间序列预测方法预测精度比较低,无法建立精确的预测模型.回声状态网络(echo state network,ESN)在非线性混沌系统预测与建模方面有着良好的性能,非常适合网络流量的预测.为了提高网络流量的预测精度,提出一种基于遗传算法(genetic algorithm,GA)优化回声状态网络的网络流量非线性预测方法.首先利用回声状态网络对网络流量进行预测;然后利用遗传算法对回声状态网络预测模型中的储备池参数进行优化,提高预测模型的预测精度.通过中国联合网络通信公司辽宁分公司采集的实际网络流量数据进行了仿真验证.与差分自回归滑动平均模型(auto regressive integrated moving average,ARIMA)、Elman神经网络以及最小二乘支持向量机(least square support vector machine,LSSVM)这3种常见预测模型进行了对比,仿真结果表明提出的方法具有更高的预测精度与更小的预测误差,更能刻画网络流量复杂的变化特点.展开更多
把坚硬顶板视为弹性梁,把煤柱视为应变软化介质并采用W e ibu ll分布描述它的损伤本构模型,本文对坚硬顶板和煤柱组成的力学系统,用突变理论方法研究了它的演化失稳过程。通过对建立的尖点突变模型的分析发现,系统失稳主要取决于系统的...把坚硬顶板视为弹性梁,把煤柱视为应变软化介质并采用W e ibu ll分布描述它的损伤本构模型,本文对坚硬顶板和煤柱组成的力学系统,用突变理论方法研究了它的演化失稳过程。通过对建立的尖点突变模型的分析发现,系统失稳主要取决于系统的刚度比k与材料的均匀性或脆性指标m值,并给出了失稳的充要条件力学判据和失稳突跳量的表达式。考虑煤柱介质的粘性或蠕变性,建立了系统演化的非线性动力学模型———物理预报模型,并给出了根据顶板沉降观测数据反演非线性动力学模型的方法和稳定性判别准则。对木城涧矿根据观测序列进行了动力学模型的反演分析,一个重要发现是:D值在临近失稳时陡增出现峰值而后急剧下降。根据材料损伤与声发射累计计数的对应关系,建立了系统演化过程中声发射率的动力学模型,并进行了声发射模拟分析和分维分析,发现m值与系统的演化路径对系统演化的声发射活动规律及分维特征有重要影响,单纯根据声发射监测和降维现象预报冲击地压是不可靠的。展开更多
By analysing the instability process of various nonlinear systems, we conclude that their instability precursors are the continual heightening of the response rate or response ratio.Applying this theory to earthquake ...By analysing the instability process of various nonlinear systems, we conclude that their instability precursors are the continual heightening of the response rate or response ratio.Applying this theory to earthquake prediction, we adopt the periodical change of the stress in crust caused by the tide-generating force as the loading and unloading. If we can measure the ratio of the response (such as crust deformation, gravity and seismicity) during the loading period to that during the unloading period, this parameter must contain some characteristic information about the seismogenic process.With nine earthquakes (M≥7) data that occurred in Chinese mainland during 1970-1988, we take the sum of square root o?energy of small earthquakes, which is called the released strain by Benioff, as the response to the tidal loading and unloading. We find that the response ratios of seven earthquakes increase obviously before the main earthquakes.展开更多
文摘把坚硬顶板视为弹性梁,把煤柱视为应变软化介质并采用W e ibu ll分布描述它的损伤本构模型,本文对坚硬顶板和煤柱组成的力学系统,用突变理论方法研究了它的演化失稳过程。通过对建立的尖点突变模型的分析发现,系统失稳主要取决于系统的刚度比k与材料的均匀性或脆性指标m值,并给出了失稳的充要条件力学判据和失稳突跳量的表达式。考虑煤柱介质的粘性或蠕变性,建立了系统演化的非线性动力学模型———物理预报模型,并给出了根据顶板沉降观测数据反演非线性动力学模型的方法和稳定性判别准则。对木城涧矿根据观测序列进行了动力学模型的反演分析,一个重要发现是:D值在临近失稳时陡增出现峰值而后急剧下降。根据材料损伤与声发射累计计数的对应关系,建立了系统演化过程中声发射率的动力学模型,并进行了声发射模拟分析和分维分析,发现m值与系统的演化路径对系统演化的声发射活动规律及分维特征有重要影响,单纯根据声发射监测和降维现象预报冲击地压是不可靠的。
基金This work is sponsored by the National Natural Science Foundation and the Joint Seismological Science Foundation of China.
文摘By analysing the instability process of various nonlinear systems, we conclude that their instability precursors are the continual heightening of the response rate or response ratio.Applying this theory to earthquake prediction, we adopt the periodical change of the stress in crust caused by the tide-generating force as the loading and unloading. If we can measure the ratio of the response (such as crust deformation, gravity and seismicity) during the loading period to that during the unloading period, this parameter must contain some characteristic information about the seismogenic process.With nine earthquakes (M≥7) data that occurred in Chinese mainland during 1970-1988, we take the sum of square root o?energy of small earthquakes, which is called the released strain by Benioff, as the response to the tidal loading and unloading. We find that the response ratios of seven earthquakes increase obviously before the main earthquakes.
