基于EMD和高斯过程-灰色预测的话务量预测..doc

基于EMD和高斯过程-灰色预测的话务量预测 宋秀秀1，贾振红1*，覃锡忠1，曹传玲2，牛洪梅1.新疆大学信息科学与工程学院 乌鲁木齐 830046； 2.中国移动通信集团新疆有限公司 乌鲁木齐 830063） 摘要：为了更准确的对忙时话务量进行预测，在考虑多因素条件下提出一种基于经验模态分解(EMD)的高斯过程和灰色预测的组合预测模型。首先对影响话务量的多因素进行相关性分析，提取出最有影响力的关键因素。然后用经验模态分解法把话务量数据在时域上分解成具有不同频率特征的多个分量。把本征模函数（IMF）分量分别和关键因素作为输入，用高斯过程进行预测，趋势分量用灰色预测方法进行预测，然后把各预测结果叠加，得到话务量预测值。通过对收集的话务量数据进行仿真实验，验证了该算法在预测话务量方面具有预测精度高，实现较容易等优越性。 关键词：话务量预测；多因素；经验模态分解；高斯过程；灰色预测；组合模型 中图分类号：TP393 文献标识码：A 文章编号： Forecasting of telephone traffic based on EMD and Gaussian processray prediction SONG Xiuxiu1, JIA Zhenhong1*, QIN Xizhong1，CAO Chuanling2, NIU Hongmei2 School of Information Science and Engineering, Xinjiang University, Urumqi 830046, China; 2. Subsidiary company of China Mobile in Xinjiang, Urumqi 830063, China) Abstract: To improve the prediction accuracy of busy telephone traffic, this paper proposes a combined forecasting model which takes the influence of multiple factors into consideration and combines Empirical Mode Decomposition and Gaussian process model and gray prediction model. Correlation analysis is firstly applied to the busy telephone traffic data to obtain the key factors which influence the busy telephone traffic. Then Empirical Mode Decomposition is used to decompose the telephone traffic data in time domain to get the components with different frequency. The IMF component and the obtained key factors are loaded into Gaussian process model to predict, while the trend component is loaded into gray prediction model to predict, finally the forecasting result is achieved by the superposition of each predictive values. Through the simulation experiments of telephone traffic data collected in practice. The simulation results show that the proposed model has the superiority of higher prediction accuracy and easier to implement. Keywords: Forecasting of telephone traffic; multiple factors; EMD; Gaussian process; gray prediction; combined model 0 引 言 据最新数据统计，中国移动通信的用户量已经突破15亿。随着中国4G时代的到来，今后移动信也会大幅度的增加。在移动通信