CAJ | 학술논문

　　话务量是一种受多重因素影响的动态非线性时间序列，采用传统的预测方法难以实现高精度的预测，为了提高预测精度，文章根据采集所得每日忙时话务量数据，分析其混沌特性和可预测性，并重构话务量时间系列的相空间，提出采用加权一阶局域预测方法来建立其预测模型，并进一步预测，仿真结果表明，该方法能够较好的拟合话务量的变化规律，具有较高的精度，为话务量的预测提供了一种有效的途径。
　　화무량시일충수다중인소영향적동태비선성시간서렬，채용전통적예측방법난이실현고정도적예측，위료제고예측정도，문장근거채집소득매일망시화무량수거，분석기혼돈특성화가예측성，병중구화무량시간계렬적상공간，제출채용가권일계국역예측방법래건립기예측모형，병진일보예측，방진결과표명，해방법능구교호적의합화무량적변화규률，구유교고적정도，위화무량적예측제공료일충유효적도경。
Traffic is a kind of dynamic nonlinear time series affected by various factors; the traditional predication methods can’t achieve higher accuracy, In order to improve its forecasting accuracy, In this paper, analyzing chaotic characteristics and predictability of traffic based on the data of daily rush-hour by collecting, and reconstructing the phase space, propose a method of building predication model by using weighed one-step local-region prediction method, and use the model to one-step forecast, the experimental results show that this method can better fit the change of traffic and has higher accuracy . It can provide an effective way for forecasting of traffic.