CAJ | 학술논문

通过交通量预测的方法,可以在突发事件情况下评估事件对高速公路网的交通影响.各种预测算法的共同特征,需要一定的历史数据支持.在同一算法中,对历史信息吸收得越多,其预测往往越准确.作者借用图像压缩的原理,从历史交通数据中抽象出历史交通状态演化模式,并用压缩码表示该演化模式.通过对压缩码的检索,可以快速提取与当前演化模式相近的历史数据,通过这些演化模式相近的历史数据,预测交通状态的变化趋势.试验证明,预测数据与真实数据具有良好的拟合度.
통과교통량예측적방법,가이재돌발사건정황하평고사건대고속공로망적교통영향.각충예측산법적공동특정,수요일정적역사수거지지.재동일산법중,대역사신식흡수득월다,기예측왕왕월준학.작자차용도상압축적원리,종역사교통수거중추상출역사교통상태연화모식,병용압축마표시해연화모식.통과대압축마적검색,가이쾌속제취여당전연화모식상근적역사수거,통과저사연화모식상근적역사수거,예측교통상태적변화추세.시험증명,예측수거여진실수거구유량호적의합도.
The effect of an accident occurred on expressway on the expressway net can be estimated by the method of traffic volume forecasting. The common characteristic of forecast algorithms needs history data support,and an algorithm always forecasts more accurate by absorbing more history data. Referred to the theory of image compression,the history traffic volume evolvement mode was abstracted from history data and the evolvement mode was expressed by compressed code. History data similar to current traffic volume evolving can be selected by searching similar compressed code and traffic volume trend can be forecasted by integrating these history data which has similar evolvement mode. Factual experiment shows that the forecasted data has good coherence with real data.