计算机工程与应用 ›› 2022, Vol. 58 ›› Issue (11): 279-286.DOI: 10.3778/j.issn.1002-8331.2106-0305

• 工程与应用 • 上一篇    下一篇

基于CEEMDAN分解的短时交通流组合预测

聂铃,张剑,胡茂政   

  1. 上海工程技术大学 航空运输学院,上海 201620
  • 出版日期:2022-06-01 发布日期:2022-06-01

Short-Term Traffic Flow Combination Prediction Based on CEEMDAN Decomposition

NIE Ling, ZHANG Jian, HU Maozheng   

  1. School of Air Transportation and Flying, Shanghai University of Engineering Science, Shanghai 201620, China
  • Online:2022-06-01 Published:2022-06-01

摘要: 短时交通流预测是实现交通流诱导与控制的重要保障,鉴于交通流的随机性和复杂性,提出基于自适应噪声完全集合经验模态分解(CEEMDAN)的短时交通流组合预测模型。利用CEEMDAN算法对非线性序列具有自适应分解的特性,将交通流时间序列通过CEEMDAN分解为频率不同、复杂度不同的多个时间序列分量;利用PE算法分析各个分量的随机特性,根据时间序列分量的不同随机特性分为高频序列分量、中频序列分量和低频序列分量,根据高频、中频和低频序列分量的随机特性分别建立GWO-BP模型、GWO-LSSVM模型和ARIMA模型进行预测;叠加高频、中频和低频各个分量的预测结果,得到短时交通流最终预测值。仿真分析结果表明,与其他预测模型相比,基于CEEMDAN分解的短时交通流组合预测模型提升了预测精度。

关键词: 短时交通流, 组合预测, 排列熵, 经验模态分解

Abstract: Short-term traffic flow prediction is an important guarantee for traffic flow guidance and control. In view of the randomness and complexity of traffic flow, a combined short-term traffic flow prediction model based on complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN) is proposed. Firstly, the CEEMDAN algorithm is used to decompose the nonlinear sequence adaptatively, and the traffic flow time series is decomposed into multiple time series components with different frequencies and complexity by CEEMDAN algorithm. Secondly, PE algorithm is used to analyze the random characteristics of each component. According to the different random characteristics of time series components, they are divided into high frequency, intermediate and low frequency sequence components. According to the random characteristics, GWO-BP model, GWO-LSSVM model and ARIMA model are established respectively. Finally, the prediction results of high frequency, intermediate and low frequency components are superimposed to obtain the final predicted value of short-term traffic flow. The simulation results show that compared with other prediction models, the combined prediction model based on CEEMDAN decomposition improves the prediction accuracy.

Key words: short-time traffic flow, combination prediction, permutation entropy, empirical mode decomposition