基于小波核的LS-SVM算法在城市交通流预测中与其他核函数对比研究

LS-SVM Algorithm Based on Wavelet Nuclear Research Compared with Other Nuclear Function in Urban Traffic Flow Forecasting

针对城市交通流短时流量预测在智能交通系统中的重要性,在以往研究基础上采用Morlet函数作为小波核函数,进一步提高了模型的预测精度及泛化能力.将其与其他几种常用核函数模型进行比较,其效果明显优于其他核函数模型,能够满足智能交通控制和诱导的要求.

For short term urban traffic flow projections, the importance of the intelligent transportation system, on the basis of previous studies using Morlet function as kernel function, and further improving the prediction accuracy and generalization ability, with several other commonly used kernel function model, the effect is significantly better than other nuclear function model, and be able to meet the requirements of intelligent traffic control and induction.