基于非参数回归的短时交通流预测研究综述

短时交通流预测是实现交通控制和诱导的关键问题之一。综述了基于非参数回归的短时交通流预测方法，指出了状态向量的选取没有考虑天气环境等存在的问题，提出了改进思路和方法，即基于动态聚类和决策树的历史数据组织方式、时空一天气环境相结合的状态向量选取方法以及基于密集度和状态向量的自适应变K机制等，期望通过这些改进能提高基于非参数回归短时交通流的预测精度，为交通控制和交通诱导建立基础。

Review of the Short-term Traffic Flow Forecasting Based on the Non-parametric Regression

Short-term traffic flow forecasting is one of the key problems of traffic control and traffic flow guidance.Having reviewed the short-term traffic flow forecasting methods based on non-parametric regression,some shortages,for example,weather environment is not considered as a component of state vector,etc.,were pointed out.Some improved methods,such as the data organization method based on the dynamic cluster and decision tree,the state vector based on the space-time and weather environment,and the K mechan...