引入支路流量的非参数回归交通流短时预测资料精.pdfVIP

引入支路流量的非参数回归交通流短时预测 杨珍 ，王慧 (浙江大学 系统工程研究所 智能交通研究中心 杭州 310027) 摘 要：先进的交通信息系统和先进交通管理系统对交通流短时预测提出了越来越高的要求，针对目前交通流短时预 测存在的问题，提出了一种引入支路流量的非参数回归城市交通流短时预测方法。在传统非参数回归方法中构建一 个支路流量突变因子，在发生由支路流量引起并且历史样本库中不存在的突发事件时，结合非参数回归预测和支路 流量突变因子准确预测干路的未来交通流量，并利用最小二乘法确定历史数据和当前支路流量数据的权值。经仿真 验证，该方法在历史样本库不足的情况下有很好的预测效果。 关键字：交通流短时预测；非参数回归；支路流量 Short-term Traffic Flow Forecasting Based on Non-Parametric Regression with Branch Flow Yang Zhen, Wang Hui (Research Center of Intelligent Transportation System, Institute of System Engineering, Zhejiang University, Hangzhou 310027, China) Abstract: Advanced Traffic Information System and Advanced Traffic Management System require precise short-term traffic flow forecasting methods. This article concluded the problems that existed in the field of short-term traffic flow forecasting and proposed a new method based on non-parametric regression with branch flow. A jumping factor was introduced to the conventional non-parametric method so that when sudden event caused by branch flow happened, whose neighborhoods cannot be found in the historical database, precise forecasting can still be reached. Least Square method was used to estimate the weight values for historical data and the jumping factor. It was validated that the method based on non-parametric regression with branch flow had more precise forecasting ability then the conventional non-parametric method when the historical database was deficient. Keywords: Short-term Traffic Flow Forecasting; Non-Parametric Regression; Branch Flow 1 概述 交通诱导是智能交通系统(Intelligent Transportation System, ITS)中一个非常重要的环节，其核心 工作是通过实时检测的交通信息数据传送到先进的交通信息系统和先进交通管理系统提供出行者合 理