基于优化SVM的路况预测分析框架  

Forecasting Framework of Urban Traffic Based on Optimized Support Vector Machine

作　　者：高凤娇[1] 吴冈[1] 田晓英[1] 杨洋[1] 

机构地区：[1]黑龙江省科学院自动化研究所,哈尔滨150090

出　　处：《黑龙江科学》2012年第6期49-50,75,共3页Heilongjiang Science

摘　　要：城市路况分析对于城市交通安全、交通疏导有着极为重要的意义。本研究首次提出城市路况预测模型及相应的多路段通行分析框架，采用基于支持向量机的智能学习技术实现城市分时多路段路况建模，采用基于粒子群算法优化预测模型，提高模型可靠性，进而利用该模型进行实时路况分析，实现通行时间的精确估计和预测。Automation Urban traffic analysis for city traffic safety, traffic channel has great significance. The paper was the first to propose city traffic prediction model and corresponding sections of traffic analysis technique. For time - sharing muhi - roads （TSMR） modeling,intelligence learning technology based on support vector machine （SVM） is utilized. Particle swarm optimiza- tion （PSO） algorithm is employed for optimizing the models and improving the reliability. With TSMR model, traffic situation could be analyzed, and the passage time is able to accurately estimate and predict.

关 键 词：路况分析 支持向量机 粒子群算法 

分 类 号：TD872[矿业工程—非金属矿开采]