基于物联网和PCA支持向量机的交通流量预测系统

Prediction System for Traffic Flow Based on Internet of Things and PCA Support Vector Machine

摘要:

为了解决已有交通流量监测系统存在的数据采集分散、车辆识别度低、实时性差和流量预测误差大等问题,设计了一种基于物联网技术和最小二乘支持向量机(Least Squares Support Vector Machine, LSSVM)的交通流量预测系统；首先,描述了系统原理和部署模型,然后对系统的硬件即车载传感器节点和Sink节点进行了设计,同时对系统的软件流程进行了描述,通过在监控中心执行PCA主成分分析方法实现对采集数据提取独立主成分,消除无关冗余数据,在此基础上采用LSSVM实现道路交流流量预测；最后,在十字路口布置实验环境,实验结果表明:文章方法能实时精确地实现交通流量预测,与其它方法相比,具有拟合精度高和的泛化能力强的优点,具有很强的实用性。

Abstract:

In order to solve the given traffic flow monitoring system existing the problems such as data collection dispersion, low vehicle identification, low in-time performance and big prediction error, a traffic flow prediction based on Internet of things and LSSVM (Least Squares Support Vector Machine) was proposed. Firstly, the principle and deployment model of system was described. Then the hardware of system includes vehicle sensor and Sink node was designed, and the system software was also introduced. The monitoring center executed the PCA method to extract the independent main information for the rude data, and then the LSSVM was operated to predict the traffic flow for the next time. Finally, the method in this paper was simulated in the environment of crossing road, and the simulation result shows:the proposed method can in time and accurately predict the traffic flow, and compared with other methods, it has the advantages of high fitting accuracy and generalizing ability. Therefore, it has big practicability.