基于移动互联网技术的出行模式识别方法

Transportation mode recognition method based on mobile internet technology

针对智能交通中出行模式识别问题,提出一种基于移动互联网技术的出行模式识别方法。通过自主开发的智能终端应用,完成对出行模式位置特征参数、运动参数的实时采集;对采集到的用户出行参数进行分析,选择具有辨识度高的特征;提出一种改进的BP神经网络算法(adaptive learning rate back propagation,ALBP)对出行数据进行训练并识别分类。实验结果表明,该方法能够有效采集出行数据,所提算法提高了出行方式的识别率。

To address the problem of transportation modes recognition in intelligent transportation system, a method based on mobile internet technology was proposed to figure out different transportation modes. Location parameters and motion parame- ters of different transportation modes were collected using developed intelligent terminal applications. Significant mode features were chosen from the collected data. An enhanced back propagation algorithm ALBP （adaptive learning rate back propagation） was proposed to recognize different transportation modes. Experimental results show the proposed method can efficiently collect transportation data and the ALBP algorithm improves the recognition quality.