摘要
解决路面气象状态的识别问题并满足一定的精度要求,是高速公路气象监测研究中的一个关键问题。本文提取、和三种彩色模型信息组成特征向量;提出一种附加动量法和自适应学习率相结合的改进BP算法;在此基础上建立了非线性识别模型,并将该模型应用于路面状态识别试验。实验结果表明该模型能准确有效地识别路面气象状态。
To resolve recognition problem of road surface conditions and satisfy measure accuracy is one key of freeway meteorology monitoring study. Three kinds of,and color models extracted constitute feature vectors. A BP neural network algorithm improved by momentum and self-adaptive learning rate is presented. On the basis of these, a model of the nonlinear system is set up and applied in road surface conditions identification. Experimental results show that the model can identify road surface meteorology more accurately and effectively than that of the traditional BP neural network.
出处
《微计算机信息》
2010年第2期3-4,46,共3页
Control & Automation
基金
科技部科研院所技术开发研究专项资金<气象通量采集系统和路面状态传感器>(NO:2007505)
