高速公路气象数据融合滤波与短时雨雾预测

Highway Weather Data Fusion Filtering and Short-term Rain and Fog Prediction

摘要:

高速公路路域内雨雾气象数据呈现多源、异构特性，对其进行短时精准预测存在很大难度。针对该问题，在利用联合概率法求得数据之间联合概率的基础上构建异构数据融合模型，并将该融合模型与卡尔曼滤波方法相结合，建立面向多源异构气象数据的协同融合滤波模型。在此基础上，利用贝叶斯最大熵方法，结合雨雾经验理论、融合滤波后的数据以及原始雨雾数据，实现了对高速公路目标路段雨雾天气的短时精准预测。实验结果表明，该方法能够为用户提供精准、稳定的高速公路短时雨雾气象预测结果，对减少交通事故，合理进行交通管制具有重要意义。

Abstract:

The rain and fog weather data in the highway area presents multi-source and heterogeneous characteristics, and it is very difficult to make short-term predictions accurately. Aiming at this problem, a heterogeneous data fusion model is constructed based on the joint probability of data obtained by the joint probability method, and the fusion model is combined with the Kalman filter method to establish a collaborative fusion filter for multi-source heterogeneous meteorological data model to fuse meteorological data. On this basis, using the Bayesian Maximum Entropy (BME) method, combined with the rain and fog empirical theory, fusion filtered data and original rain and fog data, a short-term accurate prediction of the rain and fog weather conditions of the target section of the highway is realized. Experimental results show that this method can provide accurate and stable expressway short-term rain and fog weather forecast results, which is of great significance for reducing traffic accidents and reasonably implementing traffic control.