基于改进RBF神经网络模型的气象能见度监测方法研究  

Research on Meteorological Visibility Monitoring Method Based on Improved RBF Neural Network Model

作　　者：袁超[1] YUAN Chao(Taian Meteorological Bureau,Shandong TaiAn 271000,China)

机构地区：[1]泰安市气象局,山东泰安271000

出　　处：《新一代信息技术》2019年第16期50-54,65,共6页New Generation of Information Technology

基　　金：国家自然科学基金“深部流变岩体巷道破坏机理与支护”(项目编号:51804182)。

摘　　要：气象能见度实时监测是确保气象预报贴近现实气象情况的重要组成部分,是高精确度气候预测技术手段中必不可少的环节。目前主要采用的方法为模糊识别和BP神经网络分析,在气象能见度监测方面效果并不理想。为此,提出了改进RBF神经网络模型的气象能见度监测方法,综合考虑了SO2,NO2两种气体,以及气温,气压,温度,湿度等因素采集方法,计算各因素的权重和相关系数后,在改进RBF神经网络模型的基础上确立了具体的监测方法,最后检验了两种气象能见度监测方法的准确性,新的能见度监测方法具有更高的准确性,在气象监测方面具有应用前景。Real-time monitoring of meteorological visibility is an important part of ensuring that weather forecasts are close to real-world meteorological conditions,and is an indispensable link in high-precision climate prediction techniques.At present,the main methods used are fuzzy recognition and BP neural network analysis,and the effect on meteorological visibility monitoring is not satisfactory.To this end,a meteorological visibility monitoring method based on improved RBF neural network model is proposed.The two methods of SO2 and NO2,as well as temperature,pressure,temperature,humidity and other factors are collected.After calculating the weights and correlation coefficients of each factor,Based on the improved RBF neural network model,the specific monitoring methods are established.Finally,the accuracy of the two meteorological visibility monitoring methods is tested.The new visibility monitoring method has higher accuracy and has application prospects in meteorological monitoring.

关 键 词：改进RBF神经网络模型 气象能见度监测 气象能见度影响因素 

分 类 号：P412.17[天文地球—大气科学及气象学]