基于K-means聚类的沙尘天气快速识别技术研究

依据沙尘天气判定条件分析2018年春季（2—4月）陕西省西安市环境空气质量小时监测数据,共识别9 d受沙尘影响。研究基于K-means聚类沙尘天气分析方法,分析沙尘天气监测数据特征,并对沙尘识别聚类模型进行优化研究,识别的沙尘影响天气与传统方法一致。分析表明,聚类方法可用于沙尘天气监测数据的识别,与传统方法相比较,基于K-means聚类方法能够快速、准确识别沙尘影响天气。

Research on the Fast Recognition Technology of Sand Dust Weather Based on K-means Clustering

According to judging conditions of sand dust weather,the ambient air quality hourly monitoring data in Xi''an,Shaanxi Province in the spring of 2018 (February to April) is analyzed,which identifies 9 days affected by sand dust weather.Based on the K-means clustering method for sand dust weather analysis,the characteristics of sand dust monitoring data are analyzed,and the sand dust identification clustering model is optimized.The identified sand dust weather is consistent with sand dust weather recognition by using the traditional method.The analysis shows that the clustering method can be used to identify sand dust weather monitoring data.Compared with the traditional method,the method based on K-means clustering is faster and more accurately of identifying the sand dust weather.