关于气象数据检测质量控制仿真

Quality Control Simulation of Meteorological Data Detection

针对传统气象数据质量控制算法存在的不足,首先提出了一种新的基于兴趣度模型的关联规则挖掘算法,通过真实数据验证表明,新算法不仅能够挖掘出所有相关性很强的规则,同时与同类非Apriori类的算法相比,在时间性能上更加优越。接着通过该关联规则算法挖掘历史气象数据提取出所有关联项对形成范例库,并以此构建气象数据质量控制模型,同时通过测试验证与传统质量控制方法对比,发现基于新的关联规则算法的质量控制模型在检出率和灵敏度以及性能方面得到极大地提高,非常具有可行性。

Aiming at the shortcomings of traditional meteorological data quality control algorithm, a new associa- tion rule mining algorithm b^ed on interestingness is now proposed in the paper. The comparison of experimental data sets with real ones show that the new algorithm is not only able to find all the relevant rules but also more superior to the non Apriori algorithm. Then the new algorithm was adopted to mine the historical meteor＇ological data to extract the correlation pairs to form a sample database, and construct the model of the meteorological data quality control. By comparing with the traditional quality control method, it is found that the quality control model based on the new algorithm of association rules can greatly improve detection rate, sensitivity and even time performance. So it is very feasible.