临近空间通常指距离地球表面20~200km高度的地球空间，是国家安全空间战略新高地。数据同化是一种最初来源于数值天气预报的技术，旨在为数值天气预报提供初始场的数据处理技术，已广泛应用于大气海洋领域和其他地球科学领域。对临近空间数据做质量控制是同化的关键一步，质量控制找出数据的离群点并加以处理，使其更符合物理意义。本文基于一种双权重的数据质量控制方法，该方法通过引入双权重平均值和双权重标准差BSTD，对TIMED数据的质量和控制规则的有效性进行加权，以实现对数据质量的更精确控制。首先回顾了临近空间天气变分同化数据质量控制的研究背景、原理理论，然后详细介绍了双权重数据质量控制方法的设计和实现过程，最后通过研究验证该方法的有效性和优越性。结果表明，使拟合正态分布的效果达到最好，不同高度层上对应的z-score不同。研究结果可促进临近空间数据同化技术的发展，以提高模型预报的准确性。

Near space typically refers to the Earth's space with an altitude of 20 to 200 km, which is a new high ground in national security space strategy. Data assimilation is a technology originally derived from numerical weather forecasting, aiming to provide data processing techniques for numerical weather forecasting and has been widely applied in atmospheric and oceanic fields and other earth sciences. Quality control of near-space data is a crucial step in data assimilation, which identifies outliers in the data and processes them to make them more physically meaningful. This paper introduces a double-weighted quality control method based on TIMED data, which uses double-weighted mean value and double-weighted standard deviation BSTD to weight the quality and effectiveness of control rules, thus achieving more precise control of data quality. First, the background, theoretical basis, and research progress of near-space weather data assimilation quality control are reviewed, and then the design and implementation process of the double-weighted quality control method is detailedly introduced. Finally, through research verification, the effectiveness and superiority of the method are demonstrated. The results show that the best effect can be achieved by fitting the normal distribution, and the corresponding z-score varies with the height layer. The research results can promote the development of near-space data assimilation technology, thereby improving the accuracy of model forecasts.