随着海洋研究的不断深入,对观测的数据质量要求越来越高,为了有效降低实测数据的不确定性,该文以海水CO2分压测算的关键因子温度和盐度的相关实测数据为例,针对仪器平台能够进行多参数同步观测的特点,在假设观测指标的变化在较小的时间空间范围内为一平稳过程的基础上,通过ADF平稳性检验理论证实了大部分原始观测序列满足二阶差分平稳性假设,在此基础上提出了一种基于观测序列差分统计特征的测量数据不确定性的表达方式,以及对离异点数据进行识别和多参数差分相关联合滤波的算法.与常用滤波算法对比分析表明,提出的新算法能够有效融合参数间的相关信息,降低序列的不确定性,并且最大程度地保护原始测量数据.%With the deepening of the ocean research, more and more demands on the quality of observational data are proposed. In order to effectively reduce the uncertainty of the measured data, the measured temperature and salinity data related to the seawater CO2 partial pressure was taken as an example. As the instrument platform can measure multi-parameter simultaneously, and the measurement indicators can be assumed to change stably in small time and space range. On this basis, most of the original observation sequences were confirmed to meet the hypothesis of second-order differential smoothness by ADF stationary test theory. Then, an expression for observation data uncertainty based on the differential statistical characteristic of observation sequence, and an algorithm of multi-parameter difference correlation Federated Filter based on the outliers recognition, are proposed in this paper. Comparing with the common filtering algorithm, it was found that the new algorithm can effectively integrate the relevant information of parameters, reduce the uncertainty of the sequence, and maximize the protection of the original measurement data.
展开▼
