基于半参数模型的水位数据预处理方法研究

Research on Tidal Data Pre-processing Method Based on Semi-parametric Model

为消除压力式验潮仪数据因误差而存在的抖动现象,提出了利用半参数模型对水位数据进行预处理的方法。研究了半参数模型在水位预处理中的具体实现方法,并对处理后的数据在应用方面进行了数据实验。实验表明,半参数模型既能够剔除不合理观测噪点,又能够高精度拟合观测水位,利用处理后的数据确定的余水位以及潮汐比较参数更具备合理性,精度得到了提高。

To solve the problem that observation data may exist fluctuation because of weather influence and water vibration,the model o f semi-parametric is used to smoothen the water level observation registered by pressure-type tide gauge.And some data experiments are performed to evaluate the fitting precision of this method.The results of experiments show that the model of semi-parametric is effective to remove system error and make hydrograph smoother and is more reasonable.The conclusion is that the precision of residual water level and tidal comparison parameters estimated by smoothed data is improved and semi-parametric model is useful for pre-processing of tidal data.