多点季节性变形监测资料分析和预报方法研究

Analysis and Forecast Methods of the Series Seasonal Deformation Monitoring Data

变形监测预报一直是变形测量的难点和重点,其结果的好坏直接影响人类的生产和生活。特别是复杂的季节性变化,往往给预报结果带来了波动性。提出了一种综合滑动平均去季节性和SCGM(1,m)模型的变形监测预报方法。该方法顾及了多点之间的空间关系,也兼顾了由季节性引起的时间变化,因此提高了预报的准确性。最后用美国南加州GPS网(SCIGN)提供的实测数据验证了该方法的可靠性。

The forecast of transforms monitoring is always the difficulty of deformation measuring. Its results directly affect human production and life. Especially, the influences of season always bring fluctuate to the forecasting results. Therefore, a deformation monitoring prediction method of combining the method of moving average seasonal - removal with model was proposed. As considering the change caused by season as well as the special relation between the series data, the forecasting accuracy was improved. At last, the real data provided by SCIGN of the USA were used to valid the reliability of the proposed method.