赵各庄矿地下水位时间序列数字滤波分析

Digital Filter Analysis of Time Series of Groundwater Level in Zhaogezhuang Coalmine

摘要:

为了更好地了解煤矿地下水位的变化趋势，结合工程实例，运用递归和非递归2种滤波技术进行分析，通过对比2种滤波技术的滤波效果，可以得到地下水位的变化规律和成因机理。结果表明：滤波系数的调校对滤波效果的影响至关重要。根据平滑度指标、均方根误差和信噪比对现有观测数据进行计算的结果可知，当b=0.5时递归滤波效果相对较好，当q=5时非递归滤波效果相对较好。滤波结果可以作为预测未来该地区地下水位变化趋势的依据，也可以为安全生产提供参考。

Abstract:

In order to understand the changing trend of groundwater level in a coal mine, combining with factual examples, and contrasting the results gained by using recursive filtering method and non-recursive filtering method respectively, the changing rule and the mechanism of groundwater level have been studied. It is showed that the adjustment of filtering coefficients will effect filtering results. According to the smoothness index, the root mean square error and the signal-noise ratio, and comparing with the existing observation time series of groundwater level, it is concluded that the recursive filtering effect is better when b=0.5, while the non-recursive filtering effect is proper when q=5. The filtering results can be used as the basis for predicting the changing trend of groundwater level in the future. It also can provide some references for the safe production in the coal mine.