矿坑涌水对大气降水的响应分析及预测

分析了程潮铁矿矿坑涌水对大气降水的响应特征,通过大气降水-矿坑涌水互相关函数值,研究了不同滞后期大气降水对矿坑涌水的影响程度。建立了基于时间序列分析的大气降水-矿坑涌水传递函数模型的预测方法。利用最小二乘法对传递函数部分进行参数估计,并对残差序列进行自回归处理,平均误差低于3.26%。预测模型在一定程度上能够反映出矿坑涌水对大气降水的滞后和相关关系,适用于采用崩落法或上覆围岩裂隙破碎带发育的地下矿山。

Response analysis and forecasting of the pit-water due to the rainfall

This paper is aimed to present our analysis of the correlation between the pit water and the rainfall in Chengchao Iron Mine,and in turn to introduce a scientific and effective method developed by the author for forecasting the pit water in the underground mines.As is known,rainfall is the main factor of mining water filled in the most underground mines in China,which makes it necessary to forecast the pit water situation so as to guarantee the regular mining activities.For this purpose,we have done the analysis of the response features of the pit water to the rainfall in the above said mine.Calculating the cross-correlation function of the rainfall sequence and the pit water consequence,we have studied the influential degree of the rainfall on the different intensities of the water in the different lag periods.So far as we have investigated,in the period from March to June,the pit water has been affected by the rainfall for 6 days from the 1st day to the 5th day in March and April,while 24 days from the 2nd day to the 25th day in July and August.Based on the time consequence analysis discussed above,we have put forward a forecasting method for establishing a transfer function model on the rainfall-pit water in this paper,whereas the only information we know was the data of the rainfall and pit water,with the groundwater system regarded as "the grey box model".In such a situation,we can simplify the system of the rainfall recharge of the pit water to a linear system with a single input and single output and determine the order of function-transferring model by cross-correlation function and features of pit water series.After preliminary assessment of the forecasting model with the least square method and autoregressive disposal for residual series,we have achieved a high forecasting accuracy with the average error being lower than 3.26%.Thus,the forecasting method we have gained can not only reflect the lag and correlation relationship between the rainfall and the pit water,but also provide us with a simplified approach with a few parameters suitable for the underground caving mining situations or the mines with cracked crush zone in overlaying rocks.