采用自回归移动平均(ARMA)模型、趋势拟合/ARMA组合模型、与小波多分辨率分析结合的WARMA模型,对废弃采空区残余沉降观测数据进行数学建模,并对3种预测模型的预测结果作比较分析。结果表明:3种预测模型均可用于残余沉降短期预测,其中,WARMA模型的预测精度、预测步长和稳定性明显优于其它2个模型,具有实际应用价值。
The observation data modeling for the residual subsidence in abandoned mine goafs by auto regression moving average(ARMA) model,trend regression ARMA model and WARMA model of combining with wavelet multi-resolution analysis is built,and the corresponding prediction results are analyzed respectively.The results are shown as follows: all of three models can be used for short-term prediction of residual subsidence;In contrast with other two models,WARMA model is more precise in prediction and superior in predicting step and stability.Hence,it has a practical value in application.