弓长岭露天矿采空区顶板位移超前预测算法

Roof displacement prediction algorithm of goaf in Gongchangling Open-pit Mine

针对弓长岭露天矿采空区顶板位移预测问题,基于K-means聚类算法基本原理,提出层次迭代聚类算法,并结合依拉达准则,实现了对实时采集数据中的杂乱数据和随机误差数据的自动化剔除。应用灰色预测模型对位移-时间序列进行超前预测,预测结果与实测数据基本吻合,验证了本文算法的可靠性和准确性。根据预测结果可以判定弓长岭露天矿采空区顶板位移变化趋势平稳,短期内没有失稳风险。

Aiming at the prediction of goaf roof displacement in Gongchangling Open-pit Mine,based on the basic principle of K-means clustering algorithm,the hierarchical iterative clustering algorithm is proposed,and the Puata criterion is combined to automatically eliminate the clutter data and random error data in real-time acquisition data.Then the gray prediction model is used to predict displacement-time series.The predicted results are in good agreement with the measured data,which verifies the reliability and correctness of the proposed algorithm.According to the prediction results,it can be justified that the displacement variation of goaf roof in Gongchangling Open-pit Mine is stable and there is no risk of instability in the short term.