D-InSAR技术和支持向量机算法在矿山开采沉陷监测与预计中的应用可行性研究

Studyonthe Feasible Applicationof D-InSAR Technology and Support Vector Machine Algorithms in Mining Subsidence Monitoring and Forecasting

鉴于传统的矿山沉降监测方法精度低、效率低且稳定差等弊病,本文将D-InSAR与支持向量机算法进行有机统一,从而形成一种新的矿山沉降监测方法。首先介绍了D-InSAR技术和支持向量机算法的原理,然后对重庆某矿山的测量结果进行评价。结果表明:D-InSAR技术具有监测区域面积大、靠性高、长期稳定性好、监测精度高、可以准确反映沉陷变化的动态过程等特点,而基于SVM算法建立的沉陷预计模型具有所需样本小、预计精度高等特点。将两者结合,可实现对矿山开采过程的总体、系统把握。在实践中该方案可以实现沉陷监测和预计的一体化功能,可以应用于矿山开采沉陷监测与预计中。

Considering there were disadvantages of the low accuracy, low efficiency and poor stability in traditional mine subsidence monitoring methods, this paper combined D-InSAR and Support Vector Machine(SVM) Algorithm to form a new mine subsidence monitoring method. In the article, the TWO basic principles were introduced and then to assess their application effect in a mine of Chongqing. The results showed that D-InSAR technology had large monitoring area, high reliability and long-term stability. The monitoring accuracy was high, and the dynamic process of subsidence change could be accurately reflected. The subsidence prediction model based on SVM algorithm had the characteristics of small sample size and high prediction accuracy. Combining the two could achieve an overall and systematic grasp of the mining process. Facts had proved that this scheme could realize the integrated function of subsidence monitoring and prediction, and be applied in mine subsidence monitoring and prediction.