Arabian Journal of Geosciences ( IF 1.327 ) Pub Date : 2021-02-22 , DOI: 10.1007/s12517-021-06647-w Utku Sakız, Gulhan Ustabas Kaya, Olgay Yaralı
Rock drillability characteristic is one of the important properties for mining and tunneling operations. The rock drillability can be determined by using the drilling rate index (DRI) for engineering applications. The present study attempts to develop a practical and convenient DRI estimation model by using rock strength and abrasivity properties. For this purpose, fuzzy inference system (FIS) being an accurate prediction model was applied to predict DRI by using experimental data obtained with 37 different rocks. The predictive FIS based on experts knowledge by taking mechanical and abrasivity properties as input parameters was created on MATLAB. This structure was carried out by using Mamdani extraction method. DRI values obtained experimentally and estimated from the FIS model were compared. This comparison is given with statistically reliable (R2=0.9277) results. In order to prove the validity of the FIS model for DRI prediction, a validation process has been performed by using test data as well. The performance determination coefficients (R2) are found as 0.9513 by using test data. As a result, it was found that DRI values can be predicted very efficiently and accurately with the proposed prediction method.
岩石可钻性是采矿和隧道作业的重要属性之一。可以通过在工程应用中使用钻速指数（DRI）来确定岩石的可钻性。本研究试图通过利用岩石强度和耐磨性来开发一种实用且方便的DRI估算模型。为此，通过使用在37种不同岩石上获得的实验数据，将作为准确预测模型的模糊推理系统（FIS）应用于DRI预测。在MATLAB上创建了基于专家知识的预测性FIS，该预测性FIS将机械和磨蚀性作为输入参数。该结构通过使用Mamdani提取方法进行。比较了实验获得的DRI值和从FIS模型估算的DRI值。这种比较在统计上可靠（R 2 = 0.9277）结果。为了证明FIS模型对于DRI预测的有效性，还通过使用测试数据执行了验证过程。使用测试数据发现性能确定系数（R 2）为0.9513。结果，发现利用所提出的预测方法可以非常有效且准确地预测DRI值。