基于WOA-BP神经网络的岩石可钻性预测
Prediction of Rock Drillability Based on WOA-BP Neural Network

岩石可钻性指岩石抵抗钻凿破碎的能力，准确评价可钻性有利于动态调整钻井参数，是实现高效钻井的关键。利用数据清洗、小波滤波降噪、归一化方法处理数据，通过相关性分析优选模型输入参数。鲸鱼算法优化BP神经网络，构建WOA-BP神经网络模型预测岩石可钻性，对比评价BP模型、GA-BP模型、WOA-BP模型的训练精度和预测精度。结果表明：经优化后的BP神经网络模型避免了传统BP神经网络易陷入局部最优的问题，提高了模型的预测精度和收敛速度；WOA-BP模型的训练精度最高，误差小于10%的数据占比为75.38%，相较于BP模型、GA-BP模型分别提高13.1%和4.86%，预测效果较好。
Drillability of rock refers to the ability of rock to resist drilling and crushing. Accurate evaluation of drillability is conducive to dynamic adjustment of drilling parameters and is the key to realizing efficient drilling. The data were processed by data cleaning, wavelet filtering and normalization, and the input parameters of the model were optimized by correlation analysis. The whale algorithm optimizes BP neural network, constructs WOA-BP neural network model to predict rock drillability, and compares and evaluates the training accuracy and prediction accuracy of BP model, GA-BP model and WOA-BP model. The results show that the optimized BP neural network model avoids the problem that the traditional BP neural network is easy to fall into the local optimal, and improves the prediction accuracy and convergence speed of the model. The training accuracy of WOA-BP model is the highest, and the proportion of data with error less than 10% is 75.38%, which is 13.1% and 4.86% higher than that of BP model and GA-BP model, respectively, and the prediction effect is better.