Steering drilling wellbore trajectory prediction based on the NOA-LSTM-FCNN method

Aiming at the problem that it is difficult to accurately predict wellbore trajectory under complex geological conditions, the NOA-LSTM-FCNN prediction method for steering drilling wellbore trajectory is proposed by combining NOA, LSTM and FCNN. This method adopts LSTM layer to receive input data and capture long-term dependencies within the data, extracting important information. The FCNN layer performs nonlinear mapping on the output of the LSTM layer and further extracts relevant features to enhance prediction accuracy. NOA is employed for hyperparameter optimization of the LSTM-FCNN model. The experimental results show that the prediction effect of the proposed method is better than that of other methods. Taking the prediction results of the well deviation angle of H21 as an example, compared with traditional machine learning methods (LR, SVM and BP) and deep learning methods (CNN, LSTM and GRU), the evaluation index R² of this method was improved by 0.17887, 0.03129, 0.0259, 0.00054, 0.00032 and 0.00031 respectively, showing significant prediction accuracy advantages and strong adaptability. In addition, it applies to various types of wellbore trajectory data, effectively enhancing wellbore trajectory prediction capabilities under complex geological conditions.