In the process of coal mine roadway excavation, the tunneling path of Tunnel Boring Machine (TBM) may deviate from the predetermined axis due to the uncertainty and complexity of the environment. In order to improve the accuracy of TBM pose prediction, this paper proposes a TBM tunneling pose prediction model based on optimized deep learning algorithm. The collected TBM tunneling parameters are extracted and cleaned, and the spatial features of the data are extracted by convolutional neural network (CNN). Bidirectional long short-term memory network (BiLSTM) is used to learn the time dependence of the data. In order to further optimize the performance of the model, the Sparrow Optimization Algorithm with Sine Cosine and Cauchy Mutation (SCSSA) is used to optimize the hyperparameters of the CNN-BiLSTM model. The results show that the optimized model has a significant improvement in multiple error indicators, including mean absolute error (MAE), mean absolute percentage error (MAPE), mean square error (MSE) and root mean square error (RMSE). At the same time, the performance indicators such as correlation coefficient (R) and coefficient of determination (R2) are also improved. Specifically, when the SCSSA-CNN-BiLSTM model predicts the pose parameters such as rolling angle, pitch angle, azimuth angle and horizontal deviation of TBM tunneling, the determination coefficient R2 reaches 0.9982,0.9944,0.9936 and 0.9865, respectively, which verifies the efficiency and accuracy of the model in the prediction of TBM tunneling pose in coal mine roadway.